UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Forest fragmentation and regrowth : use of riparian and upland forest by birds in managed and unmanaged… Shirley, Susan 2002

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

Item Metadata

Download

Media
831-ubc_2002-750825.pdf [ 9.24MB ]
Metadata
JSON: 831-1.0090748.json
JSON-LD: 831-1.0090748-ld.json
RDF/XML (Pretty): 831-1.0090748-rdf.xml
RDF/JSON: 831-1.0090748-rdf.json
Turtle: 831-1.0090748-turtle.txt
N-Triples: 831-1.0090748-rdf-ntriples.txt
Original Record: 831-1.0090748-source.json
Full Text
831-1.0090748-fulltext.txt
Citation
831-1.0090748.ris

Full Text

FOREST F R A G M E N T A T I O N A N D R E G R O W T H : U S E O F RIPARIAN A N D U P L A N D FOREST B Y BIRDS I N M A N A G E D A N D U N M A N A G E D M A T U R E C O A S T A L BRITISH C O L U M B I A RAINFOREST by S U S A N SHIRLEY B . S c , The University of British Columbia, 1993 A THESIS SUBMITTED IN PARTIAL F U L F I L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF DOCTOR OF PHILOSOPHY in T H E F A C U L T Y OF G R A D U A T E STUDIES (Department of Zoology) We accept this thesis as conforming to the required standard  THE UNIVERSTTY OF BRITISH C O L U M B I A September 2 0 0 2 © Susan S h i r l e y , 2 0 0 2  In  presenting  degree freely  at  this  the  available  copying  of  department publication  of  in  partial  fulfilment  of  the  University  of  British  Columbia,  I  agree  for  this or  thesis  reference  thesis by  this  for  his thesis  and  study.  scholarly  or for  her  financial  Department  DE-6  (2/88)  purposes  Columbia  gain  shall  that  agree  may  representatives.  permission.  T h e U n i v e r s i t y o f British Vancouver, Canada  I further  requirements  It not  be is  that  the  Library  permission  granted  by  understood be  for  allowed  an  advanced  shall for  the that  without  make  it  extensive  head  of  my  copying  or  my  written  Abstract Riparian ecosystems are known for their high diversity, yet they represent one of the most threatened habitats globally. I studied riparian and upland bird communities in undisturbed old-growth forest on western Vancouver Island, B.C. and in riparian buffer strips retained after harvesting. In undisturbed forest, species richness and diversity did not differ over the riparian-upland gradient. Except for riparian specialists and species associated with deciduous vegetation, there was broad overlap of species and abundances between riparian and upland habitats. To test if fragmentation of riparian habitats alters bird communities, I surveyed birds in buffers of varying widths 5-7 years post-harvest. Effects of fragmentation on species richness and overall abundance were weak to absent in these communities, which are dominated by several forest generalists. The few forestinterior species were less abundant in all but the widest buffers and were replaced by open-edge species in narrower buffers. Species assemblages in narrow buffers were also least similar to controls. Species richness and abundance increased in both buffers and adjacent clearcuts during the study period. Because populations able to use the forest matrix may persist better in buffers, I studied movements of birds across river and forestclearcut edges. Forest generalists, open-edge and ubiquitous species crossed riparian and clearcut edges most often. Movements were more frequent across clearcut edges than river edges and were positively related to densities in buffers rather than to buffer width. Movements were highest in narrow buffers, however, suggesting that birds move in and out of narrow buffers because they do not provide suitable habitat to support forestdwelling species. I suggest that narrow buffers function as foraging sites or travel corridors. Finally, I examined patterns of vegetation across varying buffer widths to test if habitat diversity better explains patterns of bird species richness and abundance than buffer width. Densities of faster-growing deciduous trees, and richness and cover of shrubs and forbs differed over the range of buffer widths. While species richness and abundances of several bird guilds and species were explained better by buffer width, three habitat specialist guilds: riparian specialists, forest interior and open-edge species and 4 of 8 individual species were best predicted by density and cover of deciduous vegetation. Deciduous vegetation was more common in wider buffers providing further evidence that retaining wide buffers enhances bird habitat.  iii  Table Of Contents Abstract  ii  Table of Contents  •  iii  List of Tables  vi  List of Figures  vii  Acknowledgements  •  viii  C H A P T E R 1 Introduction 1.1 Introduction  1 1  1.2 Literature Cited  6  C H A P T E R 2 Bird communities of temperate riparian forests 2.1 Introduction  12  2.2 Methods 14 2.2.1 Study area 14 2.2.2 Vegetation sampling 15 2.2.3 Bird sampling 16 2.2.4 Data analysis 16 2.3 Results 19 2.3.1. Vegetation 19 2.3.2. Bird abundance and diversity 20 2.3.3. Habitat associations 21 2.4 Discussion 21 2.4.1 Species diversity and abundance along the riparian gradient 22 2.4.2 Habitat selection 23 2.4.3 Management implications 25 2.5 Literature cited 36 Appendix 1 Classification of bird species into migratory status and foraging guild summarized by habitat guild 43 C H A P T E R 3 Forest Fragmentation And Regrowth: Use Of Riparian Buffer Strips By Birds In Mature Coastal British Columbia Rainforest 45 3.1 Introduction 3.2 Methods 3.2.1 Study areas 3.2.2 Bird sampling 3.2.3 Data analyses 3.3 Results 3.3.1 Bird communities in buffers and controls  45 48 48 49 50 53 53  iv  3.3.2 Species-area relationships and diversity 53 3.3.3 B ird abundance and densit y 54 3.3.4 Bird abundance by guilds 54 3.3.5 Individual species 56 3.3.6 Species turnover and community similarity 56 3.3.7 Changes to bird communities in the clearcut habitat 57 3.4 Discussion 58 3.4.1 Species richness 58 3.4.2 Patterns of abundance among guilds and species 60 3.4.3 Species turnover and community composition 62 3.4.4 Forest mosaics at a landscape level 63 3.5 Conclusions 64 3.6 Literature cited 81 Appendix 2 Species richness and abundance of guilds and species versus width of buffers and controls 94 Appendix 3 Linear regression statistics for guild abundance vs. log-buffer width in clearcut fragments for years 1996-1998 96  CHAPTER 4 Effect Of Buffer Strip Width On The Movement Of Riparian Birds Across River And Clearcut Edges 97 4.1 Introduction 97 4.2 Methods 99 4.2.1 Study area 99 4.2.2 Bird movement frequency 99 4.2.3 Bird density 100 4.2.4 Data analysis 100 4.3 Results 101 4.3.1 Temporal movement patterns over the breeding season 102 4.3.2 Movements across varying buffer widths 102 4.3.3 Comparison of movements across river and clearcut edges 103 4.3.4 Correlations of movements with density 103 4.4 Discussion 103 4.4.1 Effect of buffer width on bird movement frequency 104 4.4.2 Movements across river versus clearcut edges 105 4.4.3 Relationship of movements with density 106 4.4.4 Influence of the clearcut/forest matrix on movements 107 4.4.5 Implications for conservation of forest-dependent species 108 4.5 Literature cited 117 Appendix 4 Bird movements across river edges by species categorized by habitat guild 124 Appendix 5 Bird movements across clearcut edges by species categorized by habitat guild 126 Appendix 6 Summary of bird densities by habitat guild in buffers of varying widths 128  C H A P T E R 5 Influence Of Vegetation Structure On Bird Use Of Riparian Buffer Strips 5.1 Introduction 5.2 Methods 5.2.1 Study area 5.2.2 Vegetation sampling 5.2.3 Habitat associations 5.2.4 Data analyses 5.3 Results 5.3.1 Vegetation patterns 5.3.2 Habitat associations of birds 5.4 Discussion 5.4.1 Changes in vegetation with buffer width 5.4.2 Habitat associations of birds 5.5 Management implications 5.6 Literature cited  C H A P T E R 6 General Discussion 6.1 Are riparian bird communities more diverse than upland bird communities? 6.2 Fragmentation effects on bird communities in buffer strips 6.3 Buffer width and bird movements across habitat boundaries 6.4 Bird/habitat associations in buffer strips 6.5 Management implications 6.6 Conclusions 6.7 Literature cited  129 131 131 131 131 132 133 133 134 136 136 139 142 158  166 166 167 168 169 170 173 175  vi  List of Tables 2.1  Correlations between first three P C A factor scores and vegetation variables for old-growth sites  2.2  27  Paired t-tests comparing vegetation structure characteristics in riparian and upland habitat  2.3  28  Stepwise linear regression models for abundance of guilds and species against vegetation variables  29  3.1  Characteristics of rivers/streams sampled  67  3.2  Comparison of guild and species abundance between controls and buffers (19961998)  68  3.3  Changes individual species abundances between 1996-1998  70  4.1  Number of bird crossings for habitat guilds and individual species at river and clearcut edges  109  4.2  Paired t-tests comparing mean number of crossings at river and clearcut edges 110  4.3  Linear regressions for number of bird crossings with density of birds for each habitat guild at river and clearcut edges  5.1  Mean and standard deviation of vegetation variables in old-growth and riparian buffer strips  5.2  144  Correlation coefficients between first three P C A scores and vegetation variables  5.3  146  Linear regression statistics for vegetation variables vs. log-buffer width in riparian buffers and controls  5.4  147  Species richness and abundance of selected bird guilds and species in buffers  5.5  Ill  148  Stepwise linear regression models for abundance of guilds in buffers against vegetation variables and buffer width  .'  150  vii  List of Figures 2.1  Species abundance distributions for 50m intervals from the river edge  31  2.2  Mean abundance and standard deviations at 50m intervals from the river edge.... 33  2.3  Means and standard deviations of species richness at 50m intervals from the river edge (1996 - 1998)  35  3.1  Map of study area and sampling grid layout  73  3.2  Species-area relationships for buffers and controls for the years 1996-1998  74  3.3  Species diversity for buffers and controls for the years 1996-1998  75  3.4  Total abundance for buffers and controls for the years 1996-1998  76  3.5  Percentage changes in abundance for buffers and controls for the period 19961998  77  3.6  Species turnover of bird communities for 1996-1998 for buffers and controls.... 78  3.7  Similarity of species assemblages in buffers and controls for the years 19961998  3.8  78  Correlation of change in species richness from 1996-1998 between buffers and clearcuts  79  3.9  Changes in abundance between 1996-1998 for clearcuts  80  4.1  Schematic drawing of mapping grid established at each site  4.2  Relationship between number of movements by Swainson's Thrush across habitat  113  boundaries and time during the breeding season for river and clearcut edges.... 114 4.3  Relationship between number of crossings and habitat guild at each buffer width category for river and clearcut edges  4.4  Linear regressions of bird movements with log buffer width across rivers and clearcut edges by habitat guild  5.1  115  116  Map of study area showing approximate locations of nine watersheds and layout of each site  153  5.2  Principal component analysis of 10 vegetation variables  154  5.3  Linear regressions of principal component scores 2 and 3 against log buffer width  5.4  155  Plots of vegetation variables with high loadings on principal components 2 and 3 for buffers and old-growth  156  Acknowledgements There are many who helped make this research happen. I would like to thank Manfred Enstipp, Kevin Fort, Stephanie Frioud, Sam Hicks, Dirk Lewis, Ramona Maraj, Gaby Matscha, Stephanie Melles, Frank Pouw, Sarah Weber and several volunteers who endured rain, blackflies, flat tires and unrelenting salmonberry thickets for the enjoyment of watching, hearing, and counting birds. I would also like to thank G. Matscha and C. Ferguson for their help in processing the vegetation data and E. Harms, M . Porter, and D. Dolecki for processing many insect samples. I am grateful to W. French and the engineering group at Weyerhaeuser Canada who provided indispensable advice and cooperation in locating field sites.  Statistical advice from D. Huggard, Valerie LeMay, John Richardson, and D . Schluter was much appreciated. Thanks also to L. Barrett-Lennard, J. Hellman, D. Huggard and D. Srivastiva for critical comments on earlier drafts of manuscripts. I owe many thanks to my committee, Ken Lertzmann, Bill Neill, John Richardson, and D. Schluter for their wise comments and general interest and support. I would especially like to thank my supervisor, J.N.M. Smith for his support of this project and his tireless editing and insights that greatly improved this research. This project was supported financially by the Habitat Conservation Trust Fund and Forest Renewal B.C. M y personal support was provided by a N S E R C operating grant to JNMS, a University of B.C. graduate fellowship to SS, Forest Renewal B.C. and the David MacLean Scholarship Fund.  I would like to thank my parents and family and many friends for their support over the years. And lastly, I want to thank the faculty and students of the huts for providing the best place possible to pursue this research.  1  Chapter 1: General Introduction Natural disturbance regimes in many areas of the world have been altered increasingly by human activities. Many habitats have been heavily affected and increasing numbers of species are at risk. In forested landscapes of the Pacific Northwest, accelerated clearcutting in the last several decades has resulted in reduction and fragmentation of oldgrowth forest area. Unlike forest remnants in eastern North America that are surrounded by an agricultural or urban matrix, old-growth patches in this area are distributed in a matrix of regenerating forest (Freemark et al., 1995). This matrix is represented by early successional (5-20 years) forests with dense shrubs and deciduous/coniferous saplings that develop into dense coniferous second-growth after canopy closure (25-70 years) until reaching the mature stage. These alterations to landscapes that are historically subjected to very low frequencies of major disturbance (Lertzman et al., 1996) raise ecological questions about the ability of these landscapes to sustain the full range of species found in unmanaged forests. As well, "ecologically-based" forestry practices such as retention of buffer strips along streamsides are being widely promoted and implemented without adequate scientific theoretical and empirical testing (Simberloff et al., 1992). Scientific studies evaluating habitat alteration are often considered to be inadequate because study time scales are too short and measured variables are inadequate (Kellner et al., 1992).  Theories and evidence of forest fragmentation  Theories of the impacts of forest fragmentation on biodiversity have their roots in the species-area relationship (Connor and McCoy, 1979), a general empirical pattern where species number is positively related to area. One mechanism to explain this pattern is the dynamic equilibrium hypothesis developed in the theory of island biogeography (MacArthur and Wilson 1963, 1967). In this theory, island size and isolation are the primary factors determining the number of species on islands through their influence on immigration and extinction. Populations on islands of larger size have lower probabilities of extinction and those closer to mainland source populations have higher  2  rates of imrnigration. Another mechanism is provided by the habitat-diversity hypothesis whereby larger islands are also likely to have greater habitat diversity (Williams, 1964; Haila, 1983, Boecklen, 1986).  The species-area relationship has been well documented for forest patches in fragmented landscapes (review in Faaborg et al., 1995, Freemark et al., 1995). MacArthur and Wilson's (1963, 1967) dynamic equilibrium hypothesis has been applied to montane "habitat islands" of various sorts (reviewed in Harris, 1984) surrounded by an inhospitable matrix. The island analogy does not fully represent the complex interactions of forest fragments with their surrounding environment (Diamond, 1975; Zimmerman and Bierregaard, 1986; Harris, 1984; Wiens, 1994). Patch area, however, remains a major factor in determining species distribution and abundance in forest fragments (reviewed in Askins et al., 1990). Other mechanisms such as habitat diversity (Zimmerman and Bierregaard, 1986) and patch connectivity to facilitate dispersal (e.g. Lande, 1988; Stacey and Taper, 1992, Robinson et al., 1995, Brooker and Brooker, 2002) are also important.  Fragmentation of forest habitat results in changes to both the abiotic (Saunders et al., 1991, Chen et al., 1995) and biotic (Ranney, 1981; Wilcove et al., 1986) environments at the forest edge. These edge effects often harm forest-dwelling species (e.g. Gates and Gysel, 1978; Wilcove et al., 1986; Murcia, 1995) and can penetrate deep into forest interiors (Camargo and Kapos, 1995; Chen et al., 1995). Edge effects can be short-lived and may change depending on their interactions with the matrix and the remaining forest (Gascon et al., 2000). In the wet temperate rainforests of the Pacific Northwest, forest succession at edges may result in greater habitat diversity, in particular, a shift towards deciduous tree species associated with young regenerating forests. Higher densities of deciduous trees are positively associated with bird community diversity along wider rivers (Lock and Naiman, 1998) and with several individual species abundances (Enns et al., 1993, Hagar, 1999). The extent to which increased habitat diversity can compensate for reduced forest area, however, is unknown.  3  Fragmentation of riparian habitats Riparian ecosystems along streams have high species diversity compared to contiguous upland areas for several taxonomic groups (Stevens et al., 1977; Stauffer and Best, 1980; Brinson et al., 1981; Raedecke, 1988). This diversity depends on several features of natural riparian environments including: (1) presence of surface water and soil moisture, (2) diverse plant communities dominated by woody vegetation (3) diverse habitat structures with high edge area and (4) linear corridors that provide migration and dispersal routes. Since riparian areas are used by specialized species and also by species characteristic of upland habitats (Wauer, 1977; Wiebe and Martin, 1998), these areas can contain the majority of species within an area.  Despite their high diversity, riparian ecosystems represent one of the most threatened habitats globally (Naiman et al., 1993). Riparian habitats are currently threatened by agricultural, forestry, recreation and urban development. In forested landscapes of B.C., protection initiated by the B.C. Coastal Fisheries/Forestry Guidelines (Ministry of Forests et al., 1992) recommended leaving 10m wide buffers of uncut trees at small fish-bearing streams and 30m buffers for streams over 30m wide to protect fish populations from negative impacts of logging. These guidelines, however, have been seldom properly implemented or enforced (Tripp et al., 1992; Tripp, 1994; B.C. Ministry of Forests and Province of B.C., 1994). In addition, many buffers on coastal streams have subsequently blown down in windstorms. The B.C. Forest Practices Code (1995) created wider riparian buffer zones. Currently, a range of 30m-70m based on stream width is required for fish-bearing streams.  It is of interest to know how far buffers of varying width protect the terrestrial and semiterrestrial components of riparian ecosystems. Early studies of avian riparian systems were in arid southern forests of the United States. These studies showed that riparian habitats have high population diversity and density (e.g. Johnson et a l , 1977; Szaro, 1980; Tubbs, 1980). More recent studies from unmanaged forests in the Pacific Northwest cast doubt on whether the results from southern forests can be extrapolated to  4  the forests of coastal B.C., Oregon, Washington and Alaska (McGarigal and McComb, 1992; Pearson and Manuwal, 2001). These studies found weak (Pearson and Manuwal, 2001) or reversed (McGarigal and McComb, 1992) gradients in diversity and abundance from riparian to upland forests.  Generalizations about fragmentation effects in riparian habitats are difficult because study designs vary in two key ways. First, one or both sides of the river may be logged; second, the successional stage of both buffers and adjacent clearcuts varies. In addition, most studies to date have been conducted along small headwater streams (Darveau et al., 1995; Hagar, 1999, Pearson and Manuwal, 2000). When a riparian buffer is created, the harvested portion supports only early successional species and habitat generalists. Buffers retain a varying proportion of the original forest species with rare and areasensitive species tending to be absent locally in narrow (< 60m) buffers (Darveau et al., 1995; Hagar, 1999, Pearson and Manuwal, 2000) despite linkages with larger patches at one or both ends. In this way, riparian buffers are similar to isolated upland fragments. One important difference is the presence of riparian specialists in buffers. These studies (Darveau et al., 1995; Hagar, 1999, Pearson and Manuwal, 2000), however, do not explicitly consider fragmentation effects on riparian specialists as a group.  Objectives  My study sought to increase the ecological knowledge needed to protect riparian habitat for birds. I investigated the use of riparian habitats to forest birds in B.C. coastal forests by describing birds in natural riparian communities, and analysing the responses of birds to timber harvesting practices. In chapter 2,1 evaluated the use of riparian habitats by birds in unmanaged coastal temperate old-growth forests by comparing bird communities in riparian and adjacent upland habitats.  The knowledge gained in chapter 2 is used to evaluate the effects of forest harvesting on bird communities. In chapter 3,1 examine the value of riparian buffer strips 5-7 years post-harvest as breeding habitat to forest birds over a 3 year period. I tested if forest  5 buffer width was related to the abundance, species diversity, and density of birds by comparing communities across a spectrum of buffer widths to equal areas of undisturbed forest. I also relate temporal patterns of species richness in buffers to those in adjacent clearcuts.  The forest/clearcut and forest/river edges may act as barriers to the movement of birds (Faaborg et al., 1995, Walters, 1998, Brooker and Brooker, 2002). In chapter 4,1 study the ability of species to cross habitat boundaries by estimating movement patterns of birds across streams and between buffers and adjacent clearcuts. To test whether the frequency of movements varies with buffer width and edge type, I compare movements for several guilds and species across both edges among different buffer widths.  In chapter 5,1 investigate how bird communities may be influenced by the diversity of riparian vegetation in buffers. I compare riparian vegetation structure among varying buffer widths to test if vegetation differs with buffer width. I then test whether vegetation structure predicts species richness and abundance of several species and habitat guilds.  In chapter 6,1 summarize the major findings of my thesis and its contribution to our current understanding of fragmentation of riparian habitats. Finally, I discuss some management implications based on my results and my general experience.  6  Literature cited  Askins, R.A., J.F. Lynch, and R. Greenberg. 1990. Population declines in migratory birds in eastern North America. Current Ornithology 7:1-57.  B.C. Ministry of Forests and B.C. Environment. 1995. Forest Practices Code of British Columbia: Riparian management area guidebook. Province of British Columbia, Victoria, B.C.  B.C. Ministry of Forests and Province of B.C. 1994. Forest, range and recreation resource analysis. Crown Publications Inc., Victoria, B.C.  B.C. Ministry of Forests, B.C. Ministry of ELP, Dept. of Fisheries, and Council of Forest Industries. 1992. B.C. coastal fisheries/forestry guidelines., Victoria, B.C.  Boecklen, W.J. 1986. Effects of habitat heterogeneity on the species-Area relationships of forest birds. Journal of Biogeography 13:59-68.  Brinson, M . M . , B.L. Swift, R.C. Plantico, and J.S. Barclay. 1981. Riparian ecosystems: their ecology and status. Department of the Interior, U.S. Fish and Wildlife Service, Kearneysville, V W .  Brooker, L., and M . Brooker. 2002. Dispersal and population dynamics of the Bluebreasted Fairy Wren Malurus pulcherrimus in fragmented habitat in the western Australian wheatbelt. Wildlife Research in press.  Camargo, J.L.C., and V. Kapos. 1995. Complex edge effects on soil moisture and microclimate in central Amazonian forest. Journal of Tropical Ecology 11:205-221.  7  Chen, J., J.F. Franklin, and T.A. Spies. 1995. Growing-season microclimatic gradients from clearcut edges into old-growth Douglas-fir forests. Ecological Applications 5:74-86.  Connor, E.F., and E.D. McCoy. 1979. The statistics and biology of the species-area relationship. American Naturalist 113:791-833.  Darveau, M . , P. Beauchesne, L. Belanger, J. Huot, and P. Larue. 1995. Riparian forest strips as habitat for breeding birds in boreal forest. Journal of Wildlife Management 59:67-78.  Diamond, J.M. 1975. The island dilemma: lessons of modern biogeographical studies for the design of natural preserves. Biological Conservation 7:129-146.  Enns, K . A . , E.B. Peterson, and D. McLennan. 1993. Impacts of hardwood management on British Columbia wildlife: problem analysis. F R D A II, Victoria, B.C.  Faaborg, J., M . Brittingham, T. Donovan, and J. Blake. 1995. Habitat fragmentation in the temperate zone. Pages 357-380 in Martin, T.E., and D . M . Finch, editors. Ecology and management of Neotropical migratory birds: a synthesis and review of critical issues. Oxford University Press, New York.  Freemark, K.E., J.B. Dunning, S.J. Hejl, and J.R. Probst. 1995. A landscape ecology perspective for research, conservation, and management. Pages 381-427 in Martin, T.E., and D . M . Finch, editors. Ecology and management of Neotropical migratory birds: a synthesis and review of critical issues. Oxford University Press, New York.  Gascon, C , G.B. Williamson, and G.A.B. daFonseca. 2000. Receding forest edges and vanishing reserves. Science 288:1356-1358.  Gates, J.E., and L.W. Gysel. 1978. Avian nest dispersion and fledging success in fieldforest ecotones. Ecology 59:871-883.  8  Hagar, J.C. 1999. Influence of riparian buffer width on bird assemblages in western Oregon. Journal of Wildlife Management 63:484-496.  Haila, Y . 1983. Land birds on northern islands: a sampling metaphor for insular colonization. Oikos 41:334-351.  Harris, L . D . 1984. The fragmented forest. The University of Chicago Press, Chicago.  Johnson, R.R., L.T. Haight, and J.M. Simpson. 1977. Endangered species vs. endangered habitats: a concept. Pages 68-79 in Johnson, R.R., and D.A. Jones, editors. Importance, preservation and management of riparian habitat: a symposium. U S D A Forest Service General Technical Rep, Ft. Collins, Colorado.  Kellner, C.J., J.D. Brawn, and J.R. Karr. 1992. What is habitat suitability and how should it be measured? Pages 476-488 in McCullough, D.R., and R.H. Barrett, editors. Wildlife 2001. Elsevier Science Publishers, London.  Lande, R. 1988. Demographic models of the Northern Spotted Owl (Strix occidentalis caurina). Oecologia 75:601-607. Lertzman, K.P., G.D. Sutherland, A. Inselberg, and S.C. Saunders. 1996. Canopy gaps and the landscape mosaic in a coastal temperate rain forest. Ecology 77:1254-1270.  Lock, P.A., and R.J. Naiman. 1998. Effects of stream size on bird community structure in coastal temperate forests of the Pacific Northwest, U.S.A. Journal of Biogeography 25:773-782.  MacArthur, R.H., and E.O. Wilson. 1963. An equilibrium theory of insular zoogeography. Evolution 17:373-387.  9  Mac Arthur, R.H., and E.O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, N.J.  McGarigal, K., and W.C. McComb. 1992. Streamside versus upslope breeding bird communities in the central Oregon Coast Range. Journal of Wildlife Management 56:1023.  Murcia, C. 1995. Edge effects in fragmented forests: implications for conservation. T R E E 10:58-62.  Naiman, R.J., H . Decamps, and M . Pollock. 1993. The role of riparian corridors in maintaining regional biodiversity. Ecological Applications 3:209-212.  Pearson, S., and D.A. Manuwal. 2001. Breeding bird response to riparian buffer width in managed Pacific Northwest Douglas Fir forests. Ecological Applications 11:840-853.  Raedecke, K.J. 1988. Streamside management - riparian wildlife and forestry interactions. University of Washington Institute of Forest Resources, Seattle, Washington.  Ranney, J.W., M.C. Bruner, and J.B. Levenson. 1981. The importance of edge in the structure and dynamics of forest islands. Pages 67-96 in Burgess, R.L., and D . M . Sharpe, editors. Forest island dynamics in man-Dominated landscapes. Springer-Verlag, New York.  Robinson, S.K., F.R. Thompson, T . M . Donovan, D.R. Whitehead, and J. Faaborg. 1995. Regional forest fragmentation and the nesting success of migratory birds. Science 267:1987-1990.  Saunders, D.A., R.J. Hobbs, and C R . Margules. 1991. Biological consequences of ecosystem fragmentation: a review. Conservation Biology 5:18-32.  10  Simberloff, D., J.A. Farr, J. Cox, and D.W. Mehlman. 1992. Movement corridors: conservation bargains or poor investments? Conservation Biology 6:493-504.  Stacey, P.B., and M . Taper. 1992. Environmental variation and the persistence of small populations. Ecological Applications 2:18-29.  Stauffer, D.F., and L . B . Best. 1980. Habitat selection by birds of riparian communities: evaluating effects of habitat alterations. Journal of Wildlife Management 44:1-15.  Stevens, L.E., B.T. Brown, J.M. Simpson, and R.R. Johnson. 1977. The importance of riparian habitat to migrating birds. Pages 156-164 in Johnson, R.R., and D.A. Jones, editors. Importance, preservation and management of riparian habitat: a symposium. U S D A Forest Service, Tucson, Arizona.  Szaro, R.C. 1980. Factors influencing bird populations in southwestern riparian forests. Pages 403-418 in DeGraff, R . M . , editors. Workshop management of western forests and grasslands for nongame birds. U S D A For. Ser. Gen. Tech. Rep. INT-86,.  Tripp, D. 1994. The use and effectiveness of the coastal fisheries/forestry guidelines in selected forest districts of coastal B.C.  Tripp, D., A. Nixon, and R. Dunlop. 1992. The application and effectiveness of the coastal fish/forestry guidelines in selected cutblocks on Vancouver Island. A Tripp Biological Consultants Ltd. report to the B.C. Ministry of Forests, Integrated Resources Branch, Victoria, B.C., Nanaimo, B.C.  Tubbs, A . A . 1980. Riparian bird communities of the Great Plains. Pages 419-433 in DeGraaf, R . M . , editors. Workshop proceedings: management of western forests and grasslands for nongame birds. U S D A Forest Service General Technical Report, Odgen, Utah.  11  Walters, J.R. 1998. The ecological basis of avian sensitivity to habitat fragmentation. Pages 537 in Marzluff, J.M., and R. Sallabanks, editors. Avian conservation: research and management. Island Press, Washington, D.C.  Wauer, R.R. 1977. Significance of Rio Grande riparian systems upon the avifauna. Pages 165-174 in Johnson, R.R., and D.A. Jones, editors. Importance, preservation and management of riparian habitat: a symposium. U S D A Forest Service, Tucson, Arizona.  Wiebe, K.L., and K. Martin. 1998. Seasonal use by birds of stream-side riparian habitat in coniferous forest of northcentral British Columbia. Ecography 21:124-134.  Wiens, J.A. 1994. Habitat fragmentation: island vs landscape perspectives on bird conservation. Ibis 137:S97-S104.  Wilcove, D.S., C H . McLellan, and A.P. Dobson. 1986. Habitat fragmentation in the temperate zone. Pages 237-256 in Soule, M.E., editors. Conservation biology. The science of scarcity and diversity. Sinauer Associates, Sunderland, Massachusetts.  Williams, C B . 1964. Patterns in the balance of nature. Academic Press Inc., London.  Zimmerman, B.L., and R.O. Bierregaard. 1986. Relevance of the equilibrium theory of island biogeography and species-area relations to conservation with a case from Amazonia. Journal of Biogeography 13:133-143.  12  Chapter 2: Habitat Use By Riparian And Upland Birds In Old-Growth Coastal British Columbia Rainforest Introduction Riparian habitats along rivers are influenced by both stream channel processes and the adjacent upland vegetation (Brinson et al. 1981, Naiman et al. 1993). Topography, plant communities, hydrologic patterns and soil type distinguish riparian areas from upland areas. These features are common to all riparian habitats, but vary greatly depending on geographical location. Riparian ecosystems often support high bird diversity and abundance compared to upland habitats (Carothers and Johnson 1974, Wauer 1977) because of their complex foliage structure (MacArthur and MacArthur 1961), high tree species diversity (Rotenberry and Wiens 1980, Holmes and Robinson 1981) and proximity to water. Bird diversity, in particular, is often positively correlated with high diversity of vegetative structural forms, i.e. complex canopy, shrub and herbaceous layers (MacArthur and MacArthur 1961, Carothers and Johnson 1975, LaRue et al., 1995).  There is a strong bird diversity gradient from riparian to upland habitats in southern regions of the U.S. (e.g. Dickson 1978, Stauffer and Best 1980). In contrast, studies from mature, undisturbed stands in northern forests have shown conflicting patterns. These studies (McGarigal and McComb 1992, Murray and Stauffer 1995, Wiebe and Martin 1998) showed equal or lower diversity in riparian habitats of small streams compared to upslope habitat. To date, studies of bird-habitat relationships in Northwestern North America (McGarigal and McComb 1992, Murray and Stauffer 1995, Wiebe and Martin 1998) have focused on riparian areas associated with small mountain streams < 5m wide. Riparian areas may show greater diversity near larger streams and rivers (Knopf and Samson 1994); however, use of riparian habitat by birds along larger rivers in the Pacific Northwest has not been well examined. Several recent studies (Darveau et a l , 1995, Kilgo et al., 1998, Hagar, 1999) have focused on bird/habitat relationships in buffer strips caused by logging, a topic I explore in future papers. Here I concentrate on patterns within continuous undisturbed forest.  13  My first objective was to test the hypothesis that bird species diversity and abundance is higher in old-growth riparian habitat associated with large streams and rivers than in adjacent old-growth upland habitat. Three broad groups of bird species live in riparian to upland zones: riparian obligates strongly associated with the water edge and aquatic environment, riparian species associated with the terrestrial vegetation and/or food resources and upland species. Disproportionate use of riparian habitats by birds should be reflected in a decline in species richness and abundance with increasing distance from the river. Association with riparian habitat may vary among certain species or guilds. Riparian specialists that rely on the stream or river as a food source and nest in the adjacent vegetation should decline in abundance with distance from the river. Foliage gleaners may be most abundant in riparian areas dominated by deciduous trees and tall shrubs. Forest edges such at these are thought to support a richer invertebrate fauna (Hansson 1983, Helle and Muona 1985) presumably due to higher primary productivity (Ranney et al., 1981) and susceptibility of edge trees to insect attacks (Hansson, 1983). Similarly, aerial hawkers such as flycatchers may respond to emergent aquatic insects near streams and rivers (Gray 1993) and are predicted to occur at highest densities near the water edge. Conversely, conifer specialists may increase with distance from the river.  Second, I explore how variation in vegetation composition and structure from riparian to upland habitat explains distribution patterns of several species.  Studies  conducted in other temperate coniferous forests have found differences in habitat and vegetation structure between riparian and upland (McGarigal and McComb 1992). If the structural and species attributes of riparian vegetation communities are the dominant predictors of bird diversity and abundance, then use of riparian habitats should be related to the prevalence of these structures relative to upland areas. Alternatively, bird diversity and abundance should not differ between structurally similar riparian and upland habitats.  14  Methods  Study area  The study was conducted in three valleys on the west coast of Vancouver Island, British Columbia between Ucluelet in the north and Bamfield in the south (Latitude, 48°5', Longitude, 125°5'). Four sites of old-growth continuous forest were selected along the Nahmint, Taylor and Klanawa Rivers; two sites were located along the Nahmint River, but were separated by approximately 2 km and so can be considered independent. These rivers were classified to stream order at a 1:50,000 map scale based on their degree of branching according to Kuehne (1962). Nahmint and Taylor rivers were fourth order and Klanawa river was fifth order. The sites are located in the Western Vancouver Island ecoregion within the Coastal Western Hemlock moist to very wet maritime biogeoclimatic subzones (Klinka et al. 1991, Nuszdorfer and Boettger 1994). The forest is dominated by amabilis fir (Abies amabilis), western hemlock (Tsuga heterophylla) and western red cedar (Thuja plicata). Red alder (Alnus rubra) and big leaf maple (Acer macrophyllum) are found in highest densities adjacent to the rivers, but were also scattered throughout the forest in moister areas. The understorey is dense and highly stratified and contains shrubs such as salmonberry (Rubus spectabilis), red huckleberry (Vaccinium parvifolium), salal (Gaulthier shallon) and devil's club (Oplopanax horridus), with alaskan and oval-leaved blueberry (Vaccinium alaskense and ovalifolium) predominating in the upland areas. The climate is cool and wet in winter, warm and dry in late summer (July - Sept). Annual precipitation in the area averages 3,100 mm with mean daily temperatures in January of 3.2°C and July of 15.6°C (Environment Canada Climate Data Services).  The three valleys have been subjected to intensive logging in the last thirty years. The study sites are embedded within a mosaic of forest ages across the landscape in which the amount of primary forest varies from 50-70%.  15  Stream width and classification by order of rivers on western Vancouver Island Name  Bankfull Width (m)  Klanawa River  15  Nahmint River  32  Taylor River  57  Classification 5 order th order th  4  4 order th  Vegetation sampling I sampled vegetation during 1995 within 20m radius circular plots (0.13 ha) using the procedure modified from James and Shugart (1970) to account for the large size of trees locally by increasing the radius (Mueller-Dombois and Ellenberg, 1974, Bryant et al. 1993). Plots were placed at three stations along each riparian and upland transect at distances 20m and 160m from the river respectively for a total of six plots per site. To improve the accuracy of visual estimation of percentage cover, each plot was divided into four quadrats that were standardized using the four points of a compass; measurements were then averaged to give a single value for each plot. Trees and snags were measured within the 20m radius while shrubs were sampled in a 10m radius subplot nested in the 20m circular plot. Forbs were measured using four l m quadrats nested within the shrub 2  plots. Coarse woody debris (CWD) on the forest floor was sampled along the circumference of the plot. For trees and snags, I recorded species, diameter at breast height, decay class (based on Thomas et al. 1979) and number of cavities. For understorey plants, I recorded species richness and percent cover for forbs, ferns and shrubs at four height categories: <30cm, <lm, l-2m and 2-3m. Percentage cover of each species for over 90 species was estimated visually for each height category. Thus, the sum of coverages could exceed 100% for all strata combined. Trees <3 m were treated as shrubs. For coarse woody debris, I recorded diameter, length, height above ground and decay class for each species (Thomas et al. 1979); pieces not identifiable by species were classified as unknown. Canopy cover was estimated using a spherical densiometer (Strickler, 1959) which reflects the amount of canopy cover in a spherical concave  16  mirror. The average canopy cover of four densiometer readings in each cardinal direction was calculated for each plot.  Bird sampling  Bird censuses were conducted using the spot-mapping method (Robbins 1970, Bibby et al. 1992). A mapping grid was established at each site by running a 450m transect line parallel to the river and perpendicular lines running 200m away from the river for a total size of 9 ha; the perpendicular distance was constrained by the ubiquitous presence of logging roads upland of the grids and the steep valley topography. Grids were divided into nine perpendicular transect lines at 50m intervals and each transect was flagged every 25m. Spot-mapping was conducted at each site by at least two observers who walked the transect lines from 05:00 toll:00 on days without rain and/or high winds. Bird communities were mapped four times through each of the 1996-1998 breeding seasons. The order in which sites were sampled was varied to reduce bias except when sites were far apart. Three observers counted in 1996 and four in 1997-1998. Two individual observers were present in 1997 and 1998 and one observer was present in all years. A l l observers were rotated among sites and transect lines during each year. When mapping, I also recorded microhabitat information describing the forest structure where the bird was first heard or observed, (i.e. "salmonberry", "shrub", etc.) and its approximate location. Spot maps were evaluated following the protocol established by Robbins (1970) except that number of registrations was used rather than the number of territories (Enemar et al. 1978).  Data analysis A l l data were tested for normality using the Shapiro-Wilks statistic. Homogeneity of variances were tested using the Levene and Bartlett Box statistics. Data that violated the assumptions were rank transformed (Conover and Iman 1981, 1982) or else nonparametric statistics were used. I used an alpha value of 0.10 to adjust for low sample  17  sizes and the high biodiversity cost of making a type II error in resource management decisions (Toft and Shea 1983, Dayton 1998).  Vegetation  Because studies have demonstrated that both floristics and structural attributes influence habitat selection by birds (MacArthur and MacArthur 1961, Holmes and Robinson 1981, Robinson and Holmes 1984), I focused on broad tree classifications, snags, C W D , and the dominant shrubs representing 17 of the 158 variables measured. Forbs were analyzed as a group rather than as individual species. I converted diameter to basal area (m /hectare) for coniferous trees, deciduous trees and snags. Density (number ha' ) was 2  1  also recorded for these three categories and for saplings (trees <10cm diameter). Coarse woody debris was expressed as volume per unit area (m ha" ) based on a formula 3  1  developed by Van Wagner (1968). From the 17 variables, I examined a correlation matrix. Variables that measured the same structural feature and were highly correlated (>70%. Approx. 0.001 sign, level) were either averaged together as in the case of shrub species richness and cover for each height or eliminated from further analyses. To assess if there were clear general vegetation differences between riparian and upland habitats, I performed a principal components analysis using 10 variables. I used varimax rotation to improve the separation of components. I retained the first three components representing those with eigenvalues that exceeded the broken stick model (Legendre and Legendre, 1983). I then performed a paired t-test for the two factor scores (principal components). Where there was a significant difference in factor scores between riparian and upland habitats, I selected those variables with high loadings (> 0.5) for the factors and compared the mean values of riparian areas with upland areas using paired t-tests.  Birds  I analyzed abundances both by guilds and by individual species because patterns of abundance and distribution of species and guilds may differ between riparian and upland, e.g. riparian specialists should be more abundant near the river. For the guild analysis, I  18  grouped species abundances according to foraging method, habitat use and migratory status. Assignments for guilds were based on published literature (Hatler et al., 1978, Ehrlich et al., 1988 and Peterson, 1990) and local knowledge (Campbell et al., 1990, 1997) (Appendix 1). I analyzed the abundance for all species with > 5 observations in each year (9 species). I tested for differences in abundance among years using repeatedmeasures A N O V A (rmANOVA) with year as the repeated variable. I then tested for differences in abundance across four 50m intervals from the river edge using linear regression. For all comparisons, data were pooled across years because there were no significant distance by year interactions except for American robin. For this species, each year was analysed separately.  To estimate species richness of bird communities, I excluded all species observed in only one census session during each year as recommended by Willson et al. (1996), and calculated the mean species richness for each site. I compared species richness at four 50m intervals from the river edge using r m A N O V A to test for differences among distance from river and over time. Site variation was not tested separately as a categorical variable because of confounding effects due to low sample sizes. To analyze community diversity, I used the log-series distribution (Fisher et al. 1943). It is considered to be the best measure of diversity available (Wolda, 1983) and less sensitive to changes in abundance of the most common species. The log series relates the number of species (S) to the number of individuals in a sample (N) according to the relationship  S = alog (l + N / a ) e  where a is a measure of species diversity (Fisher et al. 1943). I calculated a values for all sections (Williams 1964) and compared these using r m A N O V A to test for differences among distance from river and over time. Values reported in the results section are means ± 1 SD.  Habitat Associations  19  To explore the relationship between vegetation composition and species distributions, I correlated bird abundance for the nine most common species with six vegetation variables. The variables included those whose means differed between riparian and upland habitats, e.g. deciduous, coniferous and snag density, and shrub cover as well as the two dominant shrub species, salmonberry and blueberry spp. I used stepwise multiple linear regression and a significance level of 0.10 to enter the model. I deleted sites from the analysis where a species was not known to occur, e.g. Hammond's flycatcher at Klanawa river.  Results Vegetation  Vegetation patterns were best described by three major axes. These first three factors together represented 79% of the variance among the vegetation variables with the first factor representing 39%. The principal components accounted for most of the variance (>70%) in individual vegetation variables except for snag density. The first factor distinguishes areas with high conifer and snag density and low deciduous tree density and average shrub cover including low cover of salmonberry cover (Table 2.1). The second factor distinguishes areas with high shrub and forb richness and low snag density, while the third factor distinguishes areas with high shrub and forb cover and low amounts of downed wood. The first factor was the only one to show a difference between riparian and upland habitats (Paired t = 4.09, P = 0.026). 3  Vegetation structure and composition both varied between riparian and upland areas. Of the six individual variables with high loadings on the first principal components axis, three differed between riparian and upland habitats (Table 2.2). Riparian areas had five times the density and volume of deciduous trees while upland areas had 1.4 and 1.9 times the density of conifers and snags respectively (Table 2.2). Sites varied considerably, with some sites having higher amounts of conifers and snags. Downed wood volume was 1.4 times higher in upland habitats compared to riparian and salmonberry cover was 2.3  20  times higher in riparian habitats. Neither of these differences was statistically significant, however, due to high variation among sites. Average shrub cover did not vary between riparian and upland habitats.  Bird abundance and diversity  In 1996-1998,1 recorded a total of 645 birds of 36 species on all visits. Nine species had >20 observations for all sites combined during the period and accounted for 80% of the total observations. The six most abundant species were winter wren {Troglodytes troglodytes), chestnut-backed chickadee (Poecile rufescens), American robin (Turdus migratorius), Swainson's thrush (Catharus ustulatus), Pacific-slope flycatcher (Empidonax difficttis), and golden-crowned kinglet (Regulus satrapa). These six species predominated at all distances from the river although their relative rankings varied slightly (Figure 2.1). Except for one forest interior species (Pacific-slope flycatcher), these species were either forest generalists or ubiquitous. Winter wren and chestnutbacked chickadee predominated interior sections and were replaced in part by American robin and Swainson's thrush near the river. There were few riparian specialists even close to the river and all of these had low abundances with the exception of Hammond's flycatcher (Empidonax hammondii).  Total abundance did not differ with distance from the river (F3.44 = 1.26, P = 0.30) (Fig. 2.2a). This result is largely due to the dominant species being forest generalists. However, as expected, abundances of riparian specialists and aerial hawkers did decline significantly with distance from the river (Fig. 2.2b,c) (riparian specialists - Fi.46 = 8.43, P = <0.01, aerial hawkers - Fije = 2.45, P = 0.08). Foliage gleaner abundance (Figure 2.2d) showed no pattern with distance (F 6 iA  = 0.42, P = 0.74). Of the nine species that  had sufficient data for analysis, only two showed significant variation across the distance intervals: Swainson's thrush (Fi,  46  = 2.82, P = 0.10) (Fig. 2.2k) and Hammond's  flycatcher (Empidonax hammondii) (Fi.46 = 3.91, P = 0.015) (Fig. 2.2i) were positively associated with the rivers.  2 1  Species richness did not differ with distance from the river (Fj, n - 0.15, P = 0.93) and there was no significant interaction between distance and year (F<j, 24 = 0.61, P = 0.62). However, I did detect yearly variation in species richness (within-subjects: year F,  2 24  = 6.28, P = 0.03) with 14% more species in 1997 than in 1996 and 1998 (Figure  2.3). A similar pattern was found for species diversity; diversity did not vary with distance from the river (F?, n =0.09, P = 0.96). This pattern remained constant among years (F ,24 =2.38, P = 0.121), but diversity was highest in 1997 (F , 4 =6.16, P = 0.03). 6  2 2  For all years, assemblages were predominated by the same five species at each distance comprising 53-58% of total observations with minor variations in their ranking: winter wren, chestnut-backed chickadee, American robin, Pacific-slope flycatcher, and Swainson's thrush. There was no difference in proportions of rare species over the four distance classes (Friedman % 3 1-55, P = 0.67) or among years (Friedman ^22  =  3.60, P  = 0.16).  Habitat associations  Four species had significant correlations with habitat variables (Table 2.3). Abundances of Swainson's thrushes and hairy woodpeckers had negative relationships with blueberry cover that explained 41-46% of the variance; a negative relationship with snag density explained a further 28% of the variance for Swainson's thrushes. Abundance of Hammond's flycatchers was strongly related to deciduous tree density which accounted for 88% of the variance. Pacific-slope flycatcher abundance was positively related to snag density which explained 84% of the variance. Riparian specialist abundance was positively related to salmonberry cover and negatively related to snag density. Abundance of aerial foragers was negatively related to conifer density that explained 80% of the variance. When distance from river was included as a variable, there was no change in the models with the exception of riparian specialists where distance was the only predictor of abundance.  Discussion  22  Species diversity a n d a b u n d a n c e along the r i p a r i a n gradient Contrary to my original predictions, species abundance and diversity of birds were similar along a distance gradient away from the river. Although low sample sizes reduced my ability to detect differences between habitats, my results support other recent studies in coniferous forests that found riparian areas do not support higher numbers of bird species or individuals (McGarigal and McComb 1992, Murray and Stauffer 1995, Kinley and New house, 1997). These studies were also conducted in areas without an obvious gradient in habitat structure from riparian to upland. In contrast, earlier studies in more arid or agricultural environments (Stevens et al. 1977, Stauffer and Best 1980) found large differences in diversity and abundance between riparian and upland habitats.  In this study, while there were differences in some vegetative attributes between riparian and upland areas, there was considerable variation among sites. In general, riparian plots had higher amounts of deciduous trees and were dominated by salmonberry in the understorey. In contrast, upland areas were dominated by coniferous trees and snags; blueberry species' were the dominant understorey shrubs. Thus the vegetation typically associated with riparian areas may be a more important influence on bird community structure than the actual presence of water provided by streams and rivers. In this study, four bird species as well as the groups of riparian specialists and aerial foragers were associated with densities of certain types of canopy and understorey vegetation independent of riparian/upland habitat. Other studies have found that higher densities of deciduous vegetation have been associated with higher avian abundance and diversity (James and Warner 1982, Willson and Comet 1996). Deciduous trees are known to provide specific habitat requirements such as nesting habitat for many species of birds. For example, black cottonwood (Populus trichocarpa) is used by a majority of birds for nesting (Hopkins et al., 1986, Strong and Bock, 1992). With the exception of the few  species that showed strong habitat relationships, complex topography combined with  consistently moist conditions provides suitable habitat for most species across the riparian-upland gradient that I studied and probably accounts for the lack of strong riparian effects at the community level.  23  The large fourth or fifth-order streams and rivers studied contrast to previous studies in northern forests that focused on smaller second-order streams (McGarigal and McComb 1992, Wiebe and Martin 1998). It might be expected that riparian areas would support higher bird density and diversity along larger streams with a more developed riparian vegetation community. Bird density in riparian habitat has been reported to increase in downstream reaches while species richness remains unchanged except during migration (Wiebe and Martin 1998). This study shows no such trend for abundance, but is consistent with Wiebe and Martin (1998) for species richness. While it remains possible that overall abundance in riparian areas is higher further downstream in this area, it is more likely that most species use both riparian and upland habitats, while only a few species specialize on either habitat. In northwestern forests, these specialists are a small fraction of the total community.  Habitat selection  Of the 36 species recorded, five were found only in sections adjacent to the river. Four of these riparian specialists: common merganser (Mergus merganser), American dipper (Cinclus mexicanus), belted kingfisher (Ceryle alcyon) and spotted sandpiper (Actitus macularia) accounted for the strong decrease in abundance of this group at further distances from the river. All these species depend on stream invertebrates and/or fish as a food resource while using the adjacent riparian vegetation or riverbanks for nesting (Enns et al. 1993, Campbell et al. 1997). The remaining species, the willow flycatcher (Empidonax traillii), is rarely found in mature forest except in riparian areas. In the Coastal Western Hemlock Zone, this species is more commonly found in marshes and early successional clearcuts (5-10 years) associated with young red alder and willow trees (Enns et al. 1993, Campbell et al. 1997).  Five species were found only in upland sections of forest: fox sparrow (Passerella iliaca), Hutton's vireo (Vireo huttoni), olive-sided flycatcher (Contopus cooperi), spotted towhee (Pipilo erythrophthalmus) and yellow warbler (Dendroica petechia). With the  24  exception of Hutton's vireo, the upland species were restricted to a single site characterized by flat, bottomland, open, mixed-wood forest. In this habitat, high levels of red alder and black cottonwood formed the dominant vegetation and there were occasional small patches of standing water from ephemeral headwater streams. This particular type of site provided the specific habitat attributes required by these species (Enns et al. 1993, Campbell et al. 1997). A l l these species are rare in mature forests, and may select large patches of deciduous vegetation in forest interiors. The rare Hutton's Vireo (Hatler et al. 1976) was also restricted to one site and has a patchy distribution (Bryant et al. 1993) in the region. It prefers mixed conifer-dominated woodland with a heavy shrub understory (Campbell et al. 1997) and is found primarily in upland areas (McGarigal and McComb 1992). Although the Pacific-slope flycatcher did not differ significantly between riparian and upland habitats, it was positively associated with higher snag density in upland habitats.  I predicted that two foraging guilds, aerial hawkers and foliage gleaners would decline in abundance away from the river. I found this result for aerial hawkers, but not for foliage gleaners.  Aerial hawkers include the Hammond's flycatcher, a species that shifts  habitats between locations (Willson and Comet 1996). It is considered to be an upslope specialist throughout much of the Northwest, where it strongly associates with conifers (Sakai and Noon 1991, McGarigal and McComb 1992). In the forests of Vancouver Island (Waterhouse and Harestad, 1999, this study) and southeastern B.C. (Kinley and Newhouse, 1997) however, this species was largely restricted to and associated with the deciduous or mixed forests of the riparian zone. While Hammond's flycatcher may use riparian habitat, it is sympatric with the Pacific-slope flycatcher in old-growth forest (Campbell et al. 1997) and its distribution may reflect habitat partitioning between the two species. In southern Colorado, densities were approximately half those of the Pacific-slope flycatcher where the two species co-occur (Beaver and Baldwin 1975) and Hammond's flycatcher used atypical aspen habitat perhaps to escape competition from the Pacific-slope flycatcher in the aspen-conifer habitat. In the one site in this study where Hammond's flycatcher was absent, the Pacific-slope flycatcher occupied the entire riparian-to-upland gradient. I found little difference in abundance of Pacific-slope  25  flycatchers along the gradient. However, when the site where Hammond's flycatcher did not occur is excluded, there was an increasing trend in abundance of Pacific-slope flycatchers with distance from the river. Interestingly, other studies in the Pacific Northwest (McGarigal and McComb 1992, Pearson and Manuwal 2001) have found that Pacific-slope flycatchers are associated with riparian habitats while Hammond flycatchers are associated with upland habitats. The reason for this difference is unclear, but may relate to differences in forest tree composition in upland habitats. The forests in these previous studies are dominated by Douglas fir (Pseudotsuga menziesii) in the upland while the forests in my study are dominated by amabalis fir and western hemlock.  My prediction that foliage gleaners would decline in abundance with distance as densities of deciduous vegetation declined was not supported. Several of the dominant species in this group such as chestnut-backed chickadee and golden-crowned kinglet showed no vegetation associations, while others such as warbling vireo (Vireo gilvus) were rare and numbers were variable among sites. One interesting exception was the Swainson's thrush which, although a forest generalist, was most abundant in riparian habitat. This species is widespread on the west coast (Campbell et al. 1997) and often forages in salmonberry and devil's club in riparian habitats to obtain fruits and invertebrates. The negative relationship with snags and blueberry suggests that these structures or some other closely associated vegetation may be an important influence on habitat selection for this species. Although salmonberry was not a significant predictor in the model for Swainson's thrush abundance, in my study area virtually all Swainson's thrush nests encountered incidentally during surveys were found in salmonberry suggesting a strong association with this shrub for food and/or nesting habitat.  Management implications  Patterns of high diversity and abundance of riparian bird communities have often been explained by dramatic gradients in microclimate and vegetation structure or composition (Stevens et al. 1977, Dickson 1978, Szaro 1980). Where these gradients are subtle as in forests of the Pacific Northwest, the patterns disappear (Wiebe and Martin 1998) or may  26  be reversed with greater diversity and abundance in upland areas (McGarigal and McComb 1992). In these moist forests, differences in structure and composition of vegetation are obscured by the natural variation among sites that I was unable to detect with a low sample size. Northwest forests generally lack a strong microclimate gradient from riparian to upland (Brosofske et al., 1997) and virtually every site has significant sources of water in upland areas in ravines, ephemeral streams or small ponds. These factors create a fine-scale habitat mosaic in which patches of dominant conifers are interspersed with deciduous trees and shrubs that provide habitat for species more typical of deciduous-dominated riparian areas.  Recent discussion of land management practices to preserve native biodiversity of forest species has suggested the use of a landscape-level approach that provides protection of both riparian and upland habitat to ensure connectivity across the landscape (McGarigal and McComb 1992, Wiebe and Martin 1998). While maintaining connectivity may prevent isolation of remnant forest patches (Fahrig and Merriam 1985, Saunders and de Rebeira 1991, Gonzalez 2000), the lack of upland specialists in my study argues against placing too much emphasis on upland areas per se, but rather for giving higher priority to the few species that depend on riparian habitat both at the stream edge and at upland ephemeral streams and ponds. M y study was limited to the breeding season; work in other seasons is also needed. The relative importance of riparian and upland habitats may vary greatly during periods of migration (Harris 1984, Wiebe and Martin 1998). For neotropical migrant birds, riparian habitats may provide critical habitat during long migrations to wintering grounds (Stevens et al. 1977; Finch 1991) and their availability may also be important in the survival and population stability of migratory species. Migrants may show greater density and diversity in riparian corridors due to a greater ease of movement along waterways or a more diverse foraging habitat (Wiens 1989, Wiebe and Martin 1998). Alternatively, some migratory bird species may show no preference for riparian habitats and disperse over broad areas of forest habitat.  Table 2.1. Correlations between first three P C A factor scores and vegetation variables for old-growth sites (n = Factor 1 Factor 2  Factor 3  0.27696  -0.21142  Deciduous density  -0.89388  Salmonberry cover  -0.87684  Conifer density Shrub cover  0.35326  0.90961 0.29503 -0.53636  0.29401  Snag density  0.53365 -0.57302  Downed wood  0.50497  Shrub richness  0.20432  0.77381  Forb cover  0.24894  0.38376  Forb richness Blueberry spp.cover 0.32897  0.71195  -0.54703  0.94603  0.88101  Table 2.2. Results of paired t-tests comparing vegetation structure characteristics in riparian and upland habitat (n = 4).  Variable  Riparian  Upland  Mean(SD)  Mean(SD)  Paired-t  P  Density (#/ha) Coniferous (#/ha)  247(166)  356(140)  -1.83  0.095  Deciduous (#/ha)  288(102)  59(37)  4.06  0.002  72(41)  134(96)  -1.77  0.104  27(6)  25(6)  0.80  0.440  Salmonberry (%)  7.74(7.03)  3.37(3.30)  1.65  0.126  Downed wood (Vol/ha)  2512(208)  3549(1512)  -0.94  0.369  Snags (#/ha) Shrub cover (%)  Bold font denotes significance at P < 0.10  60  C .SP 35  CN  CO o O  NO  oo  u  cn  e  CN cN  o o  NO o  o  r- co  o  d  o  O O  d  d  •<d-  oo oo  oo  d  d  d  d  d  o  CO o o  CO O O  i—i  ON  r-<  O  d  o d  o d  CN CN  ON  o  o  o  cu  co NO  r -  o  o  d  d  oo  oo  oo  2 >Oj cu  22  S «•* *S cu CS  in  in  o o  o  s? E E £ c  0-  i-H  d  ON NO  ^  o  o  O O  d  d  d  d  o  >  5 -  in  d  d  W  co  c  .SP cn  o o  o  *o  o  o  CO NO o  NO o o  o  d  d  d  o d  NO  oo oo  OO  d  d  d  d  d  o  CO o  CO o  T—1  ON O O  d  d  d  d  cN <N  ON  CO  oo  o  o o  <u u cu cd  s s  2 '-5  o o ©  CO  o  NO  oo  o  O d  d  o< ^ > £ H  CO  d  o  o  i—i  o  oo m  oo o  o  d  d  d  o  in O  d  d  oo  ON NO  o  I—i  o d  O d  .  CU  CU  u  e  >  88  .5  E  .2 °E  CS  >  c  CU  -a a o U  o  cu  .o c o  cd  >  a  "cd  1=1  CO 0 0  _  12  O  cs  > o  2;  cd  >  o  o  3 co  > N  T3  ii  3  •8 3  s  o 3  T3  g  >  00 cd C  •a fc;  Q  CO  'o  CO  c cd  > o  £  CU  Q  a  3  o )*  CD 60 cd  CJ CD  o  *8 cu a  cn  cu cu -O cd  cu  3  o  *-> cd cj >, cu  c o  T3 CU  oo  >  T3  o cd  <  2  £ cu  J3  U  O  ta  P.  o CO CJ  3  cd  PH  CO  cd  > o  30  Figure legends Figure 2.1. Species abundance distributions for 50 intervals from the river edge. Species are ordered on the x-axis based on the order of abundance for the 0-50m interval for comparison purposes. 0-50m (a), 51-100m (b), 101-150m (c) and 151-200m (d). Clear bars indicate riparian specialists. Refer to appendix 1 for full bird names.  Figure 2.2. Mean abundance and standard deviations at 50m intervals from the river edge for total individuals (a), selected guilds: riparian specialists (b), aerial hawkers (c), foliage gleaners (d) and individual species: American robin (e), Chestnut-backed chickadee (f), Golden-crowned kinglet (g), Hairy woodpecker (h), Hammond's flycatcher (i), Pacific-slope flycatcher (j), Swainson's thrush (k), Varied thrush (1) and Winter wren (m).  Figure 2.3. Means and standard deviations of species richness at 50m intervals from the river edge for 1996 - 1998.  VM3A OXdS TiSO IAHII dSCM HIM VX3M VMOX dSOS 1JON OMId OlOD nfaa  raaa IdPVV flNSH  ^ vsds rums Horva aaaa  E  o  LO i  o  aoia vrxs ISId aivoa vsaa IAVM OMVH HXVA  E o o  LO  lavH  raao iaaA\  HAVIA\ H3H3 HXMS OMIW  (0 O  CM CD  LO O LO ••- - i - O d CD CD  o  O  ci  O C\J  T-  1 O i-  o  o  o  I  eoirepunqe uojjjodojd  1 LO  1—  LO  o o  oo o  32  VM3A OXdS  VM3A OXdS IdSO IAftH dSOd TUM VX3M VMOX dSOS laON OMia OlOD  aaso  IAiXH  jsoa  OHM 4 VX3M VMOX dSOS xaoN OMId  oaoo nfaa raaa  nraa  ICDATV ONSH VSdS aiiriM  iawv flNffH VSdS  E o  nnnn  LO  E o o  aorra aaaa aaaa aoia vrxs  CM I  1  LO  vrxs isia awoo vsa>i  ISId  swoo vsaa  P IAVA\ OMVH HXVA 1JVH  c=) I A V M OMVH HXVA  TaaA\ »A\IM Hoao HXMS OHIMV  aaaM HMIM Hoao HXMS OHIMV  D; 'LTVH  moo  moo  I  O CM  d  1  1  1—  LO O LO i- T - O  d  d  d  o o d  •a  i O CM CD  eouepunqe uojjjodojd  1  1  r~  LO O LO O tTO O CD d> <5 ci  33  b)  a) 25.00n 20.00 15.00 10.00 5.00 0.00  Total abundance  4.00-j  " T  T  3.00-  l IIl  0-50m  l • •  0-50m  51 -100m 101 -150m151 -200m  c)  Foliage gleaners 10.00T  n  CD -.—.  1.00  3  A  8.00  3 ^ 2.00  51-100m 101-150m 151-200m  d) Aerial hawkers  4.00  Riparian specialists  i i i i  0-50m  51-100m 101-150m 151-200m  6.00 4.00 2.00  0-50m  Distance from river  51-100m 101-150m 151-200m  34  h)  g) 2.00 1.50 1.00 0.50 0.00  Golden-crowned kinglet  llll 0-50m  Hairy woodpecker  1.00 0.80 0.60 0.40 0.20 0.00  0-50m  51-100m 101-150m 151-200m  51-100m 101-150m 151-200m  J) Pacific-slope flycatcher  Hammond's flycatcher 2.00  ItUi  1.501.00 0.50 0.00 0-50m  51-100m 101-150m 151-200m  0-50m  Swainson's thrush 1.40  Varied thrush  1.20 1.00 0.80 0.60 0.40 0.20 0-50m  51-100m 101-150m 151-200m  51-100m 101-150m 151-200m  0.00 0-50m  ill  51-100m 101-150m 151-200m  m) 3.50  Winter wren  3.00 2.50 2.00 1.50 1.00 0.50 0.00  liiii 0-50m  51-100m 101-150m 151-200m  Distance from river  Distance from river  36  Literature cited Beaver, D.L. and Baldwin, P.H. 1975. Ecological overlap and the problem of competition and sympatry in the Western and Hammond's flycatchers. - Condor 77: 1-13.  Bibby, C.J., Burgess, N.D. and Hill, D.A. 1992. Bird census techniques. - Academic Press, Toronto, Ontario.  Brinson, M . M . , Swift, B.L., Plantico, R.C. and Barclay, J.S. 1981. Riparian ecosystems: their ecology and status. - Department of the Interior, U.S. Fish and Wildlife Service, Kearneysville, W V .  Brosofske, K.D., Chen, J., Naiman, R.J. and Franklin, J.F. 1997. Harvesting effects on microclimatic gradients from small streams to uplands in western Washington. Ecol. Appl. 7: 1188-1200.  Bryant, A.A., Savard, J.P. and McLaughlin, R.T. 1993. Avian communities in old-growth and managed forests of western Vancouver Island, British Columbia. - Canadian Wildlife Service, Nanaimo, British Columbia.  Campbell, R.W., Dawe, N.E., McTaggart-Cowan, I., Cooper, J.M., Kaiser, G.W. and McNall, M.C. 1990. The birds of British Columbia. - The Royal British Columbia Museum, Victoria, B.C.  Campbell, R.W., Dawe, N . K , McTaggart-Cowan, I., Cooper, J.M., Kaiser, G.W., McNall, M.C.E. and Smith, G.E.J. 1997. The birds of British Columbia. - U B C Press, Vancouver.  Carothers, S.W., Johnson, R.R. and Aitchison, S.W. 1974. Population structure and social organization of southwestern riparian birds. - Am. Zool. 14: 97-108.  37  Conover, W.J. and Iman, R.L. 1982. Analysis of covariance using the rank transformation. - Biometrics 38: 715-724.  Darveau, M . , Beauchesne, P., Belanger, L., Huot, J. and Larue, P. 1995. Riparian forest strips as habitat for breeding birds in boreal forest. - J. Wildl. Manage. 59: 67-78.  Dayton, P.K. 1998. Reversal of the burden of proof in fisheries management. - Science 279: 821-822.  Dickson, J.G. 1978. Forest bird communities of the bottomland hardwoods. - In: DeGraaf R . M . (eds), Proceedings of the workshop management of southern forests for nongame birds. U S D A F o r . Serv. Gen. Tech. Rep., Atlanta, Georgia, pp. 66-73.  Ehrlich, P.R., Dobkin, D.S. and Wheye, D. 1988. The birder's handbook: a field guide to the natural history of North American birds. - Simon and Schuster Inc., New York.  Enemar, A., Sjostran, B. and Svensson, S. 1978. The effect of observer variability on bird census results obtained by a territory mapping technique. - Ornis Scand. 9: 31-39.  Enns, K.A., Peterson, E.B. and McLennan, D. 1993. Impacts of hardwood management on British Columbia wildlife: problem analysis. - F R D A II, Victoria, B.C.  Fahrig, L . and Merriam, G. 1985. Habitat patch connectivity and population survival. Ecology 66: 1762-1768.  Finch, D . M . 1991. Population ecology, habitat requirements, and conservation of Neotropical migratory birds. - In: Finch D . M . (eds), USD A For. Serv. Gen. Tech. Rep., Ft. Collins, Colorado.  38  Fisher, R.A., Corbet, A.S. and Williams, C.B. 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. J. of Anim. Ecol. 12: 42-58.  Gonzalez, A. 2000. Community relaxation in fragmented landscapes: The relation between species richness, area and age. - Ecology Letters 3: 441-448.  Gray, L.J. 1993. Response of insectivorous birds to emerging aquatic insects in riparian habitats of a tallgrass prairie stream. - Am. Midi. Nat. 129: 288-300.  Hagar, J.C. 1999. Influence of riparian buffer width on bird assemblages in western Oregon. - J. Wildl. Manage. 63: 484-496.  Hansson, L. 1983. Bird numbers across edges between mature conifer forest and clearcuts in central Sweden. - Ornis Scand.14: 97-103.  Harris, L.D. 1984. The fragmented forest. - The University of Chicago Press, Chicago.  Hatler, D.F., Campbell, R.W. and Dorst, A. 1976. Birds of Pacific Rim National Park. British Columbia Provincial Museum, Victoria.  Helle, P. and Muona, J. 1985. Invertebrate numbers in edges between clear-felling and mature forests in northern Finland. - Silva Fennica 19: 281-294.  Holmes, R.T. and Robinson, S.K. 1981. Tree species preferences of foraging insectivorous birds in a northern hardwoods forest. - Oecologia 48: 31-35.  Hopkins, R.B., Cassel, J.F. and Bjugstad, A.J. 1986. Relationships between breeding birds and vegetation in four woodland types of the Little Missouri National Grasslands. - In: Gen. Tech. Rep. U S D A For. Serv., Fort Collins, pp. 1-11.  39  James, F.C. and Shugart, H.H. 1970. A quantitative method of habitat description. Audubon Field Notes 24: 727-736.  James, F.C. and Warner, N.O. 1982. Relationships between temperate forest bird communities and vegetation structure. - Ecology 63: 159-171.  Kilgo, J . C , Sargent, R.A., Chapman, B.R. and Miller, K . V . 1998. Effect of stand width and adjacent habitat on breeding bird communities in bottomland hardwoods. - J. Wildl. Manage. 62: 72-83.  Kinley, T.A. and Newhouse, N.J. 1997. Relationship of riparian reserve zone width to bird density and diversity in southeastern British Columbia. - Northwest Sci. 71: 7586.  Klinka, K., Pojar, J. and Meidinger, D.V. 1991. Revision of biogeoclimatic units of coastal British Columbia. - Northwest Sci. 65: 32-78.  Knopf, F.L. and Samson, F.B. 1994. Scale perspective on avian diversity in western riparian ecosystems. - Conserv. Biol. 8: 669-676.  Kuehne, R.A. 1962. A classification of streams, illustrated by fish distribution in an eastern Kentucky creek. - Ecology 43: 608-614.  LaRue, P., Belanger, L. and Huot, J. 1995. Riparian edge effects on boreal Balsam Fir bird communities. - Can. J. Forest Res. 25: 555-566.  Legendre, L . and Legendre, P. 1983. Numerical ecology. - Elsevier, Amsterdam.  MacArthur, R.H. and MacArthur, J.W. 1961. On bird species diversity. - Ecology 42: 594-598.  40  McGarigal, K. and McComb, W.C. 1992. Streamside versus upslope breeding bird communities in the central Oregon Coast Range. - J. Wildl. Manage. 56: 10-23.  Mueller-Dombois, D. and Ellenburg, H . 1974. Aims and methods of vegetative ecology. - John Wiley & Sons, Inc., New York.  Murray, N.L. and Stauffer, D.F. 1995. Nongame bird use of habitat in central Appalachian riparian forests. - J. Wildl. Manage. 59: 78-88.  Naiman, R.J., Decamps, H . and Pollock, M . 1993. The role of riparian corridors in maintaining regional biodiversity. - Ecol. Appl. 3: 209-212.  Nuszdorfer, F. and Boettger, R. 1994. Southern Vancouver Island and Sunshine Coast. Province of British Columbia, Ministry of Forests. Biogeoclimatic units of the Vancouver forest region, 5 of 6.1:250,000.  Pearson, S. and Manuwal, D.A. 2001. Breeding bird response to riparian buffer width in managed Pacific Northwest Douglas Fir forests. - Ecol. Appl. 11: 840-853.  Peterson, R.T., 1990. Field guide to western birds. - Boughton Mifflin Company, Boston.  Ranney, J.W., Bruner, M.C. and Levenson, J.B. 1981. The importance of edge in the structure and dynamics of forest islands. - In: Burgess R.L. and Sharpe D . M . (eds), Forest island dynamics in man-dominated landscapes. Springer-Verlag, New York, pp. 67-96.  Robbins, C.S., 1970. Recommendations for an international standard for a mapping method in bird census work. - Audubon Field Notes 24: 723-726.  Robinson, S.K. and Holmes, R.T., 1984. Effects of plant species and foliage structure on the foraging behavior of forest birds. - Auk 101: 672-684.  41  Rotenberry, J.R. and Wiens, J.A. 1980a. Habitat structure, patchiness, and avian communities in North American steppe vegetation: a multivariate analysis. - Ecology 61: 1228-1250.  Sakai, H E . and Noon, B.R. 1991. Nest-site characteristics of Hammond's and PacificSlope flycatchers in northwestern California. - Condor 93: 563-574.  Saunders, D.A. and de Rebeira, C P . 1991. Values of corridors to avian populations in a fragmented landscape. - In: Saunders D.A. and Hobbs R.J. (eds), The role of corridors. Surrey Beatty & Sons Pty Ltd., Western Australia, pp. 221-240.  Stauffer, D.F. and Best, L . B . 1980. Habitat selection by birds of riparian communities: evaluating effects of habitat alterations. - J. Wildl. Manage. 44: 1-15.  Stevens, L.E., Brown, B.T. Simpson, J.M. and Johnson, R.R. 1977. The importance of riparian habitat to migrating birds. - In: Johnson R.R. and Jones D.A. (eds), Importance, preservation and management of riparian habitat: a symposium. U S D A Forest Service, Tucson, Arizona, pp. 156-164.  Strickler, G.S. 1959. Use of the densiometer to estimate density of forest canopy on permanent sample plots. - Pacific Northwest Forest and Range Experiment Station, Portland, Oregon.  Strong, T.R. and Bock, C E . 1992. Bird species distribution patterns in riparian habitats in southeastern Arizona. - Condor 92: 866-885.  Szaro, R . C 1980. Factors influencing bird populations in southwestern riparian forests. In: DeGraff R . M . (eds), Workshop management of western forests and grasslands for nongame birds. U S D A For. Ser. Gen. Tech. Rep. INT-86, pp. 403-418.  42  Thomas, J.W., Maser, C. and Rodiek, J.E. 1979. Riparian zones. - In: Thomas J.W. (eds), Wildlife habitats in managed forest - the Blue Mountains of Oregon and Washington. U.S. Forest Service, Washington, D.C., pp. 40-47.  Toft, C.A. and Shea, P.J. 1983. Detecting community-wide patterns: estimating power strengthens statistical inference. - Am. Nat. 122: 618-625.  Van Wagner, C.E. 1968. The line intersect method in forest fuel sampling. - Forest Science 14: 20-26.  Waterhouse, F.L. and Harestad, A.S. 1999. The value of riparian forest buffers as habitat for birds in the coastal western hemlock zone, British Columbia: a pilot study. Research Section, Vancouver Forest Region, British Columbia Ministry of Forests, Nanaimo, B.C.  Wauer, R.R. 1977. Significance of Rio Grande riparian systems upon the avifauna. - In: Johnson R.R. and Jones D.A. (eds), Importance, preservation and management of riparian habitat: a symposium. U S D A Forest Service, Tucson, Arizona, pp. 165-174.  Wiebe, K . L . and Martin, K. 1998. Seasonal use by birds of stream-side riparian habitat in coniferous forest of northcentral British Columbia. - Ecography 21: 124-134.  Wiens, J.A. 1989. The ecology of bird communities. - Cambridge Univ. Press, Cambridge.  Williams, C.B. 1964. Patterns in the balance of nature. - Academic Press Inc., London.  Willson, M.F. and Comet, T.A. 1996. Bird communities of northern forests: patterns of diversity and abundance. - Condor 98: 337-349.  43  Appendix 1 - Classification of bird species into migratory status and foraging guild summarized by habitat guild (from Hatler et al., 1978, Ehrlich et al., 1988, Peterson, 1990 and Campbell et al., 1990, 1997). Common Name  Scientific Name  Migratory Foraging Status Guild A  B  Forest Generalists CBCH GCKI HAWO HUVI PISI RBSA RECR SWTH TOWA VATH WETA WIWR  Parus rufescens Regulus satrapa Picoides villosus Vireo huttoni Carduelis pinus Sphyrapicus ruber Loxia curvirostra Catharus ustulatus Dendroica townsendi Ixoreus naevius Piranga ludoviciana Troglodytes troglodytes  R R R R R R R N N R N R  FG FG BG FG FG BG FG FG FG GG FG GG  Brown creeper Pileated woodpecker Pacific-slope flycatcher Red-breasted nuthatch Ruffed grouse  Certhia familiaris Dryocopus pileatus Empidonax dijficilis Sitta canadensis Bonasa umbellus  R R N R R  BG BG AH BG FB  Blue grouse Dark-eyed junco Fox sparrow Northern flicker Olive-sided flycatcher Rufous hummingbird Song sparrow Spotted towhee Steller's jay Willow flycatcher  Dendragapus obscurus Junco hyemalis Passerella iliaca Colaptes auratus Contopus borealis Selasphorus rufus Melospiza melodia Pipilo erythrophthalmus Cyanositta stelleri Empidonax traillii  R R R R N N R R R N  GG GG GG GG AH HG FG GG GG AH  Cinclus mexicanus Megaceryle alcyon  R R  RG HD  Chestnut-backed chickadee Golden-crowned kinglet Hairy woodpecker Hutton's vireo Pine siskin Red-Breasted sapsucker Red crossbill Swainson's thrush Townsend's warbler Varied thrush Western tanager Winter wren  Forest Interior BRCR PIWO PSFL RBNU RUGR Open-Edge BLGR DEJU FOSP NOFL OSFL RUHU SOSP SPTO STJA WIFE  Riparian Specialists AMDI BEKI  American dipper Belted kingfisher  44  COLO COME HAFL SPSA WAVI YEWA  Common loon Common merganser Hammond's flycatcher Spotted sandpiper Warbling vireo Yellow warbler  Gavia immer Mergus merganser Empidonax hammondii Actitis macularia Vireo gilvus Dendroica petechia  R R N N N N  SD SD AH GG FG FG  American robin  Turdus migratorius  R  GG  Ubiquitous AMRO  A  Migratory Status: N = Neotropical migrant, R = Resident  Foraging: A H = Aerial hawking, B G = Bark gleaner, FB = Foliage browse, F G = Foliage gleaner, G G = Ground gleaner, H G = Hover glean, R G = Rock glean, SD = Surface dives B  45  Chapter3: Forest Fragmentation And Regrowth: Use Of Riparian Buffer Strips By Birds In Mature Coastal British Columbia Rainforest INTRODUCTION  Buffer strips are strips of forest retained along streamsides after harvesting to mitigate negative impacts of forestry on aquatic fauna and water quality. Although they may be connected to larger patches of forest at one or both ends, buffer strips resemble isolated forest patches and species inhabiting these habitats before harvesting may demonstrate negative effects of forest loss and fragmentation (Darveau et al., 1995, Hagar, 1999, Whitaker and Montevecchi, 1999). Because they have high amounts of edge relative to the total surface area, narrow buffer strips are expected to support fewer forest interior species and perhaps fewer riparian specialists. Conversely, open area specialists may increase while numbers/richness of forest generalists may not change. Forest birds are good indicators of habitat degradation since they are easy to census and respond to habitat fragmentation (review in Saunders et al., 1991). Several mechanisms may contribute to declines of forest-interior bird species (Ambuel and Temple, 1983) in small forest fragments. For example, high amounts of edge habitat in small fragments may increase bird mortality by increasing nest parasitism (Gates and Gysel, 1978; Brittingham and Temple, 1983) and predation (Wilcove, 1985). Changes in microclimate (Chen et al., 1993) within forest fragments may reduce food levels (Burke and Nol, 1998; Zanette et al., 2000).  Early studies of minimum buffer widths focused on impacts to the stream ecosystem including sedimentation, alterations to stream structure, and fisheries (Newbold et al., 1980; Budd et al, 1987). Little consideration was given to terrestrial wildlife communities and the few assessments that were made did not consider individual species requirements (e.g. Budd et al., 1987). More recent studies in the northeastern U.S.A. have recommended that riparian zones of at least 100m are needed to preserve Neotropical migrant species (Keller et al., 1993; Hodges and Krementz, 1996). Buffers of 175m retain 90% of bird species (Spackman and Hughes, 1995). Darveau et al. (1995)  46  showed that buffers > 60m wide were required to protect certain boreal forest birds. Given the variation in geography and landscape context, these varied recommendations are not surprising. Bird community response is likely to depend on the difference in vegetation structure and composition between riparian forest and upland habitats. Buffer strips in regenerating forested landscapes after harvesting are likely to have different value than those in agricultural landscapes (Stauffer and Best, 1980) because resources in the regenerating forests may complement those in the buffer. Buffers retained in agricultural landscapes may require wider buffer strips to protect bird communities. Here I examine buffer strips that were established 5 years prior to the study. Little information exists on isolated fragments in forested landscapes that have likely already experienced some faunal relaxation. Effects of varying buffer strip widths on bird richness and abundance are likely to fluctuate with time since harvest, but the dynamics over time with regeneration are not well known. By studying fragments such as these, that have been isolated for some time, we can determine if species richness reaches an equilibrium that is steady over several years, or if they continue to lose or recover species over time.  The relationship between species richness and forest fragment area in riparian habitats has also not been clearly established. Stauffer and Best (1980) found that richness in Iowa increased with the width of riparian woodland. However, work by Decamps et al. (1987) along the River Garonne in southwest France showed no effect of buffer width on the number of bird species in riparian woodlands. They attributed the lack of effect to the high moisture content of riparian forest soils that permits high forest bird densities, even when buffers are very isolated. Other evidence, however, suggests that some forestinterior species in forested landscapes become rare or absent in narrow strips several years after harvesting (Darveau et a l , 1995, Hagar, 1999, Whitaker and Montevecchi, 1999).  Classical views of fragmentation effects (Gates and Gysel, 1978, Whitcomb et al., 1981, Lynch & Whigham, 1984, Askins, 1990) may not hold for coastal montane forests of the Pacific Northwest for two reasons. First, two studies of bird abundance and diversity (Lehmkuhl et al., 1991; Schieck et al., 1995) in this region have either  47  contradicted or only weakly supported studies in eastern North America (Gates and Gysel, 1978, Lynch & Whigham, 1984). For example, Schieck et al. (1995) found that patch size of Pacific coast old-growth forest fragments does not influence the number of forest interior species. Brown-headed cowbirds, common in eastern forests, are rarely encountered in these forests (Carey et al., 1991; Bryant et al., 1993, Schiek et al., 1995), and the abundances of nest predators such as crows and jays are not related to patch size (Lehmkuhl et al, 1991, Schieck et al., 1995, Tewksbury et al., 1998). These studies suggest that birds in buffer strips in northwestern coastal forests are not vulnerable to parasitism by cowbirds and predation by corvids. Other factors such as microclimate and nest predation by small mammals have yet to be evaluated in this area, but small mammal predation may be significant (Tewksbury et al., 1998). Second, in this area, often only one side of the river is logged leaving continuous old-growth forest across the river from the buffer strips. Birds in these strips are less isolated from primary forest as in an upland forest fragment and may use buffers as travel corridors. These two basic differences suggest that generalizations about fragments should not be applied uncritically to buffer strips in the Pacific Northwest. To date, few studies have evaluated the effects of buffer strips in forested landscapes in the Pacific Northwest (but see Hagar, 1999, Pearson and Manuwal, 2001). Previous studies in North America have focused on remnants within permanently altered landscapes (review in Kaufman and Brueger, 1984, Strong and Bock, 1992, Saab, 1999) or on very recent fragmentation events (Darveau et al, 1995, Schmiegelow et a l , 1997, Pearson and Manuwal, 2001).  I compared communities of birds in varying widths of old-growth buffer strips with equal widths of undisturbed riparian forest. My primary objective was to identify shifts in species richness, diversity and abundance between buffers and controls of equal area. I hypothesized that species area relationships for buffers should show steeper slopes and lower intercepts compared to continuous forest (Connor & McCoy, 1979, Martin, 1981) reflecting lower species richness in narrow buffers and converging to richness similar to undisturbed forest in wide buffers. Non-zero intercepts would reflect the presence of open-edge bird species when the width of forest buffers is essentially zero. Buffers will also exhibit higher species turnover (Hinsley et a l , 1995) and more variation in richness  48  and population densities relative to controls; this will be particularly strong in narrow buffers. Species assemblages in buffers will be least similar to controls in the narrow buffers and most similar to controls in wide buffers. In particular, narrow buffers will support fewer forest-interior species and more open-edge species because of edge effects (Darveau, 1995). Abundances of certain guilds such as foliage gleaners and aerial foragers will be higher in buffers than in controls due to higher insect abundances (Helle and Muona, 1985, Hansson, 1994) and the growth of dense shrubs and deciduous trees at the forest-clearcut edge. Riparian specialists, because they typically maintain linear territories along the riverbank (Manuwal, 1986, Whitaker and Montevecchi, 1999) and do not increase with additional forest area (Whitaker and Montevecchi, 1999), should show little effect of fragmentation. I also hypothesized that these buffers would support more species and greater abundance of birds over time due to regrowth of vegetation in the adjacent clearcut areas. If regeneration of the adjacent clearcuts facilitates movement between remnant patches (Stouffer and Bierregaard, 1995, Kapos et al., 1997) by providing resources and cover, then increases in species richness and abundance in buffers should be positively correlated with similar increases in the adjacent clearcuts.  METHODS  Study area  The study sites lay within 9 watersheds on the west coast of Vancouver Island, British Columbia (Table 3.1, Fig. 3.1a). They cover an area between Ucluelet in the north and Bamfield in the south (Latitude, 48°5', Longitude, 125°5'). Major rivers include the Nahmint, Taylor, Kennedy and Klanawa. The area is in the Western Vancouver Island ecoregion within the coastal western hemlock moist to very-wet maritime biogeoclimatic subzones (Klinka et al., 1991,-Nuszdorfer and Boettger 1994). Amabilis fir {Abies amabilis) characterizes these subzones (Packee, 1972, Klinka et al., 1991) and occurs even at low elevations (200m). Other dominant species include western hemlock (Tsuga  49  heterophylla) and western red cedar (Thuja plicata) with Douglas-fir (Pseudotsuga menziesii) on southern aspects. Biomass accumulations in these forests are large and the undisturbed forest is highly stratified with a dense understorey, many snags, large amounts of coarse woody debris, and many mosses and lichens. Average precipitation in the area is 3,100 mm most of which falls from October to March. Summers are cool and moist in May/June with mean temperatures of 12.1° C, warm and dry in July/Aug with mean temperatures of 15.9° C (Environment Canada Climate Data Services). I selected buffers of 10-125m in width and adjacent clearcut strips of 75-190m in width found along one side of the river. Adjacent clearcuts were 3-5 years old at the start of the study. I used power analysis to calculate the number of replicates required to detect a 20% loss of species. I based the variances of the species-area relationships on those used by Schmiegelow and Hannon (1993) and modified these to include the results of that study (Schmiegelow et a l , 1997). I performed the analysis with an alpha of 0.10 and power of 0.95. Based on this analysis, I selected three replicates of each width category along with three control sites of undisturbed continuous riparian forest at each width for a total sample size of 24. All sites were at similar elevation (< 500m) estimated from 1:20,000 topographic maps. I define buffer categories as very narrow (010m), narrow (20-25m), medium (36-44m) and wide (100-144m).  Bird sampling  Birds were censused using the spot-mapping method (Robbins 1970, Bibby et al. 1992). A 9 ha. grid was established at each site by running a 450m transect line parallel to the river edge and nine perpendicular lines 200m from the river (Figure 3.1b). Transect lines were set 50m apart and flagged at 25m intervals. Spot-mapping was conducted at each site by at least two observers who walked the transect lines from 5:00am-l 1:00 am. Censuses were not conducted during heavy rain and/or strong winds. Birds were mapped four times through the 1996-1998 breeding seasons. There were three observers during 1996 and four for 1997-1998. Two of the observers worked in 1997 and 1998 and one observer was present in all years. The order in which sites were sampled was varied to reduce bias and observers were rotated among sites and among  50 transect lines. During mapping, we recorded microhabitat information for each record, describing the forest structure where the bird was first heard or observed, e.g. salmonberry shrub, and its approximate location. Spot maps were evaluated to produce data on number and abundance of bird species following Robbins (1970) except that the number of registrations/observations was used rather than number of territories (Enemar, 1978).  Data analyses  Before conducting statistical tests, all data were tested for normality using the ShapiroWilk statistic and analysis of standardized residuals. Homogeneity of variances was tested by analyzing residuals in the linear regressions and by the Bartlett-Box F statistic in the analysis of covariance procedures. Data that violated the assumptions substantially were transformed. I used significance level of 0.10 because of the high conservation cost of making a type II error (Smith, 1995, Dayton, 1998). Data were analyzed using SPSS for Windows 6.1.4 (1996).  Species-area Relationships. - To calculate species richness I included all species observed on >1 census session during each year. Species seen only once were excluded from the analysis to minimize the impacts of transient species and migrants (Willson et a l , 1994). Species-width relationships were produced using linear regressions using an exponential model with logio(x + 1) transformed width (Williams, 1964). A n alternative power function model (Connor and McCoy, 1979) fitted the data slightly less well. Species-width was used as a surrogate for species-area because width is a measure of edge effects and all transects parallel to the river were standardized to 450m in length.  In this paper, I evaluated the effects of habitat fragmentation at the patch level by comparing buffer strips with equal-sized sections of a continuous riparian forest. The subsamples from the larger forest were used to generate a species-area curve expected under the null hypothesis of undisturbed forest conditions (Simberloff, 1976, McGuinness, 1984). I compared species-area relationships between controls and buffers  51  using analysis of covariance ( A N C O V A , Sokal and Rohlf, 1998) with buffer width as a covariate and between years using repeated measures analysis of covariance (rmANCOVA, Scheiner & Gurevitch, 1993). I compared differences in species richness between years and among treatments with polynomial contrasts (Gurevitch and Chester, 1986, von Ende, 1993). To examine trends from 1996-1998,1 tested for linear (first order) and quadratic (second order) changes in richness. Where there were significant interactions between buffer vs. control and year, I performed the rmANCOVAS separately for buffers and controls. To compare richness and abundance in specific buffer widths between years, I used Friedman's test (Friedman, 1937). I used linear regression to compare species richness across a range of clearcut widths for each of the three years and r m A N O V A to examine changes in species richness in the clearcuts over time. To test if changes in species richness in buffers were correlated with those in clearcut habitats, I used the Pearson's correlation coefficient.  Bird Abundance. -1 estimated abundance for each species and site by the mean number of detections over the four census periods of each year. As with species richness, I included only those species observed on more than one census during each year. I analyzed abundances both by guild and by individual species. For the guild analysis, I grouped species abundances into foraging and habitat guilds, and migratory behavior, and performed linear regressions of guild abundance vs buffer width. I classified species into guilds using published information (Hatler et al., 1978, Ehrlich et al., 1988, Peterson, 1990, Campbell et a l , 1990, 1997) (Appendix 1). For individual taxa, I calculated regressions for species with greater than 5 observations /year. Data for any species that did not meet the assumptions of normality and homogeneity of variance were transformed using either the square root (Bartlett, 1936) or rank transformations (Conover and Iman, 1982).  I compared the regression slopes and intercepts for both guilds and species between buffers and controls with A N C O V A and between years with r m A N C O V A . Where there were significant interactions between buffers vs. controls and year, I performed the r m A N C O V A analyses on buffers and controls separately. I examined the relationship  52 between guild abundance and clearcut width using linear regression and compared these relationships between years using rmANOVA.  Species diversity. - Plots of species abundance curves (Whittaker, 1965) for buffers and controls for 1996-1998 indicated that diversity could be estimated by the log-series distribution (Fisher et al., 1943), which relates the number of species to the number of individuals in a sample as  S = cdogeQ + N/tx )  where a is species diversity (Fisher et al, 1943). To compare buffers and controls, I calculated a values for all sites (Williams, 1964).  I compared the diversity of buffers and controls using A N C O V A and among years using r m A N C O V A . Where there were significant interactions between fragment type and year, I performed the r m A N C O V A analysis on buffers and controls separately. Since changes in diversity may be associated with changes in the number of rare species, I performed similar analyses using the proportion of rare species which I calculated as the proportion of species with < 5% of total abundance.  Species turnover. - To test the prediction that turnover would differ between buffers and controls, I calculated the similarity in species composition within each plot for both buffers and controls from 1996-1998 using the Jaccard coefficient, Sj. I used the formula (1- Sj) to calculate species turnover for the same period and regressed Sj on log buffer width. Finally, I compared the regression slopes and intercepts using A N C O V A . Community similarity. -1 compared the assemblage of birds in buffers with continuous forest using Horn's index of similarity (Horn, 1966), a quantitative index that uses relative abundance data. I used Program SIMILAR (Krebs, 1989).  53  RESULTS  Bird communities in buffers and controls  Over the 3 years, I observed 43 species in the buffers and 28 in the controls. The six most common species represented 64% of total observations in the buffers and 69% in the controls. The species in the buffers, listed in descending order of abundance, were: Swainson's thrush, chestnut-backed chickadee, winter wren, American robin, Hammond's flycatcher and golden-crowned kinglet. In the controls the species were: American robin, winter wren, chestnut-backed chickadee, Swainson's thrush, Pacificslope flycatcher and golden-crowned kinglet.  Species-area relationships and diversity. - The slopes of the relationship between number of species and area differed for buffers and controls in 1996, but not in 1997 or 1998 (Figure 3.2). The controls had a steeper slope and lower intercept in 1996. Slopes ranged from 6.0 - 9.1 except for the 1996 control (11.3). The difference in 1996 was significant only in wide fragments where the wide control contained 17 species on average and the wide buffer contained 12 species on average (U = 0.5, P = 0.077, df = 2). However, this pattern was reversed in 1998 where wide buffers had 17 species compared to 14 in controls. Intercepts were significantly different in 1998 with buffers having 23% more species than controls. Variation in species richness decreased with width in both buffers and controls (Buffers: F,  J0  = 7.94, P = 0.021, Controls: F o = 4.74, P = 0.054), U  but did not differ between buffers and controls ( A N C O V A , F i lt2  = 0.50, P = 0.486).  Patterns of species richness differed over time in buffers and controls (rmANCOVA, F ,44 = 5.48, P = 0.007). There were no significant temporal changes in controls 2  (rmANOVA, F ,22 = 0.71, P = 0.351); however, buffers increased in species richness 2  from 1996-1998 (rmANOVA, F , = 5.92, P = 0.007) driven mainly by a 42% increase 2 22  in richness in the wide buffers over time (Friedman's X = 4.91, P = 0.086, df = 2). 2  54  The results for species diversity closely matched those for species richness. Diversity differed significantly between buffers and controls in 1996, where diversity increased more slowly in buffers. There were no differences inl997 and 1998 (Figure 3.3). Diversity increased in buffers over the three-year period (rmANOVA, ^2,22 = 9.10, P = 0.012), primarily due to an increase in diversity in wide buffers (Friedman's X = 4.67, P 2  = 0.10, df = 2), but remained constant in controls (rmANOVA, F  2y22  = 1.22, P = 0.293).  The proportion of rare species (those with <5% of total abundance) increased with area in both buffers and controls for all years and these proportions did not differ significantly between buffers and controls in any year ( A N C O V A , all P values > 0.35). However, the proportion of rare species increased 53% in buffers over time (rmANOVA, F , 2 - 8.02, 2 2  P = 0.016) while remaining constant in the controls (rmANOVA, F222 = 0.45, P = 0.514) consistent with increases in species diversity in buffers. The increases in buffers were driven by a 51% increase in the proportion of rare species in wide buffers (Friedman's X = 4.67, P = 0.10, df = 2). 2  Bird abundance and density. - There was a positive relationship between total abundance and buffer width for all years. Abundance increased slowly across narrower buffers, and rapidly in the widest buffers (Figure 3.4). Total abundance was similar between buffers and controls in all years (Figure 4) and showed a significant increase in both groups over the three-year period (27% for buffers, 12% for controls) (rmANCOVA, F ,  2 2]  = 13.77, P = 0.001). Bird density declined with width in buffers in  1996 and 1998 (1996: F  uo  = 10.26, P = 0.009, 1998: F  uo  = 10.34, P = 0.009) and in  controls during 1997 (1996: Fijo = 4.61, P = 0.057). Variation in density also declined with width for both buffers and controls (Buffers: Fijo = 23.64, P = 0.001, Controls: Fo U  = 9.97, P = 0.010).  Bird Abundance by Guilds. - Neotropical migrants made up 37% of all detections in buffers and 31% in controls. Numbers of Neotropical migrants in buffers were similar to those in controls for all years, and there were no differences in abundance of either buffers or controls across years (Table 3.2, Fig. 3.5a). Residents did not differ between  55 buffers and controls for any year (Table 3.2); however, abundance for both buffers and controls (Fig. 3.5a) did increase over time (36% in buffers, 21% in controls).  The slopes for abundance against width of aerial foragers differed between buffers and controls in 1996 with controls having a steeper slope; however, there were no differences between buffers and controls in 1997 and 1998 (Table 3.2). For both buffers and controls, there was a significant increase in abundance of aerial foragers (37% in buffers, 15% in controls) over the three years (Table 3.2). Foliage gleaners made up 37% of all detections in controls and 42% in buffers. Abundance of foliage gleaners was significantly higher (17%) in the buffers than in controls in 1998 only. They showed a significant (36%) increase in buffers across years, but a significant (13%) decrease over time in controls (Table 3.2, Fig. 3.5b). Ground gleaners were significantly more abundant (39%) in buffers than controls during 1996, but not during 1997-98 (Table 3.2). Ground-gleaner abundance showed a small, but significant decrease in the buffers (3%) (Fig. 3.5b) over the three years and a large increase in the controls (80%).  Both forest interior and open-edge species differed significantly in abundance between buffers and controls (Table 3.2). Abundances of forest interior species on controls averaged 50% higher compared to buffers; these species (see Appendix 1) were rare or absent in narrow and very narrow buffers. Conversely, abundances of open-edge species in buffers were almost 4 times higher than in controls (7% and 3% of total detections respectively) (Appendix 2). These species (Appendix 1) were at low densities in the medium and wide buffers (Appendix 2). During the study, abundances of forest interior species abundance increased in both buffers (81%) and controls (24%) while abundances of open-edge species increased in buffers (3%) and decreased in controls (73%) (Fig.3.5c). Abundances of riparian specialists were similar in buffers and controls for 1996 and 1997, but were 2 times greater in buffers across all widths in 1998 (Table 3.2). During 1996-1998, riparian specialists increased by 57% in buffers yet decreased 38% in controls (Fig. 3.5c). The patterns were similar after excluding the dominant riparian species, Hammond's flycatcher except that abundances were 2.4 times higher in controls compared to buffers for 1997 ( A N C O V A , F/, ; = 13.58, P = 0.001) and there 2  56  was no difference between the two in 1996 or 1998 (1996: A N C O V A , F i h2  0.409, 1998: A N C O V A , F  1M  = 0.71, P =  = 0.67, P = 0.422).  Individual Species. - There were enough observations (>3 detections in 2 of 3 years) to compare abundance between buffers and controls across different widths and years for 13 species (Tables 3.2, 3.3).  Most species showed a positive relationship between abundance and buffer width for both buffers and controls in at least one of the three years (Appendix 2). Notable exceptions were red-breasted sapsuckers, which showed no relationship with width in either fragment type, and varied thrushes which had no relationship to width in buffers, but a positive relationship with width in controls.  Three species (the American robin, golden-crowned kinglet and Pacific-slope flycatcher) were more abundant in controls than in buffers in at least one year; the flycatcher was more abundant in controls in both 1996 and 1998 (Table 3.2). Another three species: Hammond's flycatcher, Swainson's thrush and warbling vireo had higher abundances in buffers during at least one of the three years; Hammond's flycatchers showed this pattern in both 1997 and 1998.  Seven species showed significant temporal trends over the three-year period (Table 3.3). The golden-crowned kinglet and winter wren increased in both buffers and controls, while the warbling vireo decreased. The warbling vireo showed a larger decrease in controls (70%), than in buffers (8%). Three species (American robin, varied thrush, pacific-slope flycatcher) increased in controls only, while the chestnut-backed chickadee decreased in controls only.  Species Turnover and Community Similarity - Species turnover from 1996-1998 was highest in narrower buffers and control strips (Figure 3.6). Turnover decreased linearly from 80-88% over 3 years in the very narrow widths and 40-42% in wide widths (Buffers: regression F jo = 27.70, P = <0.001, Old-growth: regression Fijo = 15.42, P = {  57  0.003). There was no significant difference in turnover between buffers and controls of the same width ( A N C O V A , F,, = 0.00, P = 0.973). 21  Assemblages of species were least similar between buffers and controls at the very narrow width. Similarity between buffers and controls increased with width up to the medium width (41m) (Figure 3.7) and then leveled off. This pattern was consistent among years except that very narrow buffers showed high variation between years. A significant linear relationship between similarity and width was apparent in 1997 (Regression F  uo  = 9.92, P = 0.088).  Changes to bird communities in the clearcut habitat  Species richness - Linear regressions showed no significant relationship between species richness and width for clearcut strips in any year (all P-values > 0.14). However, there was a strong increase in species richness across all strip widths over the three-year period (rmANOVA, F 22 2i  = 31.32, P = <0.001). This change in species richness was  positively correlated with similar trends in adjacent buffer habitats (Figure 3.8)  (vp rson = ea  0.67, P = 0.018, df= 10).  Guild Abundance - The expected positive relationship of abundance with area (MacArthur and Wilson, 1967) was not found for total abundance or within several guilds. Total abundance was positively related to clearcut width in 1998 only. Several guilds showed a positive relationship between abundance and width in one year only (Appendix 3). Among foraging guilds, foliage gleaners showed a positive relationship in 1998 respectively while ground gleaners showed a positive relationship in both 1996 and 1998. Aerial foragers showed no relationship in any year. Within migratory classes, residents had positive relationships with width in all years, while Neotropical migrants had a positive relationship for 1998 only. In the habitat guild, open-edge and ubiquitous species had positive relationships with width in 1998 and 1996 respectively, while forest generalists showed no relationship in any year.  58  Guilds showed a strong increase in abundance in the clearcut fragments over the 3 years, except for aerial foragers and ubiquitous species (Fig. 3.9). Foliage gleaners showed large increases (78%) due to increases in open-edge species such as MacGillivray's warbler and common yellowthroat, while ground gleaners showed large increases for Steller's jay, white-crowned sparrow and spotted towhee. The largest increases in the habitat guilds were associated with open-edge species (63%) and a riparian species, the warbling vireo (173%).  DISCUSSION  Bird communities in buffers and controls  Species richness  I found that buffer strips did not consistently support fewer species or less diverse assemblages than unfragmented strips of forest. As expected, species richness and diversity varied both across buffer widths and among years. I predicted that narrow buffers would show the greatest differences in species richness and diversity between buffers and controls. Similar results were found in one study of short-term fragmentation effects (Schmiegelow et al., 1997). Interestingly, the largest differences between buffers and controls occurred at both the narrowest and the widest widths sampled. Species richness and diversity in the narrowest buffers was either higher or equal to that in controls. In addition, richness and diversity were lower in wide buffers than in controls in 1996, rather than remaining similar to undisturbed forest as predicted; however, this pattern was reversed in 1998.  Species-area and diversity-area relationships were positive in all years as expected (MacArthur and MacArthur, 1961, Nillson, 1986, Rosenzweig, 1995, Bellamy et al., 1996), but these relationships differed between buffers and controls in 1996 and 1998 and within buffers among years. Contrary to my prediction, slopes of buffers and controls differed only in 1996, with controls having steeper slopes, and intercepts differed only in  59 1998 with buffers having higher intercepts. Habitat fragmentation is expected to result in steeper slopes and lower intercepts (Connor & McCoy, 1979, Martin, 1981) due to a loss of forest interior species in smaller fragments. Patterns in buffers versus controls could differ from predictions if riparian specialists form a significant part of the assemblage. If this were the case, fewer species may be affected by fragmentation because a disproportionate number of species are primarily associated with the river. This expectation assumes that riparian species are not affected by changes to adjacent upland habitat. In my study, narrow buffers retained riparian specialists like American dipper and belted kingfisher (Appendix 2).  The maintenance of riparian specialists is one of three possible explanations for the patterns observed in my data. Second, the forest-clearcut edge creates new habitat for forest generalists and edge species. Patch area and diversity are often positively correlated, but the relationships reflect habitat variables that are correlated to patch area, such as vegetation structure, rather than area per se (Robbins, 1980, Lynch and Whigham, 1984, Dobkin and Wilcox, 1986, Haila, 1986). The patterns in this study may reflect the earlier successional stage of the adjacent clearcuts and the subsequent colonization of edge habitat by open-area and edge species (Darveau et al., 1995, Hagar, 1999). Third, because only one side of the river was logged, some species may have defended territories across the river, but have made forays to the buffers (see Shirley, in prep.). Very narrow buffers provide the most forest edge for those species that live on the other side of the river, but species richness across all buffer widths may increase with the creation of a forest edge. These three factors can produce narrow buffers where species richness and diversity exceeds that of controls. Darveau et al. (1995) also found higher species richness 3 years post-harvest in narrow buffers relative to controls and lower species richness in wider buffers.  As predicted, species richness and diversity in buffers increased over the three-year period while that of controls remained constant. Previous studies have shown an initial increase in species in fragments due to crowding after harvesting (Bierregaard and Lovejoy, 1989, Darveau et al., 1995, Hagan et a l , 1996), and a subsequent fall to below  60  before disturbance levels (Brown, 1971). This does not explain the increases seen here, 5-7 years postharvest. Instead, the increase in richness in buffers during this study may be due to: 1) the rapid development of edge habitat (shrubs, young conifers and red alder, big-leaf maple and black cottonwood) and/or 2) an increasing ability of adjacent regenerating forest to facilitate movement between remnant patches (Stouffer and Bierregaard, 1995). The increase in species richness and diversity during my study was associated largely with an increase in the number of rare species in the wide buffers and a decrease in Steller's Jays, a dominant species in very narrow buffers, over time. Although the percentage of edge habitat decreases as buffer width increases, the mix of forest edge and interior habitat in wider buffers may increase the number of vegetation niches and thus species.  Patterns of abundance among guilds and species  Contrary to my prediction, I detected no change in abundances of Neotropical migrants either between buffers and controls or among years. Abundances, however, were 30% higher in buffers than controls, but the power of my analyses was low. M y result differs from other studies in eastern North America (Ambuel and Temple, 1983, Lynch and Whigham, 1984, Askins et al, 1990) and boreal forests (Darveau et al, 1995, Schmiegelow et al., 1997). The difference may be due to a lower density of nest predators (Askins, 1995) in this, and other more natural landscapes (Bryant et al., 1993, Schieck et al., 1995, Song, 1999) compared to agricultural landscapes (Freemark and Collins, 1992, Robinson, 1992). Steller's Jays, a potential nest predator, were sporadically abundant in very narrow buffers. Also, Neotropical migrants in this study were dominated by species that may benefit from fragmentation such as those that preferred riparian habitats (e.g. Hammond's flycatchers and warbling vireos) (Shirley, submitted) or used early successional forest (Swainson's thrushes).  My prediction that abundance of foliage gleaners would be higher in buffers than controls was supported in one of three years (1998). The strong increase in abundance in buffers over the 3 years reflects changes in abundances of chestnut-backed chickadees  61  and golden-crowned kinglets. Although they are strongly associated with conifers (Carey et al., 1991, Campbell et al., 1997), these species typically prefer mixed coniferousdeciduous woodland in this region (Bryant et a l , 1993, Campbell et al., 1997) often inhabiting riparian areas of undisturbed forest. The development of a thick forestclearcut edge of red alder, black cottonwood and big leaf maple may provide superior habitat for these species causing a shift from continuous forest to buffer fragments.  As predicted, aerial forager abundance increased over time in buffers, but surprisingly, also did so in controls. Numbers of aerial foragers were dominated by two congeners that made up 95-99% of all detections, Hammond's and pacific-slope flycatcher. In this area, Hammond's flycatchers are associated with deciduous and mixed woodland next to streams (Waterhouse and Harestad, unpubl.). Pacific-slope flycatchers are associated with upland conifer forest, except when Hammond's flycatchers are absent (pers. obs.) when they are associated with both riparian and upland habitats. Hammond's flycatcher abundance was consistently higher in buffers (67%) while pacific-slope flycatcher abundance was higher in controls (60%). Changes in light and temperature at the riparian (Brosofske et al., 1997) and forest-clearcut edge (Chen et al., 1995) may benefit Hammond's flycatcher by increasing the density of deciduous trees and available forest insects (Helle and Muona, 1985, Hansson, 1994). The increase in Pacific-slope flycatcher abundance is more difficult to explain. Because no obvious structural changes occurred in the controls during this period, Pacific-slope flycatchers may have been responding to unexplained changes at a larger scale consistent with Breeding Bird survey trends for pacific-slope flycatchers in southwestern B.C. (Campbell et al., 1997).  The largest differences in abundance between buffers and controls were seen in the forest interior and open-edge guilds (Appendix 1). The forest interior guild made up 7% and 12% of all detections in buffers and controls respectively and is dominated by two species, pacific-slope flycatcher and brown creeper. These species were more abundant in controls across all widths and, as expected, virtually absent in narrow and very narrow buffers. Avoidance of small fragments by forest interior species is well documented  62  (Robbins, 1979, Whitcomb et al., 1981, Blake and Karr, 1984, Askins et al., 1987, Darveau et al., 1995).  As predicted, open-edge species such as Steller's jay, rufous hummingbird, blue grouse, northern flicker and MacGillivray's warbler replaced forest interior species in narrow buffers and were almost 4 times more abundant in buffers compared to controls. This result is consistent with other studies of buffer strips in the Pacific Northwest (Pearson and Manuwal, 2001) and the boreal forest (Darveau et al., 1995, Whittaker and Montevecchi, 1999). Open-edge species showed strong decreases over the three years in controls likely due in part to sampling variation at very low densities and strong increases in buffers probably associated with the forest edge.  Species turnover and community composition  As expected, bird communities in both buffers and controls were more stable in the widest fragments where 60% of species were detected consistently. In contrast, in narrow fragments only 15-20% of species were consistently detected. The negative association of turnover with area has been attributed to increased extinction risk in small fragments with small populations (Hinsley et al., 1995). In the smaller fragments of this study, small populations were represented by more common species while in larger fragments rare species represented most of the turnover. Turnover may also result from sampling error (Arnott et al., 1999) particularly in narrower fragments which not only have smaller sample sizes (Didham et al., 1998), but also include edge species in buffers which tend to forage in the adjacent clearcut (Haila, 1988).  The concept of species turnover, originally put forth by MacArthur and Wilson (1963, 1967), is problematic for forest fragments embedded in a forested landscape (McCoy, 1982). Turnover in forest fragments violates several conditions of the island biogeography model, particularly that no large-scale environmental alterations occur during the sampling period (McCoy, 1982). In particular, forest regeneration following harvesting creates a changing environment in which changes in species composition  63  occur as the forest proceeds through different serai stages, termed "successional turnover" by Lynch and Johnson (1974). Contrary to my prediction, species turnover did not differ between buffers and the controls. This result is consistent with a "sampling view" (Haila, 1993) where species occurrence is a random function of abundance in the regional species pool reflecting changes in territory placement.  Although species turnover showed the same linear decrease with area in both buffers and controls, the communities differed in their species composition. As expected with high turnover in the narrow buffers, these communities were 40% similar to controls on average while the widest treatments were 85% similar. Since the turnover in wider fragments was mainly accounted for by rare species, community composition weighted by abundance is less affected than in narrow buffers with fewer species.  Forest mosaics at a landscape level  Most studies of habitat fragmentation fall into two classes: those that consider permanent changes to agricultural and suburban areas (e.g. Whitcomb et al., 1981, Lynch and Whigham, 1984, Robbins et al., 1989) and those conducted immediately after fragmentation in forested landscapes (Bierregaard and Lovejoy, 1989, Darveau et al., 1995, Schmiegelow et al., 1997). In both cases, negative effects of fragmentation have included an array of edge effects such as changes in microclimate (Matlack, 1993, Chen et al., 1995), wind damage (Laurance et al., 1997), and increases in predation and competition (Gates and Gysel, 1978, Brittingham and Temple, 1983, Ambuel and Temple, 1983). Not all edge effects are negative, however, and areas dominated by forest generalists often show higher bird densities at forest-field edges, which may support higher insect abundances (Hansson, 1983, 1994). While local densities may vary, these communities persist over time in some fragmented landscapes (Hansson, 1994). In my study, increased abundance over time in buffers was due to open-edge species and riparian specialists such as Hammond's flycatcher (Shirley, submitted) that tolerate additional edge habitat. However, abundance also increased in controls during the same period for several guilds. I suggest that this is due to increased edge habitat on a regional  64  scale (Laurance, 2000). Other recent studies have provided compelling evidence to support such regional effects (Curran et al., 1999, Peters, 2000).  Less attention has been given to the influence of the matrix habitat in which remnant forests are embedded. Matrix habitat exerts a primary influence on the distribution of most species in fragmented landscapes (Saab, 1999). While edge effects may be persistent in permanently altered landscapes, regrowth of harvested areas in forested landscapes has increasing influence over time (Stamps et a l , 1987). Use of secondary forest can contribute to the persistence of species (Stouffer and Bierregaard Jr., 1995, Borges, 1999, Gascon et al., 1999) in primary forest remnants (Gascon et al., 1999). M y study shows a positive correlation between increases in species richness' over time in buffers and adjacent clearcuts and suggests that matrix regrowth enhances the persistence of species assemblages. However, there remain areas of concern. First, forest interior specialists such as brown creepers who cannot use the forest matrix may still suffer local declines and extinctions in narrow buffers. Most of the forest species that increased in clearcuts were those species such as the warbling vireo that prefer deciduous and mixed deciduous-coniferous woodland (Morrison and Meslow, 1983, Campbell et al., 1997). Second, for those species able to use a forest mosaic, densities in any one habitat may be misleading (Haila, 1989) and reliable data are needed on reproductive success in each habitat. Finally, although most evidence suggests that fragmentation effects are weak (Trzcinski et al., 1999) compared to effects of forest removal, their unpredictable effects over large scales (Bascompte and Sole, 1996, Curran et al., 1999, Laurance, 2000) can threaten species persistence independently of habitat loss through mechanisms such as increased predation, dispersal barriers (Simberloff, 1992) and food limitation (Zanette, 2000).  CONCLUSIONS  Bird responses to fragmentation in my study do not closely follow the patterns predicted by island biogeography theory (MacArthur and Wilson, 1967) or those seen in studies of fragmentation from eastern North America (Whitcomb et al., 1981, Lynch and  65  Whigham, 1984, Askins, 1990, Darveau et. al., 1995). This is not surprising since most previous studies have either focused on areas permanently altered or recently fragmented areas. I looked for edge effects in species richness, abundance and composition across all buffer widths (10m-125m). As predicted, fragmentation effects were greatest in very narrow buffers with higher amounts of edge. These buffers had the lowest similarity to controls in community composition. This is due in part to the replacement of forest interior species with species that use open areas and edges. M y results are consistent with other studies that show increased numbers owing to an influx of edge species (Oelke, 1966, Linehan et al., 1967, Lynch and Whigham, 1984, Gyug and Vaartnou, 1995, Whitaker and Montevecchi, 1999). The higher variability in density and species richness with declining fragment size is probably due, in part, to sampling error in small fragments as species become rare (Didham et al., 1998). However, because population variability is a good predictor of extinction probability in the forest mosaic (Karr, 1982), this variability suggests that narrow buffers may be unpredictable habitats and function more as travel corridors or foraging areas than permanent breeding habitat. While this hypothesis requires further intensive study, my results should preclude the retention of narrow buffers in the landscape in favor of wider buffers that may provide more stable habitat.  Species-area relationships are commonly used to evaluate the effects of fragmentation (Heywood et al., 1994, Brooks et al., 1999, Gonzalez, 2000), but concerns have been expressed about their usefulness (Martin, 1981, Simberloff, 1992, Gonzalez, 2000). In my study, the primary assumption of a constant slope through time in continuous habitat was met; species richness in controls remained constant during the course of this study. In contrast, there was an increase in species richness within buffers over time rather than species loss. These increases, however, occurred only in the widest buffers. Several forest interior species were found almost exclusively in these buffers and abundances dropped dramatically between wide (125m) and medium (41m) width buffers. Potential mechanisms responsible for species loss include edge effects, reduction of immigration rates and population sizes, and higher order effects (Turner, 1996), but their relative importance in this system is unknown. As yet the buffers studied here have not reached  66  a long-term equilibrium point and continued work is needed to evaluate their long-term viability.  67  Table 3.1 - Characteristics of rivers/streams sampled including length, width, stream order classification and elevation Name  Stream Length  Stream  Stream  (km)  Width (m)  Order  Elevation (m)  Beverly Creek  6.85  15  3  150  Cous Creek  18.57  15  3  250  Kennedy River  50.11  34  5  200  Klanawa River  33.16  32  5  40  Macktush Creek  12.47  21  3  35  Marion Creek  6.75  20  3  350  Nahmint River  34.61  30  4  150  Taylor River  31.54  34  4  225  3.45  9  2  250  View Main Creek  Stream length and order obtained from B.C. Fisheries website http://www.bc.fisheries.gov.bc.ca/fishinfobc.html.  68  NO  oo as as  d  d  —H  C N  d  o  C N  |5 oo O N ON  T—1  cn so SO  15  ©  o  o  C N  iri  in  O N  in oo  rN O  C N  O o  C N  in  XI  rl o o d  ON  OO  d  oo  cn cn  in r-; oo  o  rcn  cn in  oo cn cn  in in  O N C O  o d  r-  o  000"  o  •-r 00 ©  OO  ©  ©  d  cn  C N  O N  O N  C N  O  CN  C N  d  cn  d  in cn  in oo  so as  o  7—i  ^—1  m  d  O  +->  G G  U  ^ r-H  o u  '1  2=1  c  T3 G  0.292  o o  m -*  •<t  cn t-as  |3  SO  • SH  'o  O  r-  NO  ON  Os  O  rd  m d  d CN O N  cn oo oo  o q  in  N O  <N  C N  d  d 00  1 1  S O  r•>*  O N  as in  C N  in  in  OO  cn "*  d  cn  ©  OO O N  © ©  C N  in  CN'  oo in  X)  C N  m  CN  .—i  C N  in ON  d  d  cn oo  C N  d  d  r--  cn  in  d  -a G  03  G G  o3  .O  s  *tN  o  G  cn  o o  TT ©  ON  CN  ON O N  ©  ^  ©  o  in  o  d  ON ©  O N  SO  cn oo  Ui  03  G  c3  < > o u  G  a  o is c  o  O N  cn cn  cn  o  |S  03  <D so  cn oo  CN  ON  O N  ON  in  CN  C N  o  C N  oo  ON  O N  in cn  CX)  ON  Os ON ON  «s CO  ^  NO  00  CN  in  d  Cu  o  oo IS  4-1  G  -a  s  > i ->  en  o•  « • s*  an  C  c  '3D  o Pi  Tab u  BJD  es u O  6 0  a)  <:  "5b •a C P o  6  •3  o Cu  o a,  00  (O 00  DC  I  c  X  o PL,  o  Ci 4)  *—1  NO  NO  o  CO ©  d  oo  CN  OS  oo CO CN  00 00  ro  r~  NO  — ' 1  d  o d  un  NO  d  d  oo  m  ON  d  60S  d  d  o  oo m  ON  p ^H  in  d  d  CO  OO  in  cN  o  d  d  #  , 1 oo  ir,  § ©  NO  d in  CN O  oo  d in  ON  ON  oo  d  CO CN CN  d  m  OO  d  d  NO  ON  m CO  d  NO 1—1  in  TICS  0.101"  o  CO  d  d  CN O CN  CO  d  o  so  p o  t--  d  in  oo  d  CN  o NO  CO CO CN  oo CN  d  NO  CN  m  oo CO  ,—i  CO  ON  ON  oo  d  d  d  d  d  d  d  ON  CN  o  CN  o  d  d  CN  NO  o -*  oo CO  CN  d  d  ON  d oo  d  m  ON  CN  in  o Tf  CN  d  d  NO  o  d  d  o CN  o  XT  NO  o  xi 73  • .—l  <N O  CN NO  CN  O NO  CO 00  d  © ©  oo oo  CO CN  d  00  r-  o  CN  d  ©  © ©  Tt-  ON  CO  d  ©  in  -O  ON  CO CO CO  ,—i  d  in ©  m  ©  m  oo  Tt-  d  ©  ^  o  73  CD 7 3  o o CD O  in oo  as m  d  NO  o  ON  ON  o  (  d  1—1  d  ©  d  CN  in  >-<  CO  r—<  00  in i—i  ©  in  NO  CO  73  73  CO  3 C  co  U  M cj CO  o  C  CO  t/}  Xi  o  CD  Cu, 73 O  CJ C3  *—<  Xi  00  c  43 I +-»  0O  73  ft,  u  co  03  ON  3  co co  .y CJ IS  C  o V -§ IH  O  cd  +-»  CD  o 3  c« CJ  D Cu, O  CO u « C3  PH  c  3 43 3  -3  ca co  o  CD  13  .3.  to  Pi  C  <> - c  CD <D cj CD  £ 73  ^  CD  II 2 ^ .s -a xi a  >  oo  _g  o  73  -a  CD  42 00 CD •*-» O  73  73  — i-O>  c3 >  S3  ° PQ CO  70  Table 3.3 - Results of r m A N C O V A showing comparisons of individual species abundances between 1996-1998 for buffers and controls (n=12 each for controls and buffers, df =2,21).  Species  Buffers  Controls  Change  Change  American robin  +  Brown creeper  +  Chestnut-backed chickadee  +  Golden-crowned kinglet  +*  +*  Hammond's flycatcher  +  -  Hairy woodpecker  +  +  Pacific-slope flycatcher  +  Red-breasted sapsucker  0  -  Steller's jay  -  -  Swainson's thrush  -  +  Varied thrush  +  Warbling vireo Winter wren *  -P<0.10  ** - P<0.01 1 - denotes linear change q - denotes quadratic change  Form  + q l  l  i.q **  +*  _*  U  +*  l  71  FIGURE L E G E N D S  Figure 3.1. Map of study area showing approximate locations of nine watersheds (a) and grid layout at each site (b).  Figure 3.2. Species-area relationships for all species with >1 detections in buffers and controls for the years 1996-1998 (n = 12 each for buffers and controls). A l l regressions are significant. Results for probabilities of differences between treatments from A N C O V A are inset.  Figure 3.3. Species diversity calculated as log series a as a function of width for buffers and controls for the years 1996-1998 (n = 12 each for buffers and controls). A l l regressions are significant. Results for probabilities of differences between treatments from A N C O V A are inset.  Figure 3.4. Total abundance as a function of width for buffers and controls for the years 1996-1998 (n = 12 each for buffers and controls). A l l regressions are significant.  Figure 3.5. Percentage change in abundance for buffers and controls during the period 1996-1998 averaged across all widths (n = 12 each for buffers and controls). Asterisks indicate significance at P = < 0.10 (df = 2,21).  Figure 3.6. Turnover of bird communities for 1996-1998 for buffers and controls calculated as (1 - Jaccard Similarity Index) (n = 12 each for buffers and controls). Both regressions are significant.  Figure 3.7. Similarity of species assemblages in buffer and control fragments using Horn's Similarity Index for the years 1996-1998. Figure 3.8. Correlation of change in species richness from 1996-1998 for buffers and clearcuts (n = 12 each for buffers and clearcuts).  72  Figure 3.9. Results of repeated Measures A N C O V A comparing mean abundance with width between 1996-1998 for clearcuts (n=12). Asterisks indicate significance levels at <0.10 (*) and <0.01 (**), df = 2,22. Letters refer to the function of the relationship, linear (1) and quadratic (q).  73  0  50m  1996  30.0  1997  25.0  ^treatment —  0.714  20.0 15.0 10.0 5.0  • Controls ^ " Buffers  0.0 -5.0 -10.0  1998  30.0" 25.0  treatment  - 0.095  20.0 15.0 10.0 5.0 0.0 -5.0  -10.51  1.0  1.4  1.8  22  Log Buffer Width  2.6  1996  J99Z c/)  CD > C/5  CD "o  CD o.  Controls Buffers  CO  Log Buffer Width  76  Log Buffer Width  77  1.2  0.0  •  J  0.0  .4  1  .8  1.2  1.6  2.0  2.4  Loa Buffer Width  1.0  >  0  20  40  60  80  100  Buffer Width  120  140  CO CO CD  6  c _c o  ir  -i—>  3 O  o  •  •  CO 2  CD D) C CTj  o  -2 -2  4  Change in Buffer Richness  81  L I T E R A T U R E CITED  Ambuel, B., and S.A. Temple. 1983. Area-dependent changes in the bird communities and vegetation of southern Wisconsin forests. Ecology 64:1057-1068.  Arnott, S.E., N.D. Yan, J. Magnuson, and T . M . Frost. 1999. Interannual variability and species turnover of crustacean zooplankton in Shield Lakes. Canadian Journal of Fisheries and Aquatic Science 56:162-172.  Askins, R.A. 1995. Hostile landscapes and the decline of migratory songbirds. Science 267:1956-1957.  Askins, R.A., M.J. Philbrick, and D.S. Sugeno. 1987. Relationship between the regional abundance of forest and the composition of forest bird communities. Biological Conservation 39:129-152.  Askins, R.A., J.F. Lynch, and R. Greenberg. 1990. Population declines in migratory birds in eastern North America. Current Ornithology 7:1-57.  Bascompte, J., and R.V. Sole. 1996. Habitat fragmentation and extinction thresholds in spatially explicit models. Journal of Animal Ecology 65:465-473.  Bellamy, P.E., S.A. Hinsley, and I. Newton. 1996b. Factors influencing bird species numbers in small woods in south-east England. Journal of Applied Ecology 33:237-248.  Bibby, C.J., N.D. Burgess, and D.A. Hill. 1992. Bird census techniques. Academic Press, Toronto, Ontario.  Bierregaard, R.O., and T.E. Lovejoy. 1989. Effects of forest fragmentation on Amazon understory bird communities. Acta Amazonica 19:215-241.  82  Blake, J.G., and J.R. Karr. 1984. Species composition of bird communities and the conservation benefit of large versus small forests. Biological Conservation 30:173-187.  Borges, S.H., and P.C. Stouffer. 1999. Bird communities in two types of anthropogenic successional vegetation in Central Amazonia. Condor 101:529-536.  Brittingham, M.C., and S.A. Temple. 1983. Have cowbirds caused forest songbirds to decline? Bioscience 33:31-35.  Brooks, T.M., S.L. Pimm, and J.O. Oyugi. 1999. Time lag between deforestation and bird extinction in tropical forest fragments. Conservation Biology 13:1140-1150.  Brosofske, K.D., J. Chen, R.J. Naiman, and J.F. Franklin. 1997. Harvesting effects on microclimatic gradients from small streams to uplands in western Washington. Ecological Applications 7:1188-1200.  Brown, J.H. 1971. Mammals on mountaintops: nonequilibrium insular biogeography. American Naturalist 105:467-478.  Brown, J.H., and A. Kodric-Brown. 1977. Turnover rate in insular biogeography: effects of immigration on extinction. Ecology 58:445-449.  Bryant, A.A., J.P. Savard, and R.T. McLaughlin. 1993. Avian communities in oldGrowth and managed forests of Western Vancouver Island, British Columbia. Canadian Wildlife Service, Nanaimo, British Columbia.  Budd, W.W., P.L. Cohen, P.R. Saunders, and F.R. Steiner. 1987. Stream corridor management in the Pacific Northwest: 1. Determination of stream corridor widths. Environmental Management 11:587-597.  83  Burke, D . M . , and E. Nol. 1998. Influence of food abundance, nest-site habitat and forest fragmentation on breeding Ovenbirds. Auk 115:96-104.  Campbell, R.W., N.E. Dawe, I. McTaggart-Cowan, J.M. Cooper, G.W. Kaiser, and M.C. McNall. 1990. The birds of British Columbia. The Royal British Columbia Museum, Victoria, B.C.  Campbell, R.W., N.K. Dawe, I. McTaggart-Cowan, J.M. Cooper, G.W. Kaiser, M . C . E . McNall, and G.E.J. Smith. 1997. The birds of British Columbia. U B C Press, Vancouver.  Carey, A.B., M . M . Hardt, S.P. Horton, and B.L. Biswell. 1991. Spring bird communities in the Oregon Coast Range. Pages 123-142 in Ruggiero, L.F., K B . Aubry, A . B . Carey, and M . H . Huff, editors. Wildlife and vegetation of unmanaged Douglas-Fir forests. U.S. Department of Agriculture, Forest Serice. General Technical Report PNWGTR-285, Portland, Oregon.  Chen, J., J.F. Franklin, and T.A. Spies. 1993. Contrasting microclimates among clearcut, edge and interior old-growth Douglas-fir forest. Agricultural and Forest Meteorology 63:219-237.  Chen, J., J.F. Franklin, and T.A. Spies. 1995. Growing-season microclimatic gradients from clearcut edges into old-growth Douglas-fir forests. Ecological Applications 5:74-86.  Connor, E.F., and E.D. McCoy. 1979. The statistics and biology of the species-area relationship. American Naturalist 113:791-833.  Conover, W.J., and R.L. Iman. 1982. Analysis of covariance using the rank transformation. Biometrics 38:715-724.  Curran, L . M . 1999. Impact of El Nino and logging on canopy tree recruitment in Borneo. Science 286:2184-2188.  84  Darveau, M . , P. Beauchesne, L . Belanger, J. Huot, and P. Larue. 1995. Riparian forest strips as habitat for breeding birds in boreal forest. Journal of Wildlife Management 59:67-78.  Dayton, P.K. 1998. Reversal of the burden of proof in fisheries management. Science 279:821-822.  Decamps, H., J. Joachim, and J. Lauga. 1987. The importance for birds of the riparian woodlands within the alluvial corridor of the River Garonne, S.W. France. Regulated Rivers: Research and Management 1:301-316.  Didham, R.K., P.M. Hammond, J.H. Lawton, P. Eggleton, and N.E. Stork. 1998. Beetle species responses to tropical forest fragmentation. Ecological Monographs 68:295323.  Dobkin, D.S., and B.A. Wilcox. 1986. Analysis of natural forest fragments: riparian birds in the Toiyabe Mountains, Nevada. Pages 455 in Verner, J., M . Morrison, and C.J. Ralph, editors. Wildlife 2000: Modeling habitat relationships of terrestrial vertebrates. The University of Wisconsin Press, Wisconsin.  Ehrlich, P.R., D.S. Dobkin, and D. Wheye. 1988. The birder's handbook: A field guide to the natural history of North American birds. Simon and Schuster Inc., New York.  Enemar, A., B. Sjostran, and S. Svensson. 1978. The effect of observer variability on bird census: results obtained by a territory mapping technique. Ornis Scandinavica 9:3139.  Fisher, R.A., A.S. Corbet, and C.B. Williams. 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. Journal of Animal Ecology 12:42-58.  85  Freemark, K., and B. Collins. 1992. Landscape ecology of birds breeding in temperate forest fragments. Pages 44-454 in Hagan III, J.M., and D.W. Johnston, editors. Ecology and Conservation of Neotropical Migrant Landbirds. Smithsonian Institution Press, Washington.  Friedman, M . 1937. The use of ranks to avoid the assumption of normality implicit in the analysis of variance. Journal of the American Statistical Association 32:675-701.  Gascon, C , T.E. Lovejoy, R.O. Bierregaard Jr., J.R. Malcolm, P.C. Stouffer, H.L. Vasconcelos, W.F. Laurance, B. Zimmerman, M . Tocher, and S. Borges. 1999. Matrix habitat and species richness in tropical forest remnants. Biological Conservation 91:223229.  Gates, J.E., and L.W. Gysel. 1978. Avian nest dispersion and fledging success in fieldforest ecotones. Ecology 59:871-883.  Gonzalez, A . 2000. Community relaxation in fragmented landscapes: the relation between species richness, area and age. Ecology Letters 3:441-448.  Gurevitch, J., and S.T.J. Chester. 1986. Analysis of repeated experiments. Ecology 67:251-254.  Gyug, L.W., and M . Vaartnou. 1995. Riparian habitats: wildlife, vegetation and timber-harvest. Research Update. The silviculture systems wildlife habitat research and extension program, B.C. Ministry of Environment and Ministry of Forests, Victoria, B.C.  Hagan, J.M., W . M . Vander Haegen, and P.S. McKinley. 1996. The early development of forest fragmentation effects on birds. Conservation Biology 10:188-202.  86  Hagar, J.C. 1999. Influence of riparian buffer width on bird assemblages in western Oregon. Journal of Wildlife Management 63:484-496.  Haila, Y. 1986. North European land birds in forest fragments: evidence for area effects? Pages 315-319 in Verner, J., M . Morrison, and C.J. Ralph, editors. Wildlife 2000: Modeling habitat relationships of terrestrial vertebrates. The University of Wisconsin Press, Wisconsin.  Haila, Y. 1988. Calculating and miscalculating density: the role of habitat geometry. Ornis Scandinavica 19:2:88-92.  Haila, Y . , I.K. Hanski, and S. Raivio. 1989. Methodology for studying the minimum habitat requirements of forest birds. Annales Zoologici Fennici 26:173-180.  Haila, Y . , I.K. Hanski, and S. Raivio. 1993. Turnover of breeding birds in small forest fragments: the "sampling" colonization hypothesis corroborated. Ecology 74:714-725.  Hansson, L . 1983. Bird numbers across edges between mature conifer forest and clearcuts in central Sweden. Ornis Scandinavica 14:97-103.  Hansson, L . 1994. Vertebrate distributions relative to clear-cut edges in a boreal forest landscape. Landscape Ecology 9:105-115.  Hatler, D.F., R.W. Campbell, and A. Dorst. 1978. Birds of Pacific Rim National Park. The British Columbia Provincial Museum, Victoria.  Helle, P., and J. Muona. 1985. Invertebrate numbers in edges between clear-felling and mature forests in northern Finland. Silva Fennica 19:281-294.  Heywood, V . H . , G.M. Mace, R.M. May, and S.N. Stuart. 1994. Uncertainties in extinction rates. Nature 368:105.  87  Hinsley, S.A., P.E. Bellamy, and I. Newton. 1995. Bird species turnover and stochastic extinction in woodland fragments. Ecography 18:41-50.  Hodges, M.F., and D.G. Krementz. 1996. Neotropical migratory breeding bird communities in riparian forests of different widths along the Altamaha River, Georgia. Wilson Bulletin 108:496-506.  Horn, H.S. 1966. Measurement of "overlap" in comparative ecological studies. American Naturalist 100:419-425.  Kapos, V., E. Wandelli, J.L. Camargo, and G. Ganade. 1997. Understory birds and dynamic habitat mosaics in Amazonian rainforests. Pages 33-43 in Laurance, W.F., and R.O. Bierregaard Jr., editors. Tropical forest remnants: ecology, management, and conservation of fragmented communities. University of Chicago Press, Chicago, Illinois.  Karr, J.R. 1982. Population variability and extinction in the avifauna of a tropical land bridge island. Ecology 63:1975-1978.  Kaufmann, J.B., and W.C. Krueger. 1984. Livestock impacts on riparian ecosystems and streamside management implications - a review. Journal of Range Management 37:430-437.  Keller, C.M.E., C S . Robbins, and J.S. Hatfield. 1993. Avian communities in riparian forests of different widths in Maryland and Delaware. Wetlands 13:137-144.  Klinka, K , J. Pojar, and D.V. Meidinger. 1991. Revision of biogeoclimatic units of coastal British Columbia. Northwest Science 65:32-78.  Krebs, C J . 1989. Ecological methodology. Harper Collins, New York.  88  Laurance, W.F. 2000. Do edge effects occur over large spatial scales? T R E E 15:134135.  Lehmkuhl, J.F., L.F. Ruggiero, and P.A. Hall. 1991. Landscape-scale patterns of forest fragmentation and wildlife richness and abundance in the southern Washington Cascade Range. Pages 425-442 in Ruggiero, L.F., K.B. Aubry, A.B. Carey, and M . H . Huff, editors. Wildlife and vegetation of unmanaged Douglas-fir forests. U.S. Department of Agriculture, Forest Service. General Technical Report PNW-GTR-285, Portland, Oregon.  Linehan, J.T., R.E. Jones, and J.R. Longcore. 1967. Breeding-bird populations in Delaware's urban woodlots. Audubon Field Notes 21:641-646.  Lynch, J.F., and N.K. Johnson. 1974. Turnover and equilibria in insular avifaunas, with special reference to the California Channel Islands. Condor 76:370-384.  Lynch, J.F., and D.F. Whigham. 1984. Effects of forest fragmentation on breeding bird communities in Maryland, USA. Biological Conservation 28:287-324.  MacArthur, R.H., and J.W. MacArthur. 1961. On bird species diversity. Ecology 42:594-598.  MacArthur, R.H., and E.O. Wilson. 1963. An equilibrium theory of insular zoogeography. Evolution 17:373-387.  MacArthur, R.H., and E.O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, N.J.  Manuwal, D.A. 1986. Characteristics of bird assemblages along linear riparian zones in western Montana. Murrelet 67:10-18.  89  Martin, T.E. 1981. Species-area slopes and coefficients: a caution on their interpretation. American Naturalist 118:823-837.  Matlack, G.R. 1993. Microenvironment variation within and among forest edge sites in the eastern United States. Biological Conservation 66:185-194.  McCoy, E.D. 1982. The application of island-biogeographic theory to forest tracts: problems in the determination of turnover rates. Biological Conservation 22:217-227.  McGuinness, K.A. 1984. Species-area relations of communities on intertidal boulders: testing the null hypothesis. Journal of Biogeography 11:439-456.  Morrison, M . L . , and E.C. Meslow. 1983. Bird community structure on early-growth clearcuts in western Oregon. American Midland Naturalist 110:129-137.  Newbold, J.D., D.C. Erman, and K.B. Roby. 1980. Effects of logging on macro invertebrates in streams with and without buffer strips. Canadian Journal of Fisheries and Aquatic Science 37:1076-1085.  Nillson, S.G. 1986. Are bird communities in small biotope patches random samples from communities in large patches? Biological Conservation 38:179-204.  Norusis, M.J. 1994. SPSS basic and advanced statistics. SPSS Inc., Chicago, Illinois.  Nuszdorfer, F., and R. Boettger. 1994. Southern Vancouver Island and Sunshine Coast. Province of British Columbia, Ministry of Forests. Biogeoclimatic units of the Vancouver forest region. Map 5 of 6. Scale 1:250,000.  Oelke, H . 1966. 35 years of breeding-bird census work in Europe. Audubon Field Notes 20:635-642.  90  Packee, E.C. 1972. The biogeoclimatic subzones of Vancouver Island and the adjacent mainland and islands. MacMillan Bloedel Ltd. Forestry Division, Nanaimo, B.C.  Pearson, S., and D.A. Manuwal. 2001. Breeding bird response to riparian buffer width in managed Pacific Northwest Douglas Fir forests. Ecological Applications 11:840-853.  Peters, H.A. 2000. Clidernia hirta invasion at the Pasoh forest reserve: an unexpected plant invasion in an undisturbed tropical forest. Biotropica in press.  Peterson, R.T. 1990. Field guide to western birds. Boughton Mifflin Company, Boston.  Robbins, C S . 1970. Recommendations for an international standard for a mapping method in bird census work. Audubon Field Notes 24:723-726.  Robbins, C S . 1979. Effect of forest fragmentation on bird populations. Pages 198-212 in DeGraaf, R . M . , and K.E. Evans, editors. Management of north central and northeastern forests for non-game birds. U S D A Forest Service.  Robbins, C S . 1980. Effect of forest fragmentation on breeding bird populations in the Piedmont of the mid-atlantic region. Atlantic Naturalist 33:31-36.  Robinson, S.K. 1992. Population dynamics of breeding Neotropical migrants in a fragmented Illinois landscape. Pages 408-418 in Hagan III, J.M., and D.W. Johnston, editors. Ecology and Conservation of Neotropical Migrant Landbirds. Smithsonian Institution Press, Washington.  Rosenzweig, M . 1995. Species diversity in space and time. Cambridge University Press, U K .  Saab, V . 1999. Importance of spatial scale to habitat use by breeding birds in riparian forests: a hierarchical analysis. Ecological Applications 9:135-151.  91  Saunders, D.A., R.J. Hobbs, and C R . Margules. 1991. Biological consequences of ecosystem fragmentation: a review. Conservation Biology 5:18-32.  Scheiner, S.M., and J. Gurevitch. 1993. Design and analysis of ecological experiments. Chapman and Hall, New York.  Schieck, J., K. Lertzman, B. Nyberg, and R. Page. 1995. Effects of patch size on birds in old-growth montane forests. Conservation Biology 9:1072-1084.  Schmiegelow, F.A., C S . Machtans, and S.J. Hannon. 1997. Are boreal birds resilient to forest fragmentation? An experimental study of short-term community responses. Ecology 78:1914-1932.  Schmiegelow, F.K.A., and S.J. Hannon. 1993. Adaptive management, adaptive science, and the effects of forest fragmentation on boreal birds in northern Alberta. Transactions of the 58th North American Wildlife and Natural Resources Conference :584-598  Simberloff, D. 1976. Experimental zoogeography of islands: effects of island size. Ecology 57:629-648.  Simberloff, D. 1992. Do species-area curves predict extinction in fragmented forest? Pages 75-89 in Whitmore, T . C , and J.A. Sayer, editors. Tropical deforestation and species extinction. Chapman and Hall, London.  Smith, S.M. 1995. Distribution-free and robust statistical methods: viable alternatives to parametric statistics? Ecology 76:1997-1998.  Sokal, R.R., and F.J. Rohlf. 1998. Biometry. W.H. Freeman and Company, New York.  92  Song, S.J. 1999. Effects of natural and anthropogenic forest edge on songbirds breeding in the boreal mixed-wood forest of northern Alberta. Ph.D. Thesis, University of Alberta, Edmonton. 154 p.  Spackman, S . C , and J.W. Hughes. 1995. Assessment of minimum stream corridor width for biological conservation: species richness and distribution along mid-order streams in Vermont, USA. Biological Conservation 71:325-332.  Stamps, J.A., M . Buechner, and V . V . Krishnan. 1987. The effects of edge permeability and habitat geometry on emigration from patches of habitat. American Naturalist 129:533-552.  Stauffer, D.F., and L.B. Best. 1980. Habitat selection by birds of riparian communities: evaluating effects of habitat alterations. Journal of Wildlife Management 44:1-15.  Stouffer, P . C , and R.O. Bierregaard Jr. 1995. Use of Amazon forest fragments by understory insectivorous birds. Ecology 76:2429-2445.  Strong, T.R., and C E . Bock. 1992. Bird species distribution patterns in riparian habitats in southeastern Arizona. Condor 92:866-885.  Trzcinski, M . K . , L . Fahrig, and G. Merriam. 1999. Independent effects of forest cover and fragmentation on the distribution of forest breeding birds. Ecological Applications 9:586-593.  Turner, I.M. 1996. Species loss in fragments of tropical rain forest: a review of the evidence. Journal of Applied Ecology 33:200-209.  von Ende, C N . 1993. Repeated-measures analysis: growth and other time-dependent measures. Pages 113-137 in Scheiner, S.M., and J. Gurevitch, editors. Design and analysis of ecological experiments. Chapman and Hall, New York.  93  Whitaker, D . M . , and W.A. Montevecchi. 1999. Breeding bird assemblages inhabiting riparian buffer strips in Newfoundland, Canada. Journal of Wildlife Management 63:167179.  Whitcomb, R.F., C S . Robbins, J.F. Lynch, B.L. Whitcomb, M . K . Klimkiewicz, and D. Bystrak. 1981. Effects of forest fragmentation on avifauna of the eastern deciduous forest. Pages 125-206 in Burgess, R.L., and D . M . Sharpe, editors. Forest island dynamics in man-dominated landscapes. Springer-Verlag, New York.  Whittaker, R.H. 1965. Dominance and diversity in land plant communities. Science 147:250-260.  Wilcove, D.S. 1985. Nest predation in forest tracts and the decline of migratory ' songbirds. Ecology 66:1211-1214.  Williams, C B . 1964. Patterns in the balance of nature. Academic Press Inc., London.  Willson, M.F., T.L. De Santo, C. Sabag, and J.J. Armesto. 1994. Avian communities of fragmented south-temperate rainforests in Chile. Conservation Biology 8:508-520.  Zanette, L., P. Doyle, and S. Tremont. 2000. Food shortage in small fragments: evidence from an area-sensitive passerine. Ecology 81:1654-1666.  94  O c  i3 c cd  i«3 o a  a  |3 >N  Tt  oci SO  r~  oo oo  SO  OS  i/i  oo (N  CO  Tt  so in  r~  a 3 1  < i3 3  rTt ~-  CO CO  O  UO  OS  SO  d  <N  T-H  r-  OO  CN  SO  CO T-H  in  CN  oo Tt  CO  in  Os' in  (N  in  CN  i—i  ON  oo  Tt  d  d  '  oo ©  ON  ON  oo r~;  CO CO  CN  CN  d  Tt Tt  oo  SD  CN  in  CN  in  so  ON  t~  CO CO  <N  Tt Tt  T-H  CO  CO  SO  in  Tt  OO  SO in CN  oo  o  ©  © p  CO  CO  d  oo  ©  •n d  SO  oo i—i  CN  Tt Tt Tt  CN CN  oo d  O d  in T-H  SO o CN  SO d  oo  in  o CN  1—1  d in  CN  d  oo  r-  OO  CN CN  CN  ON  oo  NO  ON  SO CN  in  Tt  CN  d  d  O in IT)  O O in  CO CO  Tt Tt  CN  CN  Tt  CO CO  OS  Tt  Tt  SO in Tt  t~ Tt so  Tt  T-H  1—t  CO  CN CN  ON  ON  ON  1  oo  Tt  oo  CN  '-'  Os t~  od  CO O  OO  CO  O  CO CO  SO •n  dm  o  CN  CO  o  CN  co  CN  CO  Os  ON  Tt d 1—1  r-  '1 i5  '1  SO  T-H  IO T-H  co  o  Tt  in  CN  CN  CO  CD  o  a  T-H  T3  SO  m Tt  ON  d  T-H  CO  oo oo od  NO  Tt  CO  CN  oo  in in Tt  CN  Tt CN  oo T-H  CO T-H  OS  o  CO  oo r~  CO  Tt O Tt  o  CN  OS  o o in  o  Tt  CN CN  Tt  CN  cd  X  O  3 O  IS  c  cuu -=  .2  cu CO  s •a cS «  o  CD  a.  cd  "3 >> u o CS  u on  "3  cd O  'S. o o CD  3  M C  co 73 CD  CD  60 cd  Ml  C  "3» 13 «S  o  co  00 CD  00 cd  "o OH  "oo 73 3 3 O u  a  3  XI  CS  8 Cn CH  2  o CD  Cu  CO  S  S3  Cn  o PH  95  in TICS  i—, CN CN  CN d  NO  roo — I<  rTt d  o CN  in CN d  NO o d  ON cn d  o in d  o o d  Tt  Tt i—i  i—, d  , i T-H  in CN d  r-5  CO d  ON  Tt  d  CN  d  oo in i—i  oo o d  rON  ON CN Tt  CN  ON  oo 1—*  d  O  t--  ©  d  Tt oo d  Tt in  d  o o d  cn d  d  o o d  cn d  m rd  rON d  Tt d  m d  cn oo  O Tt d  d  d  d  Tt Tt d  r-  o o d  o o d  cn  cn NO  d  o o d  d  00 d  ON  ON  ON NO  d  d  OO >n d  >n CN d  o o d  NO cn d  cn cn d  Tf Tt — l 1  o o d  ,—i d  in CN d  Tf-  cn oo d  NO  ON  CN CN  CN d  Tt in d  cn cn d  o o d  o o d  CN Tt d  ©  ON  o d  cn d  CN d  o o d  OO CN d  — i 1  cn cn d  cn d  ON  T—1  CN ON  ON NO  ON NO  Tt <—'  ON  d  d  CN  d  cn d  in CN d  OO  ^1 cn d  ON  CN  in  o in Tt  Tt ON CN  CN Tt  cn m  Tt Tt cn  ON 00 d  — iH NO d  NO m Tt  o in  CN  cn m  Tt CN Tt  in ON CN  m  cn oo  CN cn CN  O o d  cn d  >-<  o d  NO  <N in  OO  in cn d  *-<  p>-<  oo  o Tt d  >-<  Tt Tt d  NO  rTt d  00  o d  m d  CN Tt d  d  d  ON CN  ON ON  oo ON  CN d rd  CN ON d  oo  p  CN —  1  Tt  >-*  NO O in  ON in  u u  CD  T3  a  O  IS o  T3 ^ CJ 03  1  3 3  co <D J3  u  M o  u cx  T3  O  CJ >, u  o  Cu O  u3  o o3  <D U 3  V} CL, o3  00 -O u  00 CJ <D PH  3  o  J3 c3  CD  . 3 >  3 O  s CO  T3  >  3  CD  60  c  3 S3  CD 3  in oo C M T-H O r- o oo o  o r—1 o d  00  ©\ ON  u  PS  ON ON  u  d  T-H  Os  o in d  o d  CN  Tt T (  O d  d  i—i  CO.  T-H ^H  d  d  CN  o  d  d SO  in CO o  C N Tt d r ~ ~  d  d  m  CN  d  00 in d  o CN d  m CO  CO  0 0 CO  SO  CN C N O  CN o Tt CN N T-H o C o o d d d o  d  d  d  < N CO  Tt Tt T-H in d d  o d  > n m Os oo CO CN  CO Os  o o d  o  SO i—i  d  o o  so  o  Tt T-H  o d  o  SO CO  oo in i —N i m C Tt d d  m  o  O  Tt s o d <N  O  o o so OO d CN  O d  o >n Tt Tt N O C N m Tt C d d d  >n o c o o T-H Tt d d  o OS in in T-H o o o in d d d  CN C N d  m  SO  o Os <N  so o in in s o p T-H d  T-H  T-H  1—1  d  o d  d  so  o d  SH  03 S  wo  c  WD es SH  \3  d  o d  T-H  T-H  T-H  Os  NO ON ON  d  Os  O  CD bO 03 >H O  PH  is  SO  c o o  <D  CD  03  03  C  CD  o d  s o SO d  d  3 O IH  O  ,—i CO  CO (N  CN  d  CN Tt T-H CN d d  co in Tt o Tt >n d CN Tt  Os  CN co  03  SH  CD  O -a <D 00 c  PH  o o d  T-H  C  CD  o3 O  O  'in  CD  CN  CN  wo  8 4•—» O CD  C CD 73 CD  PS  c  0 3 *H  CD  o — -t > 00 CD  lH o PH  CD  00  73 CD i  e  CD  PH  O  3  O  3 0  -1  97  Chapter 4: Effect Of Buffer Strip Width On The Movement Of Birds Across Rivers And Clearcut Edges Introduction Habitat fragmentation threatens the persistence of many species living in continuous habitat (reviewed in Wilcove and Robinson, 1990; Saunders et a l , 1991; Robinson et a l , 1995) by disconnecting contiguous populations into divided populations in remnant patches (Hanksi and Gilpin, 1996). Movements of animals between patches influence both the dynamics, spatial structure and persistence probability of populations (Kareiva, 1990; Turchin, 1991). Immigration into patches may be reduced if the landscape surrounding the patches (the "matrix") is hostile (Wilcox and Murphy, 1985). The ability of a species to move across the matrix may be critical to its persistence in fragmented landscapes (Fahrig and Merriam, 1994, Belisle et al., 2001). Understanding how animals respond to the distribution of forest remnants may be useful in combining conservation efforts with landscape harvest designs. To date, we know that the amount of forest cover can influence movements (Wegner and Merriam, 1979; Belisle et al., 2001) over a landscape, but little is known about how animals use forest remnants at the scale of home-range movements. Therefore, studies that examine movement as a function of landscape characteristics are particularly important.  Species differ in their tendency to cross the matrix. Among forest birds, habitat generalists (Sieving et al., 1996; Rail et al., 1997; Haddad, 1999), large species (Grubb and Doherty, 1999, Brooker et al., 1999) and species with larger area requirements (Dale et al., 1994) are more likely to cross gaps among forest patches.  In contrast, some forest  specialists (Rich et al., 1994; Sieving et al.; 1996, Bright, 1998) and small understory birds (Karr, 1982b) are reluctant to cross open areas and use woodland corridors to move between patches (Haas, 1995; Haddad, 1999). There are several reasons why species cross habitat boundaries (Grubb and Doherty, 1999; Norris and Stutchbury, 2001) such as natal dispersal (Desrochers and Hannon, 1997), food availability (Huhta et al., 1998) and mate selection (Norris and Stutchbury, 2001). However, factors that influence these  98  movements are not well understood. It is commonly suggested that species use corridors to avoid predators that are present or harder to hide from in habitat gaps (Lima and Dill, 1990; Harrison, 1992; Bright, 1998) since there are fitness costs associated with dispersal through fragmented areas (Matthysen and Currie, 1996, Brooker et al., 1999).  I used buffer strips, i.e. linear fragments of remnant forest retained along rivers and streams after tree harvesting, to examine short-range movements across habitat boundaries. In this paper I address the following five questions: 1) Does movement across habitat boundaries vary with buffer width? I expected the total number of movements across the boundaries of narrow buffers to be more frequent than wide buffers. Wide buffers may provide enough habitat for forest birds to establish breeding territories while narrow buffers of insufficient habitat quality may serve primarily as travel corridors or foraging perches. Alternatively, animal movements differ at varying levels of habitat fragmentation (Andreassen et al., 1998; Mclntyre and Wiens, 1999), but may not display a consistent relationship with fragment size. 2) Does movement frequency differ over the breeding season? I expected movements to increase later in the breeding season. Foraging by forest birds in adjacent habitats (Wegner and Merriam, 1979) may increase as food becomes more abundant in adjacent clearcuts. Movements may also increase with the addition of immature birds later in the breeding season. 3) Do movements differ among habitat guilds and species? I expected habitat generalists and open-edge species to cross boundaries more frequently than forest specialists because they use open areas (Rail et al., 1997; Cassady St. Clair et al., 1998). 4) Do movements differ between river edges and forest edges? I expected movements across clearcut edges to be more frequent than across river edges due to an abundance of berries and insects within clearcuts. Alternatively, gaps produced by clearcuts tend to be wider than gaps across rivers. In this study, the widest average river width was 48m. Therefore, movement across rivers may be more frequent than across clearcuts. Where only one side of the river is logged, forest-interior species may be more willing to cross rivers to reach adjacent continuous forest. 5) Is bird movement frequency related to bird density or abundance within the buffers? Wiens et al. (1985) proposed that higher densities in a patch should lead to more frequent movements across habitat boundaries and higher  99  interpatch movements are associated with higher densities for patches connected by corridors (Haddad, 1999; Haddad and Baum, 1999). Densities declined with buffer width (Appendix 3) so movements should be more frequent across narrow buffers. However, the actual relationship may be more complex if organisms also respond to habitat configurations and resources at larger scales (Orians and Wittenberger, 1991), i.e., percentage of landscape logged. Together, these five questions will help us to understand natural movements in an explicitly controlled set of habitats.  Methods Study area I selected 15 riparian sites within 8 watersheds in a forested landscape on the west coast of Vancouver Island, B.C. using 1:20,000 forest cover maps. Fuller details on forest vegetation are given in Shirley (Chapter 2). The sites (i.e. buffers) are old-growth remnant strips of varying widths (3 - 125m) retained along rivers. Adjacent to the fragments are patches of early successional forest 3-5 years old (i.e., clearcuts). Canopy height in the buffers averaged 50-60m while height of the tallest vegetation in adjacent clearcuts averaged 4m. I refer to the buffer categories based on average width of the replicates as very narrow (6m), narrow (22m), medium (41m) and wide (125m) in the analyses. I selected three replicates of each width category along with three control sites of undisturbed continuous forest for a total sample size of 15.  Movement Frequency  Three surveys were conducted from May 21 - July 15, 1998. A mapping grid was established at each site (details in Shirley, Chapter 2) (Figure 4.1b). During each survey, an observer walked around the perimeter of each fragment once stopping at four "stations" 100m apart along the river edge and the forest-clearcut edge and recorded bird movements. All surveys were conducted from 05:00 -11:00 h. Surveys were not conducted during heavy rain and/or strong winds. Each survey was conducted by one of  100  four observers and observers were rotated among and within sites. Sites were surveyed in a different order to reduce bias due to time of day.  In a survey, at each station, I observed for 15 minutes (7.5 min. each looking forward and backward along the edge) and recorded all birds crossing to and from the fragments across the river or into the adjacent clearcut. Only those individuals where we could clearly determine the species were recorded; sex was recorded where possible. Some species were less conspicuous when flying; for these, I used any songs and calls made after landing to help identify the species. Birds in large flocks were recorded as a single record. We recorded whether the bird was moving in or out of the fragment, the approximate distance moved, and if the bird was calling. If a bird was seen to land, we recorded details about the associated vegetation structure.  Bird Density Bird densities (Appendix 6) were calculated from data collected by spot-mapping censuses (Robbins, 1970; Bibby et al., 1992) conducted concurrently with movement observations but by independent observers. Details are given in Chapter 2.  Data Analysis I tested the data for normality of number of movements before conducting statistical tests using the Shapiro-Wilks statistic. Homogeneity of variances was tested by the BartlettBox F statistic in the analysis of covariance and repeated-measures A N O V A procedures. Data that substantially violated the assumptions were transformed by log (x + 1) or Vx. For strongly non-normal data that were not improved by transformation, I used a rank transformation (Conover and Iman, 1982). I used an alpha value of 0.10 to represent a high conservation cost of making a type II error (Smith, 1995; Dayton, 1998). Data were analyzed using SPSS for Windows 6.1.4 (Norusis, 1994). To determine if the frequency of movements across boundaries varied with buffer width, I summed the movements of each species for all four stations at both river and  101  clearcut edges. I classified the species into habitat guilds based on published literature (Hatler et al., 1978, Ehrlich et al., 1988, Peterson, 1990, Campbell et al., 1990, 1997) (Appendices 1-2). The following analyses were conducted using total numbers of crossings and numbers per habitat guild. I also performed analyses for 5 individual species. I tested if the number of crossings varied across survey period using repeatedmeasures analysis of covariance (rmANCOVA) with year as the within-subjects factor and buffer width as a covariate. I compared the number of crossings across the range of buffer widths during each survey using regression analyses.  Since there were no significant interactions between edge type and survey day, I summed the data for all surveys and used paired t-tests to compare the number of crossings between river and clearcut edges. Lastly, I tested whether there was a relationship between the average number of crossings and bird density for all habitat guilds combined and individual habitat guilds. For all habitat guilds combined, I calculated the slope of the number of bird crossings with density at each site and used a one-sample t-test to test if the average slope was different than zero. For individual habitat guilds, I used linear regression with the number of crossings as the dependent variable and density as the independent variable. Where data violated the assumptions of normality, I transformed the data using a log transformation.  Results At both edge types, the highest numbers of crossings were by forest generalist, open-edge and ubiquitous species (Table 4.1, Appendices 1-2). Seven out of nine riparian species crossed at low frequencies, but only one forest interior species out of five, the Pileated Woodpecker, crossed either boundary and then only did so once from the widest buffer. American Robins crossed at least 3 times as often on average as any other species. Vaux's Swift and Red Crossbill had a large number of crossings at one particular site. Fewer species crossed rivers than clearcuts. The clearcut edges were crossed by 39 species and of these, 16 species had > 10 crossings. River edges, in contrast, were crossed by 32 species in total and 30 species excluding controls. Of these 30 species, 4  102  species had >10 crossings. Forest generalists crossed both edges more frequently than other guilds (41% of species in both cases). Movements Over the Breeding Season  I compared the number of crossings over the 3 survey periods for all sites. The number of crossings did not vary over the breeding season either for total crossings or for species grouped by habitat guilds at either the river or clearcut edges (all p's > 0.10). While the number of crossings was low for many species and often aggregated at one site, I was able to compare the number of crossings over time for 5 individual species. Although the number of crossings did not vary over the breeding season for most species, movements by Swainson's Thrush increased later in the season at both river (F2,28 = 6.30, p = 0.03) and clearcut edges (Fi.n - 6.33, p = 0.03). Increases for this species were approximately equal across all buffer widths at the clearcut edge, but at the river edge the largest increases were seen in the very narrow and medium buffers. There was very little variation in the number of crossings in the narrow and wide buffers and a decrease in the number of crossings through time in continuous forest controls (Figure 4.2).  Movement Across Varying Buffer Widths  There was no evidence that the number of crossings varied with buffer width for either river or clearcut edges (all p's > 0.15) (Figure 4.3). Regressions of number of crossings against buffer width for all guilds were not significant except for generalists and riparian specialists and even these were significant in only one of three surveys. A trend toward more movement by open-edge species in very narrow buffers was obscured by high variation within this treatment at both edges (CV = 100% for clearcut edges and 126% for river edges). For individual species, there were no significant relationships between number of movements and buffer width except for the Swainson's Thrush (all p's > 0.25). For this species, there were significant relationships with buffer width at river edges in the first two surveys (survey period 1: Fij3 = 3.75, p = 0.08, survey period 2:  103  Fjj3 = 3.59, p = 0.08); however, the relationship was positive in session 1, but negative in session 2.  Movements Across River and Clearcut Edges  To compare the crossings at river and clearcut edges, I again summed data where there were no trends across survey periods. The number of crossings in total averaged 3 times higher across clearcut edges than river edges (Figure 4.4a). The number of movements differed significantly among several habitat guilds and species (Table 4.2). Forest generalists and ubiquitous species (represented mainly by the Swainson's Thrush and American Robin) crossed clearcut edges 3 times more frequently than river edges while open-edge species (primarily Rufous Hummingbirds) crossed 7 times more frequently (Figure 4.4b-d, Table 4.1).  I was able to compare crossings for only 3 individual species due to low sample sizes for other species: American Robin, Swainson's Thrush and Rufous Hummingbird. American Robins and Swainson's Thrushes crossed nearly 3 times more often at clearcut edges than at river edges across all buffer widths, while Rufous Hummingbirds crossed clearcut edges 6 times more frequently across all buffer widths (Table 4.1). Correlations of Movements with Density  For crossings for all habitat guilds combined and within each habitat guild, I tested whether the average number of crossings over the season was related to average bird density in the buffers. The average slope of the number of crossings for all habitat guilds combined with density was significantly different from zero at both river edges (r = 3.33, p = 0.01, df = 14), and clearcut edges (t = 1.89, p = 0.09, df = 11). The open-edge species at the river and clearcut edges was the only guild for which the number of crossings was significantly and positively associated with bird density (Table 4.3).  Discussion  104  Effect of Buffer Width on Frequency of Movements  Although I predicted a negative relationship between number of crossings and buffer width, the frequency with which birds crossed rivers and forest-clearcut edges was not correlated with buffer width. I hypothesized that birds would move in and out of narrow buffers more frequently. There are two reasons to expect this pattern:  First, narrow buffers might be used for nesting by species adapted to open areas and/or forest edges such as the American Robin. Such species are tolerant of the adjacent clearcut matrix and might incorporate the buffer as part of their territory. I could not detect more frequent movements by open-edge species and ubiquitous species in narrow buffers despite a trend toward more movement by open-edge species in very narrow buffers especially at clearcut edges. Movements were dominated by Rufous Hummingbirds and Steller's Jays at both edges. Variation in the frequency of movement among species could be due to differences in site characteristics such as vegetation as well as to species-specific factors such as body and territory size (Wiens et al, 1985; Dale et al., 1994). For example, Rufous Hummingbirds forage on salmonberry flowers in April/May and fireweed flowers in June in clearcuts (pers. obs.) and these plants were abundant at some sites, but scarce at others. Flocks of Steller's Jays were observed flying in and out of buffers at certain sites, particularly in late June. Steller's Jays prefer open woodlands and the edges of clearings (Campbell et a l , 1997), but their movements are erratic and may involve both altitudinal migrations during the warmer months and latitudinal migrations. This species may use narrow buffers as travel corridors that offer more protection from predators rather than flying across open spaces.  Second, certain species might nest in either the buffer or across the river in continuous forest, but forage in the adjacent clearcut. For example, American Robins, which forage on berries and nest in mixed forest (Campbell et al., 1997), crossed most frequently. There was, however, no preference for movements to and from narrow buffers, but a trend towards more crossings from medium buffers at both edges. However, because no studies of nesting locations were done, it is difficult to say if robins use the buffers and  105  adjacent clearcut to forage for berries while nesting on the opposite side of the river or if medium width buffers provide a superior combination of berry abundance and trees for nesting. Swainson's Thrush, another frugivorous species, had higher movement frequency in wider buffers initially and in narrow buffers later in the season. This pattern is difficult to explain except that Swainson's Thrush is one of the latest migrants, arriving in late May/early June, around the time of the first survey. The first survey may therefore have been dominated by birds prospecting for territories rather than by the back-and-forth foraging movements typical of nesting birds or birds with dependent fledglings.  These results contradict those of another study of bird movements in forested landscapes that show a higher frequency of crossings in narrow buffers (Darveau et al., 1995) in the first three years after harvest. In that study, crossings were dominated by forest-dwelling species that may have established territories covering both sides of the river and may reflect a short-term response of individuals as they adjust their territories to a new forest configuration. In contrast, the results of my study differed from those of Darveau (1995) perhaps because more time had elapsed since harvest.  Movements Across River Versus Clearcut Edges  I predicted more crossings at the clearcut edge compared to the river edge because I expected that the feeding and nesting habitat provided by regrowing forests would be used by many species (generalists and forest nesters that eat fruit). M y results supported this prediction by showing higher numbers of crossings at the clearcut edge for forest generalists and open-edge species. This is evidence that the frequency of local movements in and out of patches varies with the adjacent habitat. For forest generalists, there were more crossings at the clearcut edge for narrow and medium buffers. Crossings in these buffers were evenly distributed among species suggesting that these buffers represented territories or nesting areas from which birds made forays for food. Other guilds may use buffers for different purposes. For example, wide-ranging irruptive species such as Red Crossbills that travel in large.numbers likely use buffers mainly as travel corridors.  106  For open-edge species, crossings at clear-cut edges also were more frequent than at river edges. Open-edge species such as Rufous-Hummingbird and Steller's Jay are found at low densities in interior forest (Campbell et al., 1993) and probably breed near the clearcut edge. While the river edge represents a gap in this forest, these species may not find it as suitable as a clearcut edge because of differences in food resources and microclimate. There is little good habitat for edge species in the interior forest across the water. River edges themselves are significantly cooler than clearcut edges during the day (Brosofske et al., 1997) and may be avoided by Rufous Hummingbirds. They also provide more cover for songbirds against nest predators such as jays (Gates and Giffen, 1991; LaRue et al, 1995). The American robin, a ubiquitous species, had the largest number of crossings overall at both edges. This species occupies the widest range of habitats in the study area being found in interior forest as well as early successional forest. Its most preferred habitat appears to be at edges whether river, roadside or clearcut (pers. obs.). The higher number of crossings at clearcut edges for this species may reflect the abundance of food in the adjacent secondary growth.  Relationship of Movements with Density  I predicted a positive association between the number of crossings with density such that narrow buffers with higher density would have higher numbers of crossings at both river and clearcut edges. I found that movements of birds across both habitat boundaries were positively correlated with their densities in the buffers. This result is consistent with theoretical expectations that encounters with habitat boundaries (Tischendorf and Wissel, 1997) and perhaps the potential to cross boundaries are negatively related to buffer width. Since buffer width itself explained 76% of the variance in abundance during 1998 (Shirley, Chapter 2), buffer width should be correlated with movement frequency as well. Instead, buffer width does not account for the variation in movement frequency among sites. This result may be due to natural variation in density among buffers of similar width and might be corrected with larger sample sizes. Alternatively, other habitat characteristics such as tree species composition and tree density in buffers, adjacent clearcuts and forests across the rivers might influence movements more strongly. The  107  positive relationship between density in open-edge species and movements at both edges likely reflects a greater tendency to use riparian habitats on both sides of the river and to cross the forest-clearcut edge by species such as Rufous Hummingbirds and Northern Flickers. A positive association between total abundance in buffers and movements was found by Machtans et al. (1996). Machtans et al.'s (1996) result suggests that individuals with territories in the buffers may exploit resources such as food (Huhta et al, 1998) and mates (Norris and Stutchbury, 2001) in adjacent habitats or in other fragments nearby. In this study (Machtans et al., 1996) movements by forest species in adjacent clearcuts were also those species associated with open-edge habitats.  Influence of the Clearcut/Forest Matrix on Movements  I predicted that movements across habitat boundaries would increase over the breeding season in response to increased food production later in the season and the fledging of juvenile birds. Fruiting shrubs dominate both the forest understory and the adjacent clearcuts. Salmonberry (Rubus spectabilis) is the first dominant shrub to bear fruit (in early June) followed by blueberry {Vaccinium alasken and V. ovafolium) and red huckleberry (Vaccinium parvifolium) in July. In contrast to my prediction, there was no increasing trend in movements over the period. The one exception was the Swainson's Thrush. This species feeds extensively on salmonberry (pers.obs.) and also uses this shrub for nesting. Salmonberry abundance may influence movements strongly in this frugivorous species. Although juvenile birds were incidentally observed moving habitat boundaries, the sampling design did not able me to explicitly test the relative frequency of movement of adults and juveniles.  This study was conducted over one breeding season and movements between habitats might increase with increasing years since harvest. One important difference between clearcuts and agricultural landscapes is the adjacent regeneration of the forest in harvested areas (Askins, 1994). Regeneration of these woodland areas should facilitate movements between remnant patches by reducing the contrast between habitats and creating "soft edges" (sensu Stamps et a l , 1987), but this may only occur several years  108  after harvest. Darveau et al. (1995) found no temporal trends in movements in the first three years after harvesting, but Stouffer and Bierregaard Jr. (1995) found that some locally extinct species returned to fragments within 5 years by moving through the adjacent habitat. In particular, regrowth of key shrub and tree species may benefit understory species especially sensitive to fragmentation (Recher and Serventy, 1991; Stouffer and Bierregaard Jr., 1995; Sieving and Karr, 1997). However, these positive effects may not occur for all species; even subtle habitat boundaries may present a barrier for some strong habitat specialists (Haddad, 1999).  Implications for Conservation of Forest-Dependent Species M y results demonstrate that forest generalists, ubiquitous species and open-edge species move readily across habitat boundaries presumably to exploit resources in adjacent habitats. Of the forest-interior species, only one, the pileated woodpecker, was seen crossing boundaries, and it only did so at one wide buffer site that had extensive red alder regeneration. Forest generalists dominate bird communities in this region and forest type with differences among sites due to the presence of rare species (Bryant, 1997; Shirley unpubl.). While my sample size of forest-interior species is small due to their scarcity, other focal observations of these species (unpubl.) showed no movements and support the result that they avoid crossing habitat boundaries. While this paper presents only one year of data, the results confirm other studies in other forest types that show generalists and open-edge species are more willing to cross gaps (Machtans et al, 1996; Rail et al., 1997) while forest specialists are inhibited from crossing even narrow gaps (Rich et al., 1994; Rail et al, 1997; Bright, 1998). For these latter species, wide buffers may provide breeding habitat in addition to connections between unlogged patches. Given that responses to habitat gaps differ among species in this and other studies (Machtans et al, 1996; Cassady St. Clair et a l , 1998), conservation of forest-dependent species would benefit from more intensive studies of movements using radios to identify those species most constrained by habitat boundaries and the mechanisms that influence their persistence across the landscape.  109  3 O 3  Q  a  c  3  cd  id  CO in i—i  O O O O o d  ^—'  CO CO CN r-  c3  CO Q  73  3 a co  2  o d  CO  o o SO CN  o o  r—s  ,—  oo CO >  o o SO OS CN  3  O co  10 Q CO  ••8  CO C d CO la 3  O in d CN in CO so  3  o  cj Ua  (0  o  g  Q co, 3 a CO  IZ  CO  IS  3  o  co  IU  Q |Z  >  CO  a  c<d D  co >  £  10 D-  IT3 CO  ll  co co od r-so  S CO O  oo in d CO CO c>T oo d  O O  oo cn d rSD d  in*  co co co  CO in so  8  in in  1  — '  ,—  o o Tt  in T£  o p  CO C O T-H  o p in  so Tt  o p  O o © o d  in  o o CO o o  o OS, CO CO in CN CN oo d  n  —'  CO CO  o o d o o  o o d o o O  p SO © d CO CO CO O2S  o o d o o d  s  N  —^  , 1  CN CO CO CO CN SO OS CO CO o o o o T-H  SO Tt CO CN CO CO CO CN V  — '  CN SO  Tt, '  t-~ S CO  s  r  oo in d SO d C inO — -  CO CO —I T-H  —H  in CO N  — '  SO co co~ in T-H  CO CO '-'  ,  v  *•—'  T-H  rso C CN N  CO CO T-H  1  T s  so  oo m d  '  T-H  so d  OS  CO  so  so  o o  Tt  o o CN  o o CN  O inO d— CO CO d  oo oo d CN o p  T—1  ,—  C dO  CO O SO rSO in  O inO d  >n T-H —  C inO d  SO d  SD d  in in so  CN CO so  1—1 — '  oo d  T-H — '  o o  C O in T>H  ^— ©  J  o o CO  O O CN O o od  y  ,—,  o o d o o d  T-H  '  V  o o N •C — d p od  C inO ,_4 CO CO CN  s  ^  T-H ^—'  so CO o p CO in so o o  S OO CO t-~ SO CN ^—s  in S CO O CO  y  T-H  — '  SO CN  —^  V  S  oo in m T £ - i—i  ,O^ C t-;  O o in  o o CN  f—.  in CO CO CO SO  T-H *—'  o o CN , —  X  — '  S  r~ SO SO —*  T-H  K  T , ''  '  rSD  T-H S  — '  T-H  13  o  CO* rT-H -^  oo in d CO CO d  o o d o ©  Tt  oo m d rSO d  c CO  3 J3 O  '5  o  o PH  o PH  C O 00 73 CO C C O o. o  H  c cd  s  3  O 3  CT  "e o  «i 3  'cd 15 00  110  Table 4.2 - Paired t-tests comparing mean number of crossings at river (n = 15) and clearcut edges (n =12). Session  Clearcut  River SD  X  SD  X  P  1.19  0.32  1.75  0.22  <0.001  Forest Interior  0.25  0.87  0.08  0.29  0.339  Forest Generalist  5.00  4.69  18.71  12.77  <0.001  Open-edge  2.08  2.47  14.17  12.98  <0.001  Riparian  1.58  1.83  2.83  3.04  0.128  Ubiquitous  6.58  6.40  18.42  10.88  0.001  American Robin  6.08  6.64  16.33  11.30  0.002  Rufous Hummingbird  0.92  1.00  5.50  5.18  0.007  Swainson's Thrush  1.33  1.15  3.92  2.94  0.004  Total  Habitat Guild  Species  Bold font denotes P < 0.10  I.  Ill  Table 4.3 - Linear regression slopes and p-values for number of bird crossings with density of birds for each habitat guild at river (n = 15) and clearcut edges (n = 12). River Slope  Clearcut SE  P  Slope  SE  P  Habitat Guild Forest interior Forest generalist Open-edge Riparian Ubiquitous  0.03  0.91  0.976  5.09  13.77  0.719  -0.14  0.39  0.723  -4.00  15.64  0.803  1.97  0.59  0.006  98.76  26.29  0.004  -0.05  0.64  0.933  7.75  6.67  0.273  0.36  1.30  0.787  -22.15  42.84  0.616  Bold font - denotes P < 0.10  112  Figure Legends Figure 4.1. Schematic drawing of study locations (a) and mapping grid established at each site (b). Transect lines measure 450m parallel to the river edge and 200m away from the river. Bird movements were observed from 4 stations along the river and clearcut edges at 100m intervals.  Figure 4.2. Relationship between number of movements by Swainson's Thrush across habitat boundaries and time during the breeding season for: (a) movements across river edges, and (b) movements across clearcut edges. Data points overlap for medium and very narrow buffers and for narrow and wide buffers.  Figure 4.3. Relationship between number of crossings and habitat guild at each buffer width category for: (a) movements across river edges, and (b) movements across clearcut edges.  Figure 4.4. Linear regressions of bird movements with log buffer width across rivers and clearcut edges by habitat guild for (a) all species (b) forest generalists (c) open-edge species and (d) ubiquitous species.  113  •—#'''0' •  1 II I ; f f 50m  Buffer  ®  Clearcut  Survey Period  115  30.00,  20.00J  116  All Species 120.0  120.0  80.0  80.0  40.0  40.0  Forest Generalists  C/)  CD  C  0.0  .4  .8  1.2  1.6  2.0  2.4  8  "(/)  1.2  1.6  2.0  2.4  1.6  2.0  2.4  o O  Open-edge  H—  o  CD  Ubiquitous  120.0  120.0  80.0  80.0  40.0  40.0  _Q E  .0  .4  .8  1.2  1.6  2.0  2.4  .0  .4  .8  1.2  Log Buffer Width • River  Edge  • Clearcut  Edge  117  Literature Cited Andreassen, H.P., K. Hertzberg, and R.A. Ims. 1998. Space-use responses to habitat fragmentation and connectivity in the Root Vole Microtus Oeconomus. Ecology 79:12231235.  Askins, R.A. 1994. Open corridors in a heavily forested landscape: impact on shrubland and forest-interior birds. Wildlife Society Bulletin 22:339-347.  Belisle, M . , A . Desrochers, and M . Fortin. 2001. Influence of forest cover on the movements of forest birds: a homing experiment. Ecology 82:1893-1904.  Bibby, C.J., N.D. Burgess, and D.A. Hill. 1992. Bird census techniques. Academic Press, Toronto, Ontario.  Bright, P.W. 1998. Behaviour of specialist species in habitat corridors: arboreal dormice avoid corridor gaps. Animal Behaviour 56:1485-1490.  Brooker, L., M . Brooker, and P. Cale. 1999. Animal dispersal in fragmented habitat: measuring habitat connectivity, corridor use, and dispersal mortality. Conservation Ecology 3:4.  Brosofske, K.D., J. Chen, R.J. Naiman, and J.F. Franklin. 1997. Harvesting effects on microclimatic gradients from small streams to uplands in western Washington. Ecological Applications 7:1188-1200.  Bryant, A . A . 1997. Breeding songbirds and woodpeckers in Clayoquot Sound, British Columbia. Andrew Bryant Services, Nanaimo, B.C.  118  Campbell, R.W., N.E. Dawe, I. McTaggart-Cowan, J.M. Cooper, G.W. Kaiser, and M.C. McNall. 1990. The birds of British Columbia. The Royal British Columbia Museum, Victoria, B.C.  Campbell, R.W., N.K. Dawe, I. McTaggart-Cowan, J.M. Cooper, G.W. Kaiser, M . C . E . McNall, and G.E.J. Smith. 1997. The birds of British Columbia. U B C Press, Vancouver.  Cassady St. Clair, C , M . Belisle, A. Desrochers, and S. Hannon. 1998. Winter responses of forest birds to habitat corridors and gaps. Conservation Ecology 2:1-21.  Conover, W.J., and R.L. Iman. 1982. Analysis of covariance using the rank transformation. Biometrics 38:715-724.  Dale, V . H . , S.M. Pearson, H.L. Offerman, and R.V. O'Neill. 1994. Relating patterns of land-use change to faunal biodiversity in the Central Amazon. Conservation Biology 8:1027-1036.  Darveau, M . , P. Beauchesne, L. Belanger, J. Huot, and P. Larue. 1995. Riparian forest strips as habitat for breeding birds in boreal forest. Journal of Wildlife Management 59:67-78.  Dayton, P.K. 1998. Reversal of the burden of proof in fisheries management. Science 279:821-822.  Desrochers, A., and S.J. Hannon. 1997. Gap-crossing decisions by dispersing forest songbirds. Conservation Biology 11:1204-1210.  Ehrlich, P.R., D.S. Dobkin, and D. Wheye. 1988. The birder's handbook: A field guide to the natural history of North American birds. Simon and Schuster Inc., New York.  119  Fahrig, L., and G. Merriam. 1994. Conservation of fragmented populations. Conservation Biology 8:50-59.  Gates, J.E., and N.R. Giffen. 1991. Neotropical migrant birds and edge effects at a forest-stream ecotone. Wilson Bulletin 103:204-217.  Grubb Jr., T.C., and P.F. Doherty Jr. 1999. On home-range gap-crossing. Auk 116:618-628.  Haas, C.A. 1995. Dispersal and use of corridors by birds in wooded patches on an agricultural landscape. Conservation Biology 9:845-854.  Haddad, N . M . 1999a. Corridor and distance effects on interpatch movements: a landscape experiment with butterflies. Ecological Applications 9:612-622.  Haddad, N . M . 1999b. Corridor use predicted from behaviors at habitat boundaries. American Naturalist 153:215-227.  Haddad, N . M . , and K.A. Baum. 1999. A n experimental test of corridor effects on butterfly densities. Ecological Applications 9:623-633.  Hanski, I., and M . Gilpin. 1996. Metapopulation biology: ecology, genetics and evolution. Academic Press, San Diego.  Harrison, R.L. 1992. Toward a theory of inter-refuge corridor design. Conservation Biology 6:293-295.  Hatler, D.F., R.W. Campbell, and A. Dorst. 1978. Birds of Pacific Rim National Park. The British Columbia Provincial Museum, Victoria.  120  Huhta, E., J. Jokimaki, and P. Rahko. 1998. Distribution and reproductive success of the Pied flycatcher Ficedula hypoleuca in relation to forest patch size and vegetation characteristics: the effect of scale. Ibis 140:214-222.  Kareiva, P. 1990. Population dynamics in spatially complex environments: theory and data. Philosophical Transactions of the Royal Society of London, Series B 330:175-190.  Karr, J.R. 1982. Population variability and extinction in the avifauna of a tropical land bridge island. Ecology 63:1975-1978.  LaRue, P., L. Belanger, and J. Huot. 1995. Riparian edge effects on boreal Balsam Fir bird communities. Canadian Journal of Forest Research 25:555-566.  Lima, S.L., and L . M . Dill. 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology 68:619-640.  Machtans, C.S., M . Villard, and S.J. Hannon. 1996. Use of riparian buffer strips as movement corridors by forest birds. Conservation Biology 10:1366-1379.  Matthysen, E., and D. Currie. 1996. Habitat fragments reduces disperser success in juvenile nuthatches Sitta europaea: evidence from patterns of territory establishment. Ecography 19:67-72.  Mclntyre, N.E., and J.A. Wiens. 1999. How does habitat patch size affect animal movement? A n experiment with darkling beetles. Ecology 80:2261-2270.  Norris, D.R., and B.J.M. Stutchbury. 2001. Extraterritorial movements of a forest songbird in a fragmented landscape. Conservation Biology 15:729-736.  Norusis, M.J. 1994. SPSS basic and advanced statistics. SPSS Inc., Chicago, Illinois.  121  Orians, G.H., and J.F. Wittenberger. 1991. Spatial and temporal scales in habitat selection. American Naturalist 137:S29-S49.  Peterson, R.T. 1990. Field guide to western birds. Boughton Mifflin Company, Boston.  Rail, J. F., M . Darveau, A. Desrochers, and J. Huot. 1997. Territorial responses of boreal forest birds to habitat gaps. Condor 99:976-980.  Recher, H.F., and D.L. Serventy. 1991. Long-term changes in the relative abundances of birds in Kings Park, Perth, Western Australia. Conservation Biology 5:90-102.  Rich, A.C., D.S. Dobkin, and L.J. Niles. 1994. Defining forest fragmentation by corridor width: the influence of narrow forest-dividing corridors on forest-nesting birds in southern New Jersey. Conservation Biology 8:1109-1121.  Robbins, C S . 1970. Recommendations for an international standard for a mapping method in bird census work. Audubon Field Notes 24:723-726.  Robinson, S . K , F.R. Thompson, T . M . Donovan, D.R. Whitehead, and J. Faaborg. 1995. Regional forest fragmentation and the nesting success of migratory birds. Science 267:1987-1990.  Saunders, D.A., R.J. Hobbs, and C R . Margules. 1991. Biological consequences of ecosystem fragmentation: a review. Conservation Biology 5:18-32.  Sieving, K.E., and J.R. Karr. 1997. Avian extinction and persistence mechanisms in lowland Panama. Pages 455 in Laurance, W.F., and R.O. Bierregaard Jr., editors. Tropical forest remnants: ecology, management and conservation of fragmented communities. The University of Chicago Press, London.  122  Sieving, K.E., M.F. Willson, and T.L. De Santo. 1996. Habitat barriers to movement of understory birds in fragmented south-Temperate rainforest. Auk 113:944-949.  Smith, S.M. 1995. Distribution-free and Robust Statistical Methods: Viable Alternatives to Parametric Statistics? Ecology 76:1997-1998.  Stamps, J.A., M . Buechner, and V . V . Krishnan. 1987. The effects of edge permeability and habitat geometry on emigration from patches of habitat. American Naturalist 129:533-552.  Stouffer, P.C., and R.O. Bierregaard Jr. 1995. Use of Amazon forest fragmnts by understory insectivorous birds. Ecology 76:2429-2445.  Tischendorf, L., and C. Wissel. 1997. Corridors as conduits for small animals: attainable distances depending on movement pattern, boundary reaction and corridor width. Oikos 79:603-611.  Turchin, P. 1991. Translating foraging movements in heterogeneous environments into the spatial distribution of foragers. Ecology 72:1253-1266.  Wegner, J.F., and G. Merriam. 1979. Movements of birds and small mammals between wood and adjoining farmland habitats. Journal of Applied Ecology 16:349-358.  Wiens, J.A., C S . Crawford, and J.R. Gosz. 1985. Boundary dynamics: a conceptual framework for studying landscape ecosystems. Oikos 45:421-427.  Wilcove, D.S., and S.K. Robinson. 1990. The impact of forest fragmentation on bird communities in eastern North America. Pages 319-331 in Keast, A., editors. Biogeography and ecology of forest bird communities. SPB Academic Publishing, The Hague.  Wilcox, B.A., and D.D. Murphy. 1985. Conservation strategy: the effects of fragmentation on extinction. American Naturalist 125:879-887.  124  rt  C '  rt rt <  in oo  © ©  r- cn vo cn  00 00OOO00OO000000OO  o ©  W0 00  ©©d©©©'d©d©©©©^ co co co © © o co r~ © © © o o o m cn co© O co vo o o o ©'©'©'©'©©©©©©©o©  o ©  r- co VO CO  © ©  vo >  ©  o o o o o o  o  03 4J  O  © ©  vo o Tt CN  c  P 00  o o oo o © o o o o oo © o o d© © © © o o © o o o o o © o © o © o o © © © ©  o o o o  O TO Tt t-; vo H (N ^  P oo  vowo©oovo©Owowov-)Op  o v i o i n o o o o o o o o o o o  VO  a a  P 00  P  c  c3 ii  o rt o 3 1  p  00 c a a  T-t  ©  oo O H O H i n o o v i p q o v i  P 00  —5  ^^Or-^od©'©'©©©©  ©  © ©  ©  co rCO VO ©  ©  0 0 C 0 V 0 O O V 0 0 0 O 0 0 O © © o o vot~-coO©^wopvoopp  oo wo  © ©  00  ©o©r~~cooco©oo o r- © o o p © v o c o p c o p p © o vo © d ^r^dcScicidcicici'd'  CO CO  CO o © CO  o w o owooooooo©opowo oo ©©o©©° w ° So p 2 © ^H  © o  00  ©  © d  © Oo o o o co co o o o o C o © c o c o o o p p CO o o © d ^ d d d d d d d d d d ©  o ©  t~~ © vo o  1-  O  .3 00  o  ii  p. 00  a  ©  © co © © CO O O O CO CO © O co vo © O CO © © © CO CO © O CN © O CN © © © " © © © < © ©  a M o  o  VO  00 WO |  U  u -> 00 co ox +oo 00 00 I  ci g .2  > O0 >  o  o e l  4—»  ii  c o ii  H  o  O  CH -  03  E  >. 22CD ii tn CU oo o 'ccj S  S. ^ £  u  T3 ii S e <L> no o  oo C  p K >  32  o  o o T3  CO CO D "o C  o  oo I-I  o ft  ©  © d  o ii c  wo  wo O  1)  0)  ©  ii  OX) D <D  O  125  o o  OO o o om o o o o o o o o o  O O  o o o —i  o  0  m  o >n ino Tm -H Tt T-H  oo oo o o o  T-H m m o o o ^ 0 0 0 0 0  CO COo oo oo d © d •© © © d CD  O SO  co co  p  o o o o o o o o o o  O O O oo in O oo o o o in T—i o m  o o o o o o o o p p  o o o  CD CD CD CD CD  O O O  o  C^) co c5 <d> <d> <d>  CD  C^5 C^5 ^H C^5  o o  r - r— o sq so p  ui  CD CD CD  o oo oo o o m 'odd CO r - co o o p p SO C O CN d d d  so  C^  co r- o co CO SO O CO d d d d d d d  o o o o  o  o o o o  o  o o o o  o  o  CO No O o o o o o  O O co CN O m O O  o O  CD CD CD CD  Oo CD C o o o o co r~ CD CO p soCD CD CD d d CO CD  o o o o o o o o CN d d d  o  >n  o o oo oo in c> d d  O CD CD CD CD O CD CD CD CD CD O CD T-H O O O O O  T-H O O 00 in CD CD m Tt d d d C-- o o S O o o Tt CD CD CD  r I OO o oo —  oo 00 o in o o co p O CD CD CD CD C CD  o d  r- co o o SO o o C oO oo oo  O O  o o o o o o o o p p  00  00 00  O CD CD m  CD CD CD CD CD^  rsi co co  o o o o o o o o o o CD CD CD CD  O O O O O O O O O O CD CD CD CD CD  d>  <n p  O O  o  CD CD CD CD CD CD CD CD  O  'odd o o o op p p p oo d d CD  p  CN  d 6 6 6 d ci co o o o o co o co co o o o  CN O O O O O  O  03 CO  SH  ID  SH CD  1 j3  <D  "J3  c  03 5  < > -3 ^ O PH 00 c T3 T3 00 oo T3 C T3 (D  <u, 00  03 S'H CD c/> w  35 ffl O O  CD C O  3 HO  c  00 -bC O 3  • i-H  HO,  3 c  .2 'EH  03 PH  • i-H  O  '3 cr  O  o rt c  1  * rt  PH  ^  'EH  00 rt  CJ (D  O aJ  co U  126  T3 3  00  03 JS  O O  Q  rt  o co  >  O o O  c CD  o o o  o o o  o o  o o o o o o o o o o o o r - ( O O O O O O  o o d  O O C O O C N T - H S O O  o o t ^ c N s o c o T t o oor-imTJ-cNcoT—< o o Tf- o  o o o o o  o r-~ o o so o fl M H N  o  o o o  o o o  o o o  >, X) T3 CD N  'D o  00 CD  O  t3 u  T - H O O S O O O O C O O O  Q  sqiOT-HOOor-;00  CO  CD  I n O T t O r - H O ' - H C  'o  o o  o o  c5 CD d> ci  o o CD  o o d  cn co m o co co co o  o o  o o  o o  o o  cOTtoos'-HO'-HOomoooco incNOoocoincomr-Hmor^  o o  O oo o m  o o  c o r - c o o o o o o o c o s q c o p p p p o p M i N r O O f N O r H O ' - i CN  S3  o3 CD  GO  T—I OO 0O o co i n i n o  c5 c5 CD c5 c5 d>  CD  PH C/3 JP  o o  Q  CO  CD 00 T3 CD  O r - l O N O C N C N r - I C N O ^ O ' - H r - l O O O  CO  Ot^COOt~^COOCOCOt^COOOOOCO ONOcoosocoincocoNqcoppppco  c  *•—» d o  03 CD  OTtind^'-HCNrHddd'^'Hddd  s03 4}  o o 03 co  G o  O O  Q  W CO  Q  o o  T—i oo as i n i n o Tt i n o c o o o o c o i n o i n o oo T-H so o o T—lor-^inor^T-HO i n so i n T-H CN d Tt T - H O r - H O d T - H ^ H O  CO  o o o o o o r~ r - o o co r» o o o o o o O SO O so SO o o CO SO OCOTtO\OCNOCNOO>-HOO'—lOO  c  03 CD  >  O o  CD  T3  03  -a  § T3  o3  Q  o o m oo O O O O O O O O O O o o T—t i n O O O O O O O O O O O O O T-H d d d d d d d d d d d  O O  CO  O o o  G  o o  o so o  O o  o o  C C  O O  C C  O O  O O  O O  O O  O O  O O  O O  O O  d> c5 c5 ci—<CD'C5C5CSC5C5C5C5CDC5C5  CD  O O  O O  O O  G  a CD  >  I T3 u  S-H  CD M O CD  CD  .H o • i-H 1-H  03  lH CD C T3 G CD P.  oa t> '5 CD PCO  CD  CD  o  c/l CD  CD  l-H O PH  rt co  lH O  PH  PH  T3 O O  T3 CD 4 -—* 03  T-H  CD 50  -a  CD i  C CD PH  o  127  o o o o p dd co r~- o o o ^ o o VO o o ro vo o o o VOfNOCOOOOOO  00 WO O CO i—i © o wo o wo co p p d H d H ci d o co — r O CO CO o o co vo© CO CO O O o o O © © i-H  COWO^HOOOOOOO  oo © o o o o o wo o p p p p p  vdi-Hwodddddd c^i>r~-coooooo vovovorooopop wo d c o d d d d d o  r-O oo oo oo op op oo V  h o H oo in o if) M O N <(1 H o in ir~j © ro C> i—i H  O'—t wo wo o p p p p  osoovooomooooo r-woTtovowoppo co' d n i N r N d d d d o o o O CO r- r- o O co O O O o CNd d d d V vo cO N O  OO o co co co p p p TJ- N H d d d oo o © o o o wo o o o d o © © d d w o o o o © o o vo o VO o o o o o CO  ci ci ci ci ci ci ci  CN WO  o o o o o — < o o o o CN CN O --I O © r-o »o o o r-q o v o v o o p CN CN d d d d  © 00 © oo o o © wo © wo o ©  CN d d d d dd o o CO o o o o o CO o CN O O O O O O  © vo O WO O OO 00 pop © wo wo d co d —J d d d r- o r- o co co o v p o O vo O co co d CN © d d d ©  00OOOOOOOO m o o p p p p p p d o d o odd do COOOOOOOOO coOOOppppp  o o 00 o o o o © © 1-1 d d wo c o o o p o vrco oo p p q d d CN d © d  o o o o o o odd  CO CO CO  d d d d d d d  m oo o oo o vo o in p co CN in O r-l p d c-© <-H o o © d VO COCOt~-voCOCO o o o r^oo^oo^oo  d d d d d d d d d  Tt  wo vd o  CN  CO  wo  o o  O oo oo © wo wo odd  r- o o o CO CO co co T-H H O V O OO—O OO O o o o o co O r- o © o p p  o o d o o o  d d d d ci  o o o o o o o o o o o o d d d d d  o 13 oo T3  < 0D 0  c  3  o '3  cr  3  2 ^ 0  pi oo Q £ O  2  c IS 00 o c Pi '% c  x:  00  o oo '3 Pi : X oo V tiE s O <L> o < U oo PiZ o  T3  128  Appendix 6 - Summary of mean and standard deviations of bird densities by habitat guild in buffers of varying widths. Guild/Species  Very Narrow  Narrow  Medium  Wide  Old-growth  Mean(SD)  Mean (SD) Mean (SD) Mean (SD) Mean (SD)  Forest Generalists  0.37 (0.32)  0.27 (0.22)  0.25 (0.06)  0.21 (0.04)  0.20 (0.03)  Forest Interior  0.00 (0.00)  0.00 (0.00)  0.01 (0.01)  0.04 (0.02)  0.05 (0.02)  Open-edge  0.32 (0.38)  0.04 (0.02)  0.04 (0.03)  0.03 (0.02)  0.02 (0.02)  Riparian  0.24 (0.25)  0.10(0.04)  0.10(0.05)  0.05 (0.03)  0.02 (0.01)  Ubiquitous  0.29 (0.29)  0.01 (0.02)  0.05 (0.02)  0.03 (0.01)  0.04 (0.01)  129  Chapter 5: Influence Of Vegetation Structure On Use Of Riparian Buffer Strips by Birds  Introduction Numerous studies have shown that both structural diversity (e.g. MacArthur 1961; MacArthur et al. 1962; Willson 1974; Rotenberry and Wiens 1980; James and Warner 1982; Hodges and Krementz 1996, Kilgo et al. 1998) and plant taxonomic diversity (Holmes and Robinson 1981; Robinson and Holmes 1984; Rice et al. 1984; Rotenberry 1985) influence the composition of bird communities. Many earlier studies of bird/habitat relationships were conducted in non-riparian areas (e.g. MacArthur et al. 1962; James and Warner 1982). It is therefore of interest to ask: Do riparian areas show similar patterns? If riparian birds do respond positively to increasing diversity in the riparian plant community, what explains this response?  The composition of riparian bird communities in logged areas typically shifts away from forest-interior species towards open-edge species (Bryant et al. 1993; Darveau et al. 1995; Hagar 1999; Pearson and Manuwal 2001; my Chapter 2), but relationships with the associated plant community in buffers are not well understood. One recent study of bird/habitat associations in buffer strips of riparian forests of the Pacific Northwest (Hagar 1999) found that half of the species in small headwater streams that increased or decreased with buffer width were predicted by densities of hardwoods or conifers. The bird/habitat associations in riparian plant communities along larger streams and rivers still need to be investigated. Broad riparian zones downstream may provide a more diverse plant community to support higher bird diversity and abundance than in headwater streams.  Buffer strips provide some protection from the negative effects of logging (Newbold et al. 1981). It is therefore useful to know to what extent plant communities within riparian buffers must remain intact to preserve avian distributions similar to those in undisturbed  130  forests. Plant communities within buffers after logging adjacent areas should initially be similar to plant communities in old-growth forest; however, altered microclimate at the buffer edges and perhaps extending into the buffer should eventually produce changes in plant communities particularly in narrower buffers (Ranney et al. 1981; Williams-Linera 1990; Camargo and Kapos 1995). Examining the associations between bird community composition and plant structure within a range of buffer widths provides one approach to answering this question.  This study has two main objectives. First, I examine how vegetation structure varies over a range of buffer widths and whether differences in vegetation patterns in buffers are due to: a) pre-existing differences between the buffers or b) due to vegetation dynamics in the buffers? If fragmentation alters microclimate and light regimes, these changes should vary with buffer width and the greatest differences should be observed in the edge-dominated narrow buffers. In some forest ecosystems, second growth at forestclearcut edges develops quickly within the first decade after cutting (Ranney et al. 1981; Williams-Linera 1990; Camargo and Kapos 1995) particularly shade-intolerant species (Wales 1972; Ranney et al. 1981). Narrow buffers 5-7 years post-harvest should have higher densities of deciduous trees and shrub cover due to greater light at edges (Ranney et al. 1981; Oliver and Larsen 1990).  Second, I used correlations to compare the relative power of vegetation attributes and buffer width to explain patterns of bird richness and abundance. In particular, I expected species richness and abundances of aerial foragers, foliage gleaners, open-edge species and riparian specialists to be positively associated with growth of deciduous trees and shrubs. Aerial foragers such as flycatchers are associated with deciduous trees (Gilbert and Allwine 1991) that often dominate river bottoms (Puettmann and Hibbs 1996). Both flycatchers and foliage gleaners may respond to the richer insect fauna at forest edges (Hansson 1983; Helle and Muona 1985) while open-edge species may respond to increases in vegetation diversity (Freemark and Merriam 1986).  131  Methods  Study area The study sites lay within 9 watersheds on the west coast of Vancouver Island, British Columbia (Fig. 5.1a). Details of study sites and vegetation are given in Chapter 2. I selected old-growth buffers 3-125m in width left along one side of the river and at least 1 km away from each other. Adjacent clearcuts were 3-5 years old at the start of the study. Old-growth controls were located in sites of undisturbed continuous forest along rivers. Sites were standardized for elevation (< 500m). I refer to the buffer categories as very narrow (3-10m), narrow (20-25m), medium (36-44m) and wide (100-144) in the analyses.  Vegetation Sampling During 1997 I sampled vegetation in 20m radius circular plots (0.13 ha). Plots were placed at three stations 150 m apart along a 500m riparian transect running parallel to the river (Fig. 5.1b). Details of vegetation sampling are given in Chapter 1. In addition, I converted diameter to basal area (m /hectare) for coniferous trees, deciduous trees and 2  snags. Density (number ha" ) was also recorded for these categories and for saplings 1  (trees <10cm diameter).  Habitat Associations  Details of bird sampling are found in Shirley (Ph.D thesis, Chapters 2 and 3). A summary of species richness and abundance is summarized in Table 5.4. Buffer width for wide buffers was truncated to 50m for this analysis to include only birds whose distributions reasonably overlapped the vegetation plots.  132  Data Analyses  Before conducting statistical tests, all data were tested for normality using the ShapiroWilk statistic and homogeneity of variances by analysis of standardized residuals from linear regression. Data that violated the assumptions for linear regression were successfully transformed using a log (x + 1) transformation or arcsine transformation for percentage cover. I used an alpha value of 0.10 to represent a high conservation cost of making a type II error (Smith 1995; Dayton 1998). Data were analyzed using SPSS for Windows 6.1.4 (Norusis 1994).  Vegetation  To minimize the number of vegetation variables, I combined 158 vegetation variables to produce 10 variables that focused on broad tree classifications, snags, coarse woody debris (CWD), richness of shrubs and forbs, and cover of shrubs, forbs, grasses and mosses for the analyses (Table 5.1). Variables for shrub species richness and cover for each height category were averaged. Coarse woody debris was expressed as volume per unit area (m ha" ) based on a formula developed by Van Wagner (1968). To assess if 3  1  there were general vegetation gradients that were related to buffer width, I performed a principal components analysis on the correlation matrix using the 10 variables. I used varimax rotation to improve the separation of components. I retained the first three components with eigenvalues that exceeded the broken stick model (Legendre and Legendre 1983). The Bartlett's sphericity test was rejected (P = <0.001) which confirmed that the components were not random. I then performed linear regressions against buffer width for the first three principal components. Seven vegetation variables had high correlations (> 0.5) to components 2 and 3 (Table 5.2) and I performed linear regressions for these individual variables against buffer width and controls.  Habitat Associations  133  I examined associations between vegetation variables and bird diversity and abundance using a multiple regression analysis with bird species richness and abundance as the dependent variables. Seven guilds and 10 species were recognized from the total set of species detected on these plots (Chapter 2 of thesis). I chose six species that differed in abundance between buffers and controls of equal width during 1996-1998. These were American Robin (Turdus migratorius L.), Golden-crowned Kinglet {Regulus satrapa Lichtenstein), Hammond's Flycatcher {Empidonax hammondii Xantus), Pacificslope Flycatcher {Empidonax difficilis Baird), Swainson's Thrush {Catharus ustulatus Nuttall), and Warbling Vireo {Vireo gilvus Vieillot). I also included 4 other common species (Chestnut-backed Chickadee (Parus rufescens Townsend), Hairy Woodpecker {Picoides villosus L.), Varied Thrush (Ixoreus naevius Gmelin) and Winter Wren {Troglodytes troglodytes L.)). I deleted those sites where a species was not known to occur, (e.g., Hammond's Flycatcher in Klanawa valley) (Bryant et al. 1993) for analysis of that species' associations. I restricted the number of vegetation variables to those with high correlations to components 2 and 3 in the principal components analysis since these components varied with buffer width. I used stepwise regression and a significance level of 0.10 for each variable to enter the model. I performed multiple regressions of bird richness and abundance against vegetation variables with and without buffer width as a variable, to compare their relative explanatory power.  Results Vegetation  Vegetation patterns could not be explained by a single gradient since each axis accounted for only 18-33% of the variation. Three major axes described riparian vegetation (Fig. 5.2). These first three principal components together represented 78% of the variance among the vegetation variables, with the first component representing 33%. The principal components accounted for most of the variance (>70%) in individual vegetation variables except for percent cover of grasses. Principal component 1  134  distinguished deeply shaded areas with high conifer and snag density and low deciduous sapling density and grass cover (Table 5.2). Principal component 2 distinguished areas with tall shrub cover, high deciduous tree density and amounts of downed wood, while principal component 3 distinguished areas with low tree density, but high shrub richness and forb cover. Principal component 1 showed no relationship to buffer width; principal component 2 was positively related to buffer width, and principal component 3 showed a negative relationship (Fig. 5.3).  Of the seven individual variables, four showed relationships with buffer width (Table 5.3). Deciduous tree density declined from 288 stems/ha. in controls to 56 stems/ha. in narrow buffers (Fig. 5.4a). Shrub cover at 2-3m declined from 16% in controls to 7% in the narrowest buffers with maximum cover of 19% in wide buffers (Fig. 5.4b). In contrast, shrub richness reached its highest levels in very narrow buffers with 12 species and declined with increasing width (Fig. 5.4c) to 7 species in wide buffers and controls. The decline in shrub cover with decreasing buffer width and increase in species richness were due in part to reduced dominance by salmonberry as light levels increased (Fig. 5.4d). Forb cover followed a similar pattern to shrub richness decreasing from 64-68% in narrow buffers to 37% in controls (Fig. 5.4e); cover was also more variable in wide buffers and controls (Pearson Chi-square = 144.57, P = <0.001). Conifer density, downed wood and moss cover showed no significant relationship to buffer width (Fig. 5.4f-h). There was a trend for moss cover to decrease in narrow buffers, but variation among sites was high.  Habitat Associations of Birds  Bird species richness in riparian buffers was negatively related to shrub species richness and positively related to downed wood volume. These variables explained 69% of the variance (Table 5.5). When I added buffer width to the model, however, it was the only predictive variable, explaining 62% of the variance.  135  Abundances of Neotropical migrants and 3 of 7 guilds: aerial foragers, foliage gleaners and riparian specialists, were positively related to deciduous tree density (Table 5.5). Foliage gleaners and riparian specialists were also positively related to conifer density and negatively related to tall shrub cover respectively. These variables explained 48% and 66% of the variance in the abundance of birds in the above two guilds. Abundances of ground gleaners and forest interior species were negatively related to shrub richness and positively related to downed wood and tall shrub cover respectively, accounting for 59-63% of the variance. Abundance of open-edge species was positively associated with downed wood and negatively associated with moss cover, together accounting for 53% of the variance. When I included buffer width in the model, it was the only predictor for foliage and ground gleaners, explaining 65-71% of the variance in abundance. Abundances of Neotropical migrants were positively related to buffer width and downed wood volume that together accounted for 83% of the variance. Aerial forager abundance was positively related to buffer width and deciduous tree density; these 2 variables together explained 70% of the variance. Abundances of riparian, forest interior and open-edge species were not correlated with buffer width.  At the species level, abundances of Chestnut-backed Chickadees and Hairy Woodpeckers were weakly, but positively related to conifer density (Table 5.5). Abundance of Hammond's Flycatchers was positively related to deciduous tree density and negatively related to shrub species richness; together these variables explained 89% of the variance. Winter Wren abundance was also negatively related to shrub richness and positively related to downed wood volume, accounting for 60% of the variance. Abundances of Pacific-slope Flycatcher, Swainson's Thrush and Varied Thrush all had positive relationships to tall shrub cover which explained 38-66% of the variance. Two species, Golden-crowned Kinglet and Warbling Vireo, had no significant predictors of abundance. Buffer width did not help to predict abundances of Hammond's Flycatcher, Pacific-slope Flycatcher, Swainson's Thrush and Varied Thrush. Buffer width was the only significant predictor for American Robin (68%) and Chestnut-backed Chickadee (37%) abundance, and the first entry for Hairy Woodpecker and Winter Wren abundance.  136  Hairy Woodpecker abundance was also negatively related to tall shrub and positively related to forb cover; together the variables accounted for 84% of the variance. Abundance of Winter Wrens was also positively related to downed wood and negatively related to deciduous tree density, which together accounted for 95% of the variance.  Discussion Changes in Vegetation With Buffer Width  My research demonstrates that vegetation structure varies with buffer width. Deciduous trees and shrubs showed clear differences among buffer width comparisons, while other variables associated with coniferous forest, e.g. conifer density and snag density, showed no pattern across varying buffer widths.  Deciduous Trees Contrary to my prediction, deciduous tree density increased with buffer width. Deciduous trees are a prominent feature of riparian forests (Pabst and Spies 1999; Pearson and Manuwal 2001; Shirley, Chapter 1). I attributed the low density of deciduous trees in narrow buffers to the incidental removal of trees during harvesting operations; however, there are several alternate hypotheses. Fast-growing deciduous tree species such as red alder {Alnus rubra (Bong.)) (Burns and Honkala 1990; Puettmann and Hibbs 1996) in wider buffers may benefit from increased light near the river and reduced competition from shrubs present in edge-dominated narrow buffers (Comeau et al. 1996). Alternatively, deciduous trees could be more prevalent along downstream, wider reaches of rivers with lower gradients (Campbell and Franklin 1979; Lock and Naiman 1998; Pabst and Spies 1999). That is, wider buffers may have higher densities of deciduous trees because they represent sites further downstream. There was, however, no correlation between deciduous tree density and stream width. Finally, forest managers may have chosen to retain wide buffers at sites with a high deciduous component to allow  137  them to harvest the more valuable conifers. In this case, the patterns represent the forest composition prior to harvest. Further research and discussions with forest managers would help resolve the three alternate hypotheses.  Deciduous and Conifer Sapling Density I predicted that sapling density would be higher in narrow buffers because growth rates would increase with more light (Wales 1972; Williams-Linera 1990). This was not the case for either type of sapling. There was no direct relationship between deciduous sapling density and buffer width. Saplings may be prevented from reaching higher densities in narrow buffers due to competition with shrubs many of which grow rapidly under high light conditions (Comeau et al. 1996; Messier 1996). Chen et al. (1992) found that edge effects on conifer saplings along clearcut edges in the Cascade Range of the western United States varied by species. Regeneration in my forests is mostly of amabalis fir (Abies amabilis (Dougl) Forbes) (unpubl. data), a species that does not grow well in edge environments (Chen et al. 1992).  Shrubs and Forbs  Contrary to my prediction, species richness of shrubs and forbs decreased with increasing buffer width. Two mechanisms might explain this pattern. First, competition for light is a primary factor controlling the organization of forest plants (Spurr and Barnes 1980). Several forest shrubs may be outcompeted for light by dominants and thus respond positively to canopy removal. Typically, growth of woody shrubs and seedlings dominates the early stage of edge development within 5 years of harvest (Ranney et al. 1981; Alaback 1982). Salmonberry (Rubus spectabilis Pursh) was the dominant shrub in wide buffers and controls in this study (unpub. data). This species also predominates riparian areas in coastal Oregon (Pabst and Spies 1998) by displacing other shrub species to different habitats. The decline in cover of tall shrubs in narrow buffers is consistent with a reduction in dominance and is reflected in decreases in salmonberry cover. Hodges  138  and Krementz (1996) also noted a similar decrease in tall shrub cover with decreasing buffer width over a wider range of buffer widths in bottomland hardwood forest. Second, higher species richness in narrow buffers may reflect colonization by non-forest species such as fireweed (Epilobium angustifolium L.) (Halpern and Spies 1995) that tolerate higher light levels near the clearcut edge (Ranney 1981; Chen et al. 1995). Conifers and Snags  I expected conifer and snag density to be lower in narrower buffers where plots included part of the clearcut and were affected by tree removal. While the principal components analysis defined a gradient of shaded coniferous riparian forests, surprisingly, there was no clear association with buffer width. M y prediction assumed a random distribution of conifers with riparian forests; however, conifers along larger streams with high fluvial disturbance have lower densities (Poage 1994; Pabst and Spies 1999) and become more dominant in uplands. A previous study (Hagar 1999) found lower conifer density in logged areas, but this study was located along small, headwater streams where conifers occur in higher density in riparian forests (Minore and Weatherly 1994).  There was a trend for snag density to increase in medium and narrow buffers. In contrast, Hagar (1999) found higher snag densities in unlogged areas. The number of snags likely varies with factors such as tree species composition, forest age before cutting, topography and time since harvesting. Amabilis fir (Abies amabilis (Dougl) Forbes), a dominant tree species in these forests, is particularly susceptible to wind damage (S. Mitchell, pers. comm). M y results are consistent with other studies that found high snag densities near clearcut edges due to wind damage (Steinblums 1977; Williams-Linera 1990; Chen et al. 1992) and in narrow buffers in boreal forest (Darveau et al. 1995; Ruel, unpubl. data).  139  Habitat Associations of Birds  I expected vegetation structure to be associated with species richness across varying buffer widths. Instead, buffer width alone explained patterns of species richness. This result is consistent with studies that have examined the influence of vegetation on a microhabitat scale in homogenous habitats (Willson 1974; Martin 1981) and in newly isolated fragments (Nilsson et al. 1988; Schmiegelow 1997). Two other studies that examined the influences of both area and vegetation in buffer strips (Hagar 1999; Saab 1999) both found that vegetation played a more important role than buffer width in riparian areas. In particular, bird species richness was positively associated with large hardwoods (Lock and Naiman 1998; Hagar 1999). In coastal coniferous forests, large hardwoods reach higher densities in riparian areas (Pearson and Manuwal 2001; Shirley, submitted) so that buffer strips may have disproportionately high numbers of deciduous trees. Increases in bird species richness in my study were mainly among the resident foliage gleaners, ground gleaners and bark gleaners. Many of these birds use large hardwoods for nesting and/or foraging (Enns et al. 1993), but also use a variety of forest habitats, and would therefore be expected to increase with the amount of total forest habitat represented by buffer width. While hardwood density may be important, densities were highly variable among sites, with coefficients of variation ranging from 27-97%. This high variation may obscure a clear association with the number of bird species.  Vegetation variables predicted the abundance of several guilds and species. In three guilds, vegetation was the best predictor of abundance. As predicted, deciduous tree density and tall shrub cover explained some of the variance in abundance for riparian specialists after correcting for area, but did not do so for aerial foragers, foliage gleaners or open-edge species.  Abundances of aerial foragers and foliage gleaners were positively related to buffer width. On my sites, aerial foragers are composed almost entirely of flycatchers, a group more prevalent in older forests than in younger forests in eastern North America (Maurer  140  et al. 1981). Hammond's Flycatcher was the only aerial forager found in narrower buffers. In contrast to other areas studied (Sakai and Noon 1991; McGarigal and McComb 1992; Willson and Comet 1996), this species is strongly associated with riparian areas in the forests of western Vancouver Island (Waterhouse and Harestad, Simon Fraser University, British Columbia, unpubl.; Shirley, submitted). In wide buffers, the Pacific-slope Flycatcher is more abundant and contributes to the strong positive association with buffer width. Foliage gleaners included dominant species such as Swainson's Thrush and Chestnut-backed Chickadee that were more common in wide buffers. These were associated with both hardwoods and conifers; however, 3 of the 6 species in this guild, Chestnut-backed Chickadee, Warbling Vireo, and Western Tanager (Piranga ludoviciana Wilson), are hardwood-associated (Carey et al. 1991; Lock and Naiman 1998) contributing to higher guild abundance in wider buffers.  My findings support my prediction that abundances of riparian specialists would increase with increased densities of deciduous trees and shrubs. Riparian specialists were indeed positively associated with areas with more deciduous trees, but the relationship was weak. This association is not surprising as this guild is dominated by the Hammond's Flycatcher (73% of observations) whose abundance is strongly predicted by deciduous tree density. When I removed this species from the analysis, there were no significant predictors for the abundances of other riparian species. This last result may be due to 1) the rarity of these species and resulting low sample sizes, and 2) species like the American Dipper (Cinclus mexicanus Swainson) and Belted Kingfisher (Megaceryle alcyon L.) that are strongly associated with water may not be tied to specific terrestrial vegetation characteristics. For example, dippers need fast-flowing streams and overhanging vegetation while kingfishers need mud banks along rivers for nest sites. These species typically maintain linear territories along rivers and streams (Manuwal 1986), and my sampling design may not adequately reflect their habitat associations if they are more strongly influenced by stream rather than terrestrial habitats.  141  Vegetation associations were strong for riparian specialists, open-edge species and forest interior species. In general, I expected that vegetation associations to be strong for habitat specialists, but not for generalists. This latter pattern was found in grassland/steppe birds (Wiens and Rotenberry 1981) and riparian forest (Rice et al. 1983, Hagar 1999). The pattern here, however, varied among individual species. Vegetation associations with abundance were significant for 4 of 8 species. These included the Hammond's Flycatcher, a riparian specialist and Pacific-slope Flycatcher, a forest interior species. Two other species, Swainson's Thrush and Varied Thrush, used both riparian and upland habitats. During the breeding season, the two thrushes are closely associated with deciduous, berry-producing shrubs that are distributed throughout these moist forests. Of the 4 species whose abundance was best predicted by buffer width, all were forest generalists. I know of only one other study which explicitly compared the relative importance of area and habitat to breeding birds in buffer strips within western coniferous forest landscapes (Hagar 1999). In Hagar's (1999) study, tree density explained more variation than buffer width for five species, of which three were open-edge species and two were forest-interior species.  M y study is among the first to explicitly examine area-dependent changes in vegetation structure in buffer strips of varying widths along rivers and their associations with riparian bird distributions and abundance. Nevertheless, it is limited in several ways. First, I examined bird associations with broad structural vegetation attributes that could be clearly defined in a management application. Other characteristics such as within habitat heterogeneity (Freemark and Merriam 1986; Nilsson et al. 1988), plant taxonomic diversity ((Holmes and Robinson 1981; Robinson and Holmes 1984; Rice et al. 1984; Rotenberry 1985), and foliage height diversity (MacArthur 1962; Willson 1974) may also vary with buffer width. In addition, certain characteristics associated with bird distributions may reflect an important association through a correlation with another attribute (Young 1996).  142  Second, I considered the influence of vegetation structures at a patch level. Future studies should determine how local bird distributions are influenced by landscape or regional factors such as the distribution of serai stages across the landscape (Wiens 1989b; Wiens et al. 1993; McGarigal and McComb 1995), or land-use patterns in the surrounding matrix (Stauffer and Best 1980; Saab 1999). Third, I used foraging and habitat guilds to group species with low individual abundances assuming that species within a guild use resources in a similar way. Species within the same guild may, however, vary in their use of specific vegetation structures or species (Block et al. 1984). For example, foliage gleaners may forage primarily on deciduous or coniferous structures. Further research should incorporate finer categories of habitat use (Short and Burnham 1982; Verner 1984) on which to base decisions.  Management implications Overall, my results argue that the area retained in buffers is the most important factor for maintaining natural riparian bird communities. Forest bird communities in my study were dominated by forest generalists that, by definition, do not generally have tight associations with particular habitat elements. Some species and guilds such as riparian specialists, forest interior and open-edge species however, were associated with certain vegetation features such as deciduous trees and shrubs that vary with buffer width. Structures associated with conifers were of secondary importance for most species; however, I did not examine habitat associations in upland habitats where conifer and snag densities are higher (McGarigal and McComb 1992; Shirley, submitted). Snags may influence the distribution of forest generalists and upland specialists, particularly cavitynesters. Narrow buffers with high shrub richness and forb cover may provide habitat for open-edge species and certain species not closely associated with the forest vegetation. Wider buffers (>100m) may benefit not only those species associated with coniferous upland forests and forest generalists sensitive to buffer width, but also those species whose abundance is best predicted by deciduous trees and tall shrub understorey. For  143  these latter species, retention of large deciduous trees and tall shrubs (even) in narrow buffers would be useful.  144  o co  o CD  SD  e o  LO CM Tt  LO Is  PH  •  t-H  CD O 00  CM CNJ  Tt  CD  00 LO LO LO  d CM Tt  CO  r-co  cn o LO LO CM  o CO  Tt 1—  oo CO  o I-  CD CO CM  T—'  CD  d  Tt  LO CNJ CM O CO  00  O  00 LO r--" CD  CD  LO CM CM CNJ  co oq  LO I-  o oo  CD cq  d  Tt  T—  co oo  oo q  co Tt  Tt  ,—-  s  cn CD  SH  ,u  5  o  LO o LO  q  SD  J*H -P C  o Tt  'G  <03  c\i  if  PH  Tt  CNJ  T3  s  1—  CM  .3  CD I-'  LO Tt  d CM  oo o CD LO  co cq  I-  LO  CD CD CM  d  CM oo l<  CM CNJ I-' oo  co LO  cn  cn  s  T—  -  oo  s  cn  co co CD  ICD CO CD  CD O co  Tt  LO co  Tt  00 LO CO  LO cn I-  Tf  cn I-  CO  d i —  LO co d CM  cn  CM  oo LO co CM  O LO CO CNJ  oo Ico CM  cn  s  Tt  C 03  S3  CO  II  CO Tt  SD  a  o CM  I-H  60 i  LO I-  o LO  LO  1 1  i —  co  o .g  ro  co CO CM I-  o oq  LO Ico LO  00 00  CD oo CO LO  oo LO CD I-  CO q  Tt  LO  o  co I-  CM 00  o  CD  I00 CM  00 CM  IX  CM  oo CM LO oo  CNJ r-.  O  T j  s  s  d  o cq  00 CD LO  LO  r-CD  d  d  LO  Tt  CM  cn CM  o CM CD  CM co 1^  0)  e .o  -*—»  io3  CD DJ) CD > <4-H  ^3  IX  o c  .2 O  CD  56202.  >  SD  s  CD  T3  lx  03  Tt  CM  r-  s  CD  ^  O) CO  CD Is  I-  s  s  co  CM CM  00 CM 00  s  co  CM I-  cn  s  cn CD CO  s  o  s  Tt  II-  d  .11239.  s  .33118.  io3 >  78 41.  'G  1—  56120.  SD  .s  s  o  Is  d  Tt  oo CM  s  cn cn  Tt  CD LO  oo co co  s  o cn  CM co od  CD T-  c  03  T3 C  ON ON  S2. «  c3  C  03  CD  03  CD > =3 CD  CD P  c  O  .o E  1  O  CO  03  c Q  a  >  p. CO  § p  o  O  CO  co  CO P  P  o  rj -g o Q  CO CD CO  c CO  o © p o  O  CD P  O.  co co rj o r; •p o  CO  CD  Q  JP P CD >  o  O  i  CO CO CO  O  CO CO  o  co  W  U-  I-  _o b co  145  TT  CD  CM  co  LO  co  Tt Tt  CD  00  Tt  o  co O  C\i  CM  CD  CD  ^  co  CD  CD  Tt Tt  oo  CD  oo  cvi  Tt  d oo CM  LO CD  o  CO CM  r--  CO m o  co cn oo  oo Tt  co CD  -i-  LO  CO  CJ)  T -  oo  Tt  c\i  co  CD  LO  o  oo  LO LO  LO  d co Tt  CM  CD  oo co  CD oo oo  CD co CD oo CD  CO CO CD  c .c o lir CD  o CD  I*  CO  LO CM  cd x:  be  .Q  o >  CO  o o  rj k_  CO  CD LO  c  x: o  CO CO CD  CD CD  CD CD CO CD > <  c o D  Table 5.2 - Correlation coefficients between first three principal component scores and vegetation variables (n=15). Only coefficients >0.20 are shown. Components  Vegetation Variables  Deciduous sapling density  1 -0.89789  2  3  Snag density  0.79604  -0.27818  Moss cover  0.71875  0.54738  Grass cover  -0.66508  0.31111  0.63891  -0.51430  Conifer density Deciduous density  -0.31011  0.87085  -0.21247  0.85943  Tall shrub cover (2-3m) Downed wood  -0.21182  0.75794  0.29525  Shrub richness  0.25520  -0.37146  0.85554  Forb cover  0.51127  0.30141  0.75410  Table 5.3. Linear regression statistics for vegetation variables vs. log-buffer width in riparian buffers and controls* (n = 15). Variables selected were those with correlations >0.50 with principal components 2 and 3. R  P  SE  F  68.77  68.26  1.02  0.07  0.332  0.55  0.18  9.66  0.43  0.008  b  -0.21  0.11  3.76  0.22  0.075  a  6.53  3.12  4.36  0.25  0.057  a  Shrub richness  -5.67  0.90  39.61  0.75  0.000  b  Downed wood  539.18  500.97  1.16  0.08  0.301  7.59  6.17  1.51  0.10  0.240  Attribute  Conifer density Deciduous density Forb cover Shrub 2-3m cover  Moss cover  Slope  2  Bold font: a  -F<0.10  b  -P<0.01  * - width for controls was set at 200m representing the entire grid width  148  o  P  ,—1  ,—i  o  O O  CO  cn  cn cn cn  oo O CN  O  IX  Q  o o cn  vo r-;  oo m d  O p od  CN  cn cn  '  OO  cn  oo in d  cn vq  cn cn ©  o  vo d  vq  m  i—i  o o CN  oo in d  OO OO  VO d  o p cn  vq  O O  O  o  cn CN  o  o  in ©  O o d  12  >  CO  o  ll  Q on  oo  IX  vo  T t  O  T—1  cn  Tt  o  o  CN  m  ^H.  i—1  t3 CD  Q co  "3  Tt  in vo CN  o cn *—,  o  CN  Tt  CN  d  d  cn cn vd  cn oo CN  r-  Tt  O  IX  rVO CN  in  i—i  p  1—1  d  •  3 O  s  o  O d  oo cn d  o  CN  T t  cn cn cn  o  cn  o  CN  o  VD CN  ,—1  ^H  CN ON  VO  T t  T t  o CN  oo d  c<3  C_>  O  U  <o Cu CO  3  o  'o. o a o CD 2  T t  d  o '  1  CN  oo  CN  CN  T t  o  o <D CD CD  J3  oo in d  cn cn d  feb  CD  o d  CN o\  3 p .  o  a  CD  3 o3  <D  3 o3  CD  _o>  "5b  "5b  SD  S3  m  CD  "o  o  CD  Cu  T3  3 3 O in  a  CL,  3  3  O PH  149  o Os d  o  OS  d  d  o o d  o o  o © d  T 1  OS cN d  oo  co SO d  o  o  o  o  d  d  d  d  oo m d  o o d  o o d  o o d  CN  o  SO  o o  d  d  CO  oo d  o o d  00  in  o  >n d  d  o p  CN T t  d  oo  CO  o  o  o  i—i  T-H  ~*  d  oo O  cN  CN  d  OO CO  oo in d  T t  d  oo o  CO CO  CN  d  d  >n  SO  CN  d  <N  T t  in d  d  oo m  so d  -<  T t  SO d  o © d  m  d  d  o o d  o  Tt  CO  Os  oo d  d  o o d  oo p  Os  Os p  m  1—1  CN  d  o o  T-H  CN  d  O  o  CO  d  T t  d  CO  o o d  d  o  d  d  oo in d  (N  CN  SO SO T^  O m  o  Tt  d  T-H  CO  d  o o  co  O  o  CN  CO  d  T t  o  CO CO  CO CO  -i  CN  CN  T t  (N  d  d  in  CN  CO CO  d  d  o o d  o o d  Tt  o  o  SO d  o d  d  o d  CO CO T-H  o in d  CN  m d  O  o in d  CN T t  oo o  d  CN  Tt  CD  4) cd ^4  T3  CJ  le CJ T3  o <u  CD  3 3j  t 3  (U OH  o  t3 CD  HO I  CD 00  c  c  cd  c/)  CD  "ob  O  DH  u cu  a  C/3 CD  u  <D  -H  CJ CD OH  T3  3  .3 3  3 o o  CD OL.  _o  3 O  CJ  'cd  3  CD  T3  "3  O  cd  3  CO  cd  OH  O CD .3. > 00 3  CD  >  S3  .S  IH CD SH  o  <N  CN O  O  o  o  o  so v i o o o o  CN O  m  O in r—  in  r-~  Os CO  oo o o d  o  rco O d  i—i  SO CO S D  d  d  d  so Os  [-o o d  o d  o  SO T—I  o d  CM O d  SO  oo O d  v CN SO  CO  SO  oo  so  d  1 DC  fl  c/i  o o  os  d V  Os  O d  T t  Os  O d  r-~ Os  o d  O d  i n CO SO  d  d  o  SO  o o d  o d  CN d  o q  o V  0\ Tt  o  d  od  SO CO CO  Os Os T-H  d  d  OS  in  SO  CN  T t  o o  o o  os so  d  co  T t  so  co  T-H  d  o d  co i n d d  - H  d  S  d  oo T-H  OO O s  VO OO  O  o  T t  d  Os  CO  oo oo CN  d  o o  W  -H  — t m o o  d  co d  d  o  »-H  o P  O d  'I  o  . X  CU  a, _©  OO  oo  9  co  CO  o o  S O CN  o d  d  o  o  d  d  V  O x"> .-i o cn O  0\  p~ oo  <0.001  to  CO  q  0.020 -0.032  a _©  0.232 -0.020  cj S  0.485  'l  g g  m  CO  Tt  d  d  (N  T-H  CO  O  d  d  m o o P d P  r- o  T t  V  *5 <o CD  ny  e o  -C  2  •SH T3  h « o  cs  4H  3  ^  r~ so CO so d  d  so rON  O  d  d  o  co <N  o  CN  P  d  9  ^  VH  1> 13 2 £ 3  co  -a  _H  T-H T-H  in  co  T t  so  d  d  r-~ so o o o o d  d  o  ^ £ P  CN  O  9  d  T3  o  3  P  -a  to  O O CU 3  o C O  rs a, § s-  O u  a  co  cu  cj cu  a co  JS  oo cd  Xi o  s  cu  0)  e  O a o u O  CJ  cu  5-  00  T3  "o  O  cd cu  " cu  PH  3 3  VH  a  J3 'o  a co c  cd  a  o  '5  00|  TJ  CO I 3  CU  o t-H  151  cn CO O ©  cn in  CN  *—i  8 ON  OO  Tt  d  cn o o  NO CN o NO o  ON O o  d  d  d  d  OO  in  o  cn CN  o  o d  ON m  o  ON NO NO o o o  d  CN CN d  CN  NO  O  d  cn in  d  d  cn o d d  NO  d  d  i  cn  cn o d  ON  © d  NO  cn d  CN d  cn o o d  ON  o o d  o o d  r-  OO  CN o o d  o o d  i  o d  CD  >  o  o  d  oo cn CN  r-  Tt  cn cn  d  o  Tt  cn  i  Tt  d  OO  cn cn © d  T-H  CN O d  O  oo  cn  cn  V  d  o d  o  OO  NO  d  d  o  r-  d  o o d  o d  d  O d  CN O o d  Tt  o o d  d  O  O d  o o d  cn  o o d  "3 c >  T3  c  3 OQ O  u  o  d  T3  CD O oj  o o d  o  ON  3 3  CD XI  U  d  cn ©  oo  Tt  ©  oo  d ON o o  © O d  © © ©  oo  NO  O ©  ©  ON  CN  V  d  d  © CN © d  o © d  cn cn © d  CN © d  Tt ©  oo  cn  in  Tt  in  ON  o  o  © d  8  Tt  ON © Tt  d  V  CN  © d  V  d  ON  d  ON NO NO o  o d  NO NO  CN d  OO  8  o d  V  oo cn CN  o o d  O cn d d  3  <D  T3  X  «> Si  8 CO  §  o ©  R  U  "2 " §  «3 > 8  O  o  X 3  xi •£>  <+H  J3  P3  8^  Q  oo  cD >  3  3  ON  Tt  XJ  in  cn d  ©  T-H  T-H  ©  o © d  d  u  <D  O. -O O O  •a  o  oo  <D >  x>  CJ  3  03  03 H  & E CD  a. o  PH  oo ©  in ON  Xj  o X 3  J2  3  CO  ©  o © ©  CD 3 X  3 CD  x> C/>  & fe 3  03  X 3 CO  3  O  3 T3  'u CD  Q  3  O  o ©  V  3  3  NO  V  H  'S3  '3  cn  O ©  CD >  o u  ©  ©  © © d  CN O d  cn  Tt  o o ©  Ja  CD X! O  x  .—•  i  NO NO  o  V  v  Tt  Tt  o o d  CD  o x: U  OO  o o  o d  H  CD  >  3  152  Figure legends Figure 5.1. Map of study area showing approximate locations of nine watersheds (a) and grid layout at each site (b). Transect lines measure 500m parallel to the river edge and 200m away from the river. Each site contained three 20m radius circular plots. Trees were measured in the 20m plots. Shrubs and forbs were visually estimated in nested plots of 10m radius subplots and l m quadrats nested in the center of the shrub plots respectively. Coarse woody debris (CWD) 2  was sampled at the point of intersection along the circumference of the plot.  Figure 5.2. Principal component analysis of 10 vegetation variables: forb cover, average shrub richness, cover 2-3m, conifer density, deciduous sapling density, downed wood, grass cover, deciduous density, moss cover and snag density. Loadings for individual variables on each axis are in Table 2. Letters above each data point refer to the buffer width and replicate.  Figure 5.3. Principal component scores 2 and 3 regressed against log buffer width (n=12) for western Vancouver. Island, 1997.  Figure 5.4. Mean and standard deviations of vegetation variables that had high loadings on principal components 2 and 3 plotted against buffers and old-growth (n = 15).  154  fc > O  U  -a o  1  *w  s ^ o  N  U  W  OG  PH CD  o CD  &. 00  s  >  P i'  •  -aO  -20  •  -1.0  1  QO  1.0  20  30  CO  Deciduous Tree Density/Tall Shrub Cover PCA2  155  c  CD C  o  CL  E o o  "03  Q.  o c  .6  .8  1.0  1.2  1.4  1.6  1.8 2.0 2.2  2.4  Log Buffer Width 2.0CM  -»—*  c c o  LO-  CD  CL  CO •  "cC Q. O  -1.0-  E o o  c  P = 0.026 -2.0" -3.0  = 0.33  R .8  1.0  1.2  1.4  1.6  1.8 2.0 2.2 2.4  Log Buffer Width  CO -4—'  c CD c o  CL  E o o  05 D_ O  c  .6  .8  1.0  1.2  1.4  1.6  1.8 2.0 2.2 2.4  Log Buffer Width  o o  LO  o o  -ST  o o  CO  o o  CM  o o T  -  (•eu/ou) Ansuaa  o  yyvNA  yyvNA  yyvN  yyvN  L O O L O O L O O L O O CO CO C\J C\J i - i -  J9AOO%  157  ddVNA  ddVNA  yavN  ddVN  Q3IAJ  Q3IAI  3QIM  3QIAA  00  (•eu/ou) Aiisuaa  o  o  o  o  o  CM  O  CO  CO  TJ-  J9AOO %  o CM  J3AO0%  o  ddVNA  ddVNA  ddVN  ddVN  aaiAi  Q3IAJ  3QIAA  3QIAA  o o o - o o o LO  Tf  o o o  CO  o o o  CM  o o o 1-  •eq/aiun|OA  o  158  Literature cited Alaback, P.B. 1982. Dynamics of understory biomass in Sitka Spruce-Western Hemlock forests of southeast Alaska. Ecology 63: 1932-1948.  Block, W . M . , Brennan, L.A., and Gutierrez, R.J. 1984. The use of guilds and guildindicator species for assessing habitat suitability. In Wildlife 2000 - Modeling habitat relationships of terrestrial vertebrates. Edited by Jared Verner, Michael L. Morrison, and C.John Ralph. The University of Wisconsin Press, California p. 455.  Burns, R . M . , and Honkala, B.H. 1990. Silvics of North America: Volume 2, Hardwoods, ed. Forest Service, United States Department of Agriculture, Washington, D.C.  Camargo, J.L.C., and Kapos, V. 1995. Complex edge effects on soil moisture and microclimate in central Amazonian forest. Journal of Tropical Ecology 11: 205-221.  Campbell, A.G., and Franklin, J.F. 1979. Riparian vegetation in Oregon's western Cascade mountains: composition, biomass, and autumn phenology. Coniferous Forest Biome, Ecosystem Analysis Studies, U.S./International Biological Program, Seattle, Washington.  Campbell, R.W., Dawe, N.K., McTaggart-Cowan, I., Cooper, J.M., Kaiser, G.W., McNall, M.C.E., and Smith, G.E.J. 1997. The birds of British Columbia. U B C Press, Vancouver.  Carey, A.B., Hardt, M . M . , Horton, S.P., and Biswell, B.L. 1991. Spring bird communities in the Oregon Coast Range. In Wildlife and vegetation of unmanaged Douglas-Fir forests. Edited by Leonard F. Ruggiero, Keith B. Aubry, Andrew B. Carey, and Mark H . Huff. U.S. Department of Agriculture, Forest Service. General Technical Report PNWGTR-285, Portland, Oregon pp. 123-142.  159  Chen, J., Franklin, J.F., and Spies, T.A. 1992. Vegetation responses to edge environments in old-growth Douglas-fir forests. Ecol. Appl. 2: 387-396.  Chen, J., Franklin, J.F., and Spies, T.A. 1995. Growing-season microclimatic gradients from clearcut edges into old-growth Douglas-fir forests. Ecol. Appl. 5: 74-86.  Comeau, P.G., Harper, G., Biring, B., and Thomas, K.D. 1996. Effects of red alder on stand dynamics and nitrogen availability (EP-1121). In Silviculture of temperate and boreal broadleaf-conifer mixtures. Edited by P.G. Comeau and K.D. Thomas. Province of British Columbia, Ministry of forests research program, Victoria, B.C. p. 79.  Darveau, M . , Beauchesne, P., Belanger, L., Huot, J., and Larue, P. 1995. Riparian forest strips as habitat for breeding birds in boreal forest. J. Wildl. Mgmt. 59: 67-78.  Dayton, P.K. 1998. Reversal of the burden of proof in fisheries management. Science 279: 821-822.  Enns, K . A . , Peterson, E.B., and McLennan, D. 1993. Impacts of hardwood management on British Columbia wildlife: problem analysis. F R D A II, Victoria, B.C.  Freemark, K.E., and Merriam, H.G. 1986. Importance of area and habitat heterogeneity to bird assemblages in temperate forest fragments. Biol. Cons. 36: 115-141.  Gilbert, F.F., and Allwine, R. 1991. Spring bird communities in the Oregon Cascade Range. In Wildlife and vegetation of unmanaged Douglas-fir forests. Edited by L.F. Ruggiero, K . B . Aubry, A.B. Carey, and M . H . Huff. U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Portland, Oregon p. 533.  160  Hagar, J.C. 1999. Influence of riparian buffer width on bird assemblages in western Oregon. J. Wildl. Mgmt. 63: 484-496.  Halpern, C.B., and Spies, T.A. 1995. Plant species diversity in natural and managed forests of the Pacific Northwest. Ecol. Appl. 5: 913-934.  Hansson, L . 1983. Bird numbers across edges between mature conifer forest and clearcuts in central Sweden. Ornis Scand. 14: 97-103.  Helle, P., and Muona, J. 1985. Invertebrate numbers in edges between clear-felling and mature forests in northern Finland. Silva Fennica 19: 281-294.  Hodges Jr., M.F., and Krementz, D.G. 1996. Neotropical migratory breedng bird communities in riparian forests of different widths along the Altamaha River, Georgia. Wilson Bull. 108: 496-506.  Holmes, R.T., and Robinson, S.K. 1981. Tree species preferences of foraging insectivorous birds in a northern hardwoods forest. Oecologia 48: 31-35.  James, F.C., and Warner, N.O. 1982. Relationships between temperate forest bird communities and vegetation structure. Ecology 63: 159-171.  Kilgo, J . C , Sargent, R.A., Chapman, B.R., and Miller, K . V . 1998. Effect of stand width and adjacent habitat on breeding bird communities in bottomland hardwoods. J. Wildl. Mgmt. 62: 72-83.  Legendre, L., and Legendre, P. 1983. Numerical ecology. Elsevier, Amsterdam.  Lock, P.A., and Naiman, R.J. 1998. Effects of stream size on bird community structure in coastal temperate forests of the Pacific Northwest, U.S.A. J. Biogeog. 25: 773-782.  161  MacArthur, R.H., and MacArthur, J.W. 1961. On bird species diversity. Ecology 42: 594598.  MacArthur, R.H., MacArthur, J.W., and Preer, J. 1962. On bird species diversity. II. Prediction of bird census from habitat measurements. Am. Nat. 96: 167-174.  Manuwal, D.A. 1986. Characteristics of bird assemblages along linear riparian zones in western Montana. Murrelet 67: 10-18.  Martin, T.E. 1981. Limitation in small habitat islands: chance or competition? Auk 98: 715-734.  Maurer, B.A., McArthur, L.B., and Whitmore, R.C. 1981. Effects of logging on guild structure of a forest bird community in West Virginia. American Birds 35: 11-13.  McGarigal, K., and McComb, W.C. 1992. Streamside versus upslope breeding bird communities in the central Oregon Coast Range. J. Wildl. Mgmt. 56: 10-23.  Messier, C. 1996. Managing light and understorey vegetation in boreal and temperate broadleaf-conifer forests. In Silviculture of temperate and boreal broadleaf-conifer mixtures. Edited by P.G. Comeau and K.D. Thomas. Province of British Columbia, Ministry of Forests Research Program, Victoria p. 79.  Minore, D., and Weatherly, H.G. 1994. Riparian trees, shrubs, and forest regeneration in the coastal mountains of Oregon. New Forestry 8: 249-263.  Newbold, J.D., Erman, D.C., and Roby, K.B. 1980. Effects of logging on macroinvertebrates in streams with and without buffer strips. Canadian Journal of Fisheries and Aquatic Science 37: 1076-1085.  162  Nilsson, S.G., Bengtsson, J., and As, S. 1988. Habitat diversity or area per se? Species richness of woody plants, carabid beetles and land snails on islands. J. Anim. Ecol. 57: 685-704.  Norusis, M.J. 1994. SPSS basic and advanced statistics. 6.1 SPSS Inc., Chicago, Illinois.  Oliver, C D . , and Larson, B.C. 1990. Forest stand dynamics. McGraw-Hill, New York.  Pabst, R.J., and Spies, T.A. 1998. Distribution of herbs and shrubs in relation to landform and canopy cover in riparian forests of coastal Oregon. Canadian Journal of Botany 76: 298-315.  Pabst, R.J., and Spies, T.A. 1999. Structure and composition of unmanage riparian forests in the coastal mountains of Oregon, U.S.A. Can. J. Forest Res. 29: 1557-1573.  Pearson, S., and Manuwal, D.A. 2001. Breeding bird response to riparian buffer width in managed Pacific Northwest Douglas Fir forests. Ecol. Appl. 11: 840-853.  Poage, N.J. 1994. Comparison of stand development of a deciduous-dominated riparian forest and a coniferous-dominated riparian forest in the Oregon Coast Range. M.S. Thesis, Oregon State University, Corvallis.  Puettmann, K.J., and Hibbs, D.E. 1996. Ecology and dynamics of mixed red alder-conifer stands. In Silviculture of temperate and boreal broadleaf-conifer mixtures. Edited by P.G. Comeau and K.D. Thomas. Province of British Columbia, Ministry of Forests Research, Victoria, B.C. pp. 82-97.  163  Ranney, J.W., Bruner, M.C., and Levenson, J.B. 1981. The importance of edge in the structure and dynamics of forest islands. In Forest island dynamics in man-dominated landscapes. Edited by R.L. Burgess and D . M . Sharpe. Springer-Verlag, New York pp. 67-96.  Rice, J . C , Ohmart, R.D., and Anderson, B.W. 1983. Habitat selection attributes of an avian community: a discriminant analysis investigation. Ecol. Monogr. 53: 263-290.  Rice, J . C , Anderson, B.W., and Ohmart, R.D. 1984. Comparison of the importance of different habitat attributes to avian community organization. J. Wildl. Mgmt. 48: 895911.  Robinson, S.K., and Holmes, R.T. 1984. Effects of plant species and foliage structure on the foraging behavior of forest birds. Auk 101: 672-684.  Rotenberry, J.R., and Wiens, J.A. 1980a. Habitat structure, patchiness, and avian communities in North American steppe vegetation: a multivariate analysis. Ecology 61: 1228-1250.  Rotenberry, J.T. 1985. The role of habitat in avian community composition: physiognomy or floristics. Oecologia 67: 213-217.  Saab, V . 1999. Importance of spatial scale to habitat use by breeding birds in riparian forests: a hierarchical analysis. Ecol. Appl. 9: 135-151.  Sakai, H.F., and Noon, B.R. 1991. Nest-site characteristics of Hammond's and PacificSlope flycatchers in northwestern California. Condor 93: 563-574.  164  Schmiegelow, F.A., Machtans, C.S., and Hannon, S.J. 1997. Are boreal birds resilient to forest fragmentation? A n experimental study of short-term community responses. Ecology 78: 1914-1932.  Short, H.L., and Burnham, K.P. 1982. Technique for structuring wildlife guilds to evaluate impacts on wildlife communities. Special Sci. Report - Wildlife 244. USDI Fish and Wildlife Service.  Smith, S.M. 1995. Distribution-free and Robust Statistical Methods: Viable Alternatives to Parametric Statistics? Ecology 76: 1997-1998.  Spurr, S.H., and Barnes, B.V. 1980. Forest ecology. John Wiley and Sons, New York.  Stauffer, D.F., and Best, L.B. 1980. Habitat selection by birds of riparian communities: evaluating effects of habitat alterations. J. Wildl. Mgmt. 44: 1-15.  Steinblums, I.J. 1977. Streamside buffer strips: survival, effectiveness, and design. Master of Science Thesis, Oregon State University, Corvallis, Oregon. 180 p.  Van Wagner, C.E. 1968. The line intersect method in forest fuel sampling. Forest Science 14: 20-26.  Verner, J. 1984. The guild concept applied to management of bird populations. Environmental Management 8: 1-14.  Wales, B.A. 1972. Vegetation analysis of north and south edges in a mature oak-hickory forest. Ecol. Monogr. 42: 451-471.  Wiens, J.A. 1989. The ecology of bird communities. Cambridge Univ. Press, Cambridge.  165  Wiens, J.A., and Rotenberry, J.T. 1981. Habitat associations and community structure of birds in shrubsteppe environments. Ecol. Monogr. 51: 2.1-41.  Wiens, J.A., Stenseth, N.C., Van Home, B., and Ims, R.A. 1993. Ecological mechanisms and landscape ecology. Oikos 66: 369-380.  Williams-Linera, G. 1990. Vegetation structure and environmental conditions of forest edges in Panama. J. Ecol. 78: 356-373.  Willson, M.F. 1974. Avian community organization and habitat structure. Ecology 55: 1017-1029.  Willson, M.F., and Comet, T.A. 1996. Bird communities of northern forests: patterns of diversity and abundance. Condor 98: 337-349.  Young, J.S. 1996. Nonlinear bird-habitat relationships in managed forests of the Swan Valley, Montana. MSc. Thesis, Montana. 105 p.  166  Chapter 6: General Discussion My objectives in this thesis were to: 1) describe the riparian and upland bird communities in undisturbed old-growth forests and to test whether riparian bird communities have greater diversity and abundance compared to upland communities, 2) test whether bird communities in buffer strips left after logging show similar fragmentation effects to those seen in other areas of the Pacific Northwest; 3) examine species' movements across habitat boundaries and test whether these differ between river and forest-clearcut edges and across a range of buffer widths, 4) test whether structural features of the vegetation vary across buffer widths; and 5) test the relative importance of area and vegetation as predictors of bird species richness and abundance of several foraging and habitat guilds and individual species.  Are Riparian Bird Communities More Diverse Than Upland Bird Communities?  Species richness and diversity did not differ over the riparian-upland gradient. This pattern was consistent over the three years. Except for riparian specialists and the Hammond's flycatcher, abundances of guilds and individual species also did not differ from riparian to upland. Two other studies (Murray and Stauffer, 1995; Pearson and Manuwal, 2001) also found no differences in bird diversity or species composition (Murray and Stauffer, 1995; Pearson and Manuwal, 2001) across the riparian-upland gradient, but McGarigal and McComb (1992) found higher diversity in upland habitats. Bird communities in my study were dominated by five or six species, with a small number of uncommon species specializing on either riparian or upland habitats. These findings disagree with earlier studies in arid and agricultural environments that found higher diversity and abundance in riparian habitats (Stevens et al., 1977; Stauffer and Best, 1980; Szaro, 1980; Tubbs, 1980, reviewed by Knopf and Samson, 1994). In those studies, there was a strong structural contrast between riparian and upland habitats (Murray and Stauffer, 1995) that is lacking in moist forests. In northwestern forests, densities of deciduous trees and cover of tall, berry-producing shrubs are strong indicators of the abundances of several species including certain riparian specialists and  167  neotropical migrants (McGarigal and McComb, 1992; Pearson and Manuwal, 2001, Shirley, Chapter 1). Deciduous trees in western forests are often associated with early successional habitats and disturbed areas such as riparian zones and landslides (Willson and Comet, 1996). M y study and the other two studies in the Pacific Northwest (McGarigal and McComb, 1992; Pearson and Manuwal, 2001) were done along streams < 50 m in width. Sites along rivers >70 m wide have more deciduous trees and support more diverse bird communities than upstream sites (Lock and Naiman, 1998).  Fragmentation Effects On Bird Communities In Buffer Strips  Differences in bird communities between riparian buffer strips and undisturbed controls were small except for lower and higher abundances of forest-interior and open-edge species respectively in buffers. Species richness in buffers and controls showed little variation. Both buffers and controls were dominated by 6 common species while 1-6 rarer species made up similar proportions of each community. Abundances between buffers and controls did not differ between migratory classes. Two foraging guilds, foliage gleaners and ground gleaners were slightly more abundant in buffers than in controls in two of the years. Forest interior species such as pacific-slope flycatcher and brown creeper were 50% more abundant in controls than in buffers and were rare or absent in narrower buffers. Conversely, open-edge species such as rufous hummingbirds and Steller's jays were 4 times more abundant in buffers than in controls and rare in wider buffers. Riparian specialists, particularly Hammond's flycatchers, were 2 times more abundant in buffers in one of the three years. Species turnover over three years was similar between buffers and controls, but varied inversely with buffer width. Consequently, species assemblages in narrow buffers were less similar to controls than in wide buffers. Other studies in buffer strips (Darveau et al., 1995; Gyug, 1995; Hagar, 1999; Whitaker and Montevecchi, 1999) have found similar reductions in abundances of forest interior species in buffers < 60-70m wide and losses of these species in buffers < 40m wide along with corresponding increases in open-edge species. The pattern is evident despite the buffers remaining attached to larger forest patches at one or both ends. These results are consistent with previous work in forest fragments. As in earlier studies,  168  area is the major factor influencing species distributions in forest fragments and isolation is relatively unimportant (reviewed in Askins et al., 1990).  Interestingly, I found that species richness and abundance of several guilds and species increased during my study. Species richness increased in buffers and adjacent clearcuts and abundances of several guilds increased in buffers, clearcuts and controls. I attributed these increases to forest regrowth adjacent to buffers and to regional changes in the supply of breeding or foraging habitat. Forest regrowth may also facilitate movements between patches for area-sensitive species (Stouffer and Bierregaard Jr., 1995; Borges and Stouffer, 1999; Gascon et al., 1999). Although the adjacent clearcuts in my study were only 3-5 years old at the start of the study, re-growth may have facilitated colonization by forest interior species and forest generalists in buffers (unpubl. data).  Buffer Width and Bird Movements Across Habitat Boundaries  Disrupted dispersal behavior may cause species loss in fragmented habitats (Stacey and Taper, 1992; Villard et al., 1993; Faaborg et al., 1995; Walters, 1998). A n important factor in successful dispersal is the ability of species to move within fragmented landscapes. The ability to move across and perhaps exploit resources in adjacent habitats may also allow a species to persist in marginal habitats until a better territory is found. To date, studies have examined movements of birds within permanently fragmented habitats (Stacey and Taper, 1992; Villard et al., 1993; Faaborg et al., 1995; Walters, 1998), but movement patterns may differ in temporarily fragmented landscapes of northwestern forests. I examined the frequency of movements across river and forestclearcut edges for several habitat guilds.  As expected, I found that forest generalists, open-edge and ubiquitous species crossed riparian and clearcut edges most often. Riparian species seldom crossed boundaries while forest interior species almost never did so. These results are not surprising, however, because riparian and forest interior species were both relatively rare and absent in narrower buffers. Although key information on movements of these species remains to  169  be gathered, other observations of focal individuals (unpubl. data), however, indicate that crossings by forest interior species are less frequent than would be expected from their densities, suggesting some edge-avoidance behavior. The number of crossings was not negatively related to buffer width as I predicted; however, the number of crossings was positively related to density in the buffers at both river and clearcut edges. A significant negative relationship between density and buffer width is, therefore, consistent with my hypothesis that narrow buffers have higher numbers of crossings because they do not provide suitable habitat to support forest-dwelling species, but rather function as foraging sites or travel corridors. As predicted, the number of crossings at forest-clearcut edges was much greater than at river edges particularly for wider buffers. This result suggests that forest species with established territories in buffers frequently make forays into the adjacent clearcuts. For some forest-dwelling species, the habitat in buffers may be enhanced by their juxtaposition to early successional forest while for others, buffers might represent sub-optimal habitat. Information on demographic parameters other than density and movements of marked individuals is needed to answer these questions.  I expected movements to increase with increases in food in adjacent clearcuts and the additional movements of juvenile birds. Interestingly, no increases in number of crossings occurred later in the breeding season when numbers must have peaked after young became independent from parental care. Perhaps food resources in buffers and clearcuts remain relatively constant during the study period. Also, increases in numbers of movements due to dispersing juveniles are likely to be offset by a reduction in adult movements once young have fledged.  Bird/Habitat Associations in Buffer Strips  Studies of the sizes of unharvested patches or buffers may overlook local habitat associations that strongly influence the suitability of a site for some forest bird species. In particular, edge effects that alter plant communities may alter the assemblages of birds in narrow buffers (Ranney et al., 1981). My study of vegetation structure in buffer strips revealed differences in densities of faster-growing deciduous trees, and cover of shrubs  170  and forbs over a range of buffer widths. These differences may be due to either variation among sites before harvesting or to the response of fast-growing plants to higher light levels after harvesting; however, given the rapid growth of red alder and black cottonwood in these forests, I am confident that differences in these plants are partly attributable to vegetation dynamics within buffers since harvest.  Although buffer width explained patterns of abundance for several species groups and species, abundances of three habitat guilds: riparian specialists, forest interior and openedge species and 4 of 8 individual species were best predicted by deciduous tree density, and by species richness and cover of shrubs. Species richness was also explained by buffer width. This result agrees with studies in homogenous habitats (Willson, 1974; Martin 1981) and in newly isolated fragments (Nilsson et al., 1988; Schmiegelow, 1997). In contrast, other studies in buffer strips (Hagar, 1999; Saab, 1999) have found that bird species richness was positively associated with the presence of large hardwoods (Lock andNaiman, 1998; Hagar, 1999). These studies (Hagar, 1999; Saab, 1999) were conducted in headwater streams where deciduous tree densities are lower than downstream (Lock and Naiman, 1998). The presence of deciduous trees may have a larger impact on species richness in those areas than it did in the larger streams that I studied.  Management Implications  M y study, and three other studies in the Pacific Northwest (McGarigal and McComb, 1992; Murray and Stauffer, 1995; Wiebe and Martin, 1998) all found only weak gradients of diversity and abundance from riparian forests to upland forests. This result contrasts with patterns observed in the southern United States (Stevens et al., 1977; Dickson, 1978; Szaro, 1980). Species restricted to riparian habitats here were either strongly associated with the rivers (e.g. American dipper and belted kingfisher) or with dominant deciduous vegetation in riparian areas (e.g. Hammond's flycatcher and warbling vireo). There was no evidence that harvesting had strong negative impacts on this group. For the few forest-interior species such as brown creepers, Pacific-slope flycatchers, pileated  171  woodpeckers and red-breasted nuthatches found primarily in wide buffers, however, harvesting in these areas is unlikely to provide suitable habitat unless buffers >100m are retained.  The majority of species in these riparian forests are forest generalists that use both riparian and upland habitats. For these species, the non-linear relationship of abundance and buffer width shows a greater loss of individuals than would be expected below a width of 125 m. I was, however, unable to find any buffers to study in the range of 50125m. Coincidentally, the width prescribed by the B.C. Forest Practices Code Riparian Guidebook (B.C. Ministry of Forests, 1995) for these rivers (70m) falls within this range. Additional surveys of existing sites and/or experimental studies are therefore warranted to fully characterize the relationship between abundance and buffer width. A n important limitation of my study is that it was restricted to the breeding season. Resident birds, which make up two-thirds of community abundance in summer, may be more restricted in their habitat use during the winter. Surveys over winter in buffers and adjacent clearcuts would provide information on distributions and habitat relationships of resident birds during a period of harsher environmental conditions and reduced food supply.  Retention of wide buffers along all rivers would remove much of the easily accessible land from the harvestable land base and is therefore unlikely to happen. The status quo prescribes buffers of 70m depending on restrictions in the management zones (Ministry of Forests, 1995). This policy can result in the retention of buffers many of which do not provide suitable habitat for forest-interior species. A better guideline for protection of vulnerable species might be to alternate buffers that are wide enough (45m) to maintain stream microclimate (Brosofske et al., 1997) with wide (>100m) buffers that can protect vulnerable forest-interior bird species. Narrow buffers of 40-50m still support some forest-dwelling species and provide habitat and connections between isolated old-growth patches (Stacey and Taper, 1992, Faaborg et al., 1995).  In addition to establishing adequate buffer widths along large streams, I offer several other management recommendations to reduce the negative impacts of forest removal.  172  First, harvesting along the larger streams in this region is usually limited to one side of the stream leaving the unlogged side as a potential source population for immigrants to buffers. Logging both sides of a stream in one pass should be discouraged at least until canopy closure in the logged site has occurred. Preliminary surveys of sites where both sides are logged indicate that reductions in species richness and abundance due to edge effects are likely to be more severe than where one side is logged (Shirley, unpubl. data).  Second, strong positive associations have been found between several bird species and large deciduous tree species and shrubs in this and other studies in the Pacific Northwest (Enns et al., 1993; Lock and Naiman, 1998; Hagar, 1999). The vegetation attributes important to birds in this study are also positively correlated with buffer width. Retention of adequate unmanaged buffers should also provide suitable vegetation structures for those species. In narrow buffers and harvested areas, these features, particularly mature deciduous trees, could be retained to enhance habitat suitability for birds.  Third, silvicultural practices such as size of cutblock can have significant effects on the local bird community (Thompson et al., 1995). Small openings are generally thought to be beneficial for area-sensitive species, but they also create more edge (Thompson, 1993). The value of small openings will depend on the strength of negative edge effects in the study area. Although there have been few studies of how birds respond to selective harvesting (Thompson et al., 1995), large negative effects on population size (Thompson, 1993) have been demonstrated in small clearcuts in midwestern U.S. forests (Brittingham and Temple, 1983, Robinson and Robinson, 1999). Much less is known about edge effects in the forested landscapes of the Pacific Northwest (Freemark et al., 1995; Walters, 1998) although they are commonly believed to be less severe (Freemark et al., 1995, Schieck et al., 1995). Most studies report densities or abundance. These are not the most reliable indicators of suitable breeding habitat (Van Home, 1983).  Finally, use of secondary forest by tropical birds enhances their ability to persist in primary forest patches (Stouffer and Bierregaard, 1995, Borges and Stouffer, 1999, Gascon et a l , 1999), but little information exists on how forest birds actually use habitat  173  mosaics. Correlations between species richness in fragments and adjacent clearcuts over time (Shirley, Chapter 2) and the fact that many forest species move across the forestclearcut edge of buffers (Shirley, Chapter 3) also imply some interaction, but the exact nature of these movements is unclear and will require data on movements of individuals. Retaining mature deciduous trees, snags and patches of conifers in harvested areas may facilitate movements between patches and/or improve habitat and should be explicitly tested.  Conclusions  The beginning of my study coincided with the introduction of the B.C. Forest Practices Code (Ministry of Forests, 1995). The 70m buffers currently mandated for streams with fish provide more habitat for forest-dwelling birds than the narrow buffers recommended in the past. Although some species are closely associated with riparian habitat, bird communities of old-growth forests in this region broadly overlap riparian and upland habitats. Wider buffers, therefore, improve habitat for many forest species. It is unclear, however, whether 70m buffers, which I could not study as they were not available at the time I began, will maintain abundances, particularly of area-sensitive species, similar to those in the >100m buffers in this study. Streams without fish require buffers only 30m wide, making them unsuitable as habitat for several forest bird species. The issue is critical because several watersheds surveyed in this study have only 50% old-growth forest remaining overall and some have less than 30% old-growth (Shirley, unpubl. data). Estimates of the residual amount of low-elevation old-growth are unavailable, but my experience suggests it is substantially lower than the overall percentage remaining. To maintain watershed and regional populations of forest bird species, management policy should incorporate wildlife requirements for taxa other than fish into forest development plans. Specifically, we need to know how abundances of individual species relate to the percentage of old forest remaining.  Current harvesting plans specify that much of the remaining 30-50% of older forests will also be taken. There are many ecological uncertainties associated with this "experiment".  174  Riparian areas worldwide are important for several functions other than breeding habitat including maintaining migratory habitats for birds and stream habitats. The value of linkages between riparian and upland habitats also have not been well studied. M y research in the very moist forests of the Pacific Northwest did not find fragmentation effects as dramatic as those found in several areas of the central and eastern U.S. (review by Freemark et al., 1995), which may be partly due to a smaller proportion of Neotropical migrants in western coniferous forests (Sauer and Droege, 1992; Willson and Comet, 1996). Forests in this region have historically been subjected to low levels of natural disturbance (Lertzman et al., 1996) and so may not be resilient to large-scale disturbances beyond some unknown threshold of fragmentation.  Much information on demographics  and movements by individuals within the landscape remains to be gathered to provide a detailed assessment of threats to forest birds of this relatively understudied system. We should, therefore, take a precautionary approach in forest land-use planning. This approach should include a review of forest policy in general. Political rhetoric about job losses in the short-term should not deter sound scientific research and implementation of policies that preserve these ecosystems and their benefits to society in perpetuity.  175  Literature cited Askins, R.A., J.F. Lynch, and R. Greenberg. 1990. Population declines in migratory birds in eastern North America. Current Ornithology 7:1-57.  B.C. Ministry of Forests, and B.C. Environment. 1995. Forest Practices Code of British Columbia: Riparian management area guidebook. Province of British Columbia, Victoria, B.C.  Borges, S.H., and P.C. Stouffer. 1999. Bird communities in two types of anthropogenic successional vegetation in Central Amazonia. Condor 101:529-536.  Brittingham, M.C., and S.A. Temple. 1983. Have cowbirds caused forest songbirds to decline? Bioscience 33:31-35.  Brosofske, K.D., J. Chen, R.J. Naiman, and J.F. Franklin. 1997. Harvesting effects on microclimatic gradients from small streams to uplands in western Washington. Ecological Applications 7:1188-1200.  Darveau, M . , P. Beauchesne, L. Belanger, J. Huot, and P. Larue. 1995. Riparian forest strips as habitat for breeding birds in boreal forest. Journal of Wildlife Management 59:67-78.  Dickson, J.G. 1978. Forest bird communities of the bottomland hardwoods. Pages 6673 in DeGraaf, R . M . , editors. Proceedings of the workshop management of southern forests for nongame birds. U S D A Forest Service General Technical Report, Atlanta, Georgia.  Enns, K.A., E.B. Peterson, and D. McLennan. 1993. Impacts of hardwood management on British Columbia wildlife: problem analysis. F R D A II, Victoria, B.C.  176  Faaborg, J., M . Brittingham, T. Donovan, and J. Blake. 1995. Habitat fragmentation in the temperate zone. Pages 357-380 in Martin, T.E., and D . M . Finch, editors. Ecology and management of Neotropical migratory birds: a synthesis and review of critical issues. Oxford University Press, New York.  Freemark, K.E., J.B. Dunning, S.J. Hejl, and J.R. Probst. 1995. A landscape ecology perspective for research, conservation, and management. Pages 381-427 in Martin, T.E., and D . M . Finch, editors. Ecology and management of Neotropical migratory birds: a synthesis and review of critical issues. Oxford University Press, New York.  Gascon, C , T.E. Lovejoy, R.O. Bierregaard Jr., J.R. Malcolm, P.C. Stouffer, H.L. Vasconcelos, W.F. Laurance, B. Zimmerman, M . Tocher, and S. Borges. 1999. Matrix habitat and species richness in tropical forest remnants. Biological Conservation 91:223229.  Gyug, L.W., and M . Vaartnou. 1995. Riparian habitats: wildlife, vegetation and timber-harvest. Research Update. The silviculture systems wildlife habitat research and extension program, B.C. Ministry of Environment and Ministry of Forests, Victoria, B.C.  Hagar, J.C. 1999. Influence of riparian buffer width on bird assemblages in western Oregon. Journal of Wildlife Management 63:484-496.  Knopf, F.L., and F.B. Samson. 1994. Scale perspective on avian diversity in western riparian ecosystems. Conservation Biology 8:669-676.  Lertzman, K.P., G.D. Sutherland, A. Inselberg, and S.C. Saunders. 1996. Canopy gaps and the landscape mosaic in a coastal temperate rain forest. Ecology 77:1254-1270.  Lock, P.A., and R.J. Naiman. 1998. Effects of stream size on bird community structure in coastal temperate forests of the Pacific Northwest, U.S.A. Journal of Biogeography 25:773-782.  177  Martin, T.E. 1981. Limitation in small habitat islands: chance or competition? Auk 98:715-734.  McGarigal, K., and W.C. McComb. 1992. Streamside versus upslope breeding bird communities in the central Oregon Coast Range. Journal of Wildlife Management 56:1023.  Murray, N.L., and D.F. Stauffer. 1995. Nongame bird use of habitat in central Appalachian riparian forests. Journal of Wildlife Management 59:78-88.  Nilsson, S.G., J. Bengtsson, and S. As. 1988. Habitat diversi area per se? Species richness of woody plants, carabid beetles and land snails on islands. Journal of Animal Ecology 57:685-704.  Pearson, S., and D.A. Manuwal. 2001. Breeding bird response to riparian buffer width in managed Pacific Northwest Douglas Fir forests. Ecological Applications 11:840-853.  Ranney, J.W., M.C. Bruner, and J.B. Levenson. 1981. The importance of edge in the structure and dynamics of forest islands. Pages 67-96 in Burgess, R.L., and D . M . Sharpe, editors. Forest island dynamics in man-dominated landscapes. Springer-Verlag, New York.  Robinson, W.D., and S.K. Robinson. 1999. Effects of selective logging on forest bird populations in a fragmented landscape. Conservation Biology 13:58-66.  Saab, V . 1999. Importance of spatial scale to habitat use by breeding birds in riparian forests: a hierarchical analysis. Ecological Applications 9:135-151.  178  Sauer, J.R., and S. Droege. 1992. Geographical patterns in population trends of Neotropical migrants in North America. Pages 26-42 in Hagan III, J.M., and D.W. Johnston, editors. Ecology and conservation of Neotropical migrant landbirds. Smithsonian Institution Press, Washington.  Schieck, J., K. Lertzman, B. Nyberg, and R. Page. 1995. Effects of patch size on birds in old-growth montane forests. Conservation Biology 9:1072-1084.  Schmiegelow, F.A., C S . Machtans, and S.J. Hannon. 1997. Are boreal birds resilient to forest fragmentation? A n experimental study of short-term community responses. Ecology 78:1914-1932.  Stacey, P.B., and M . Taper. 1992. Environmental variation and the persistence of small populations. Ecological Applications 2:18-29.  Stauffer, D.F., and L . B . Best. 1980. Habitat selection by birds of riparian communities: evaluating effects of habitat alterations. Journal of Wildlife Management 44:1-15.  Stevens, L.E., B.T. Brown, J.M. Simpson, and R.R. Johnson. 1977. The importance of riparian habitat to migrating birds. Pages 156-164 in Johnson, R.R., and D.A. Jones, editors. Importance, preservation and management of riparian habitat: a symposium. U S D A Forest Service, Tucson, Arizona.  Stouffer, P . C , and R.O. Bierregaard Jr. 1995. Use of Amazon forest fragments by understory insectivorous birds. Ecology 76:2429-2445.  Szaro, R . C 1980. Factors influencing bird populations in southwestern riparian forests. Pages 403-418 in DeGraff, R . M . , editors. Workshop management of western forests and grasslands for nongame birds. U S D A For. Ser. Gen. Tech. Rep. INT-86.  179  Thompson III, F.R. 1993. Simulated responses of a forest-interior bird population to forest management options in central hardwood forests of the United States. Conservation Biology 7:325-333.  Thompson III, F.R., J.R. Probst, and M . G . Raphael. 1995. Impacts of silviculture: Overview and management recommendations. Pages 201-219 in Martin, T.E., and D . M . Finch, editors. Ecology and management of Neotropical migratory birds: a synthesis and review of critical issues. Oxford University Press, New York.  Tubbs, A . A . 1980. Riparian bird communities of the Great Plains. Pages 419-433 in DeGraaf, R . M . , editors. Workshop proceedings: management of western forests and grasslands for nongame birds. U S D A Forest Service General Technical Report, Odgen, Utah.  Van Home, B. 1983. Density as a misleading indicator of habitat quality. Journal of Wildlife Management 47:893-901.  Villard, M . A . , P.R. Martin, and C.G. Drummond. 1993. Habitat fragmentation and pairing success in the Ovenbird (Seiurus Aurocapillus). Auk 110:759-768.  Walters, J.R. 1998. The ecological basis of avian sensitivity to habitat fragmentation. Pages 537 in Marzluff, J.M., and R. Sallabanks, editors. Avian conservation: research and management. Island Press, Washington, D.C.  Whitaker, D . M . , and W.A. Montevecchi. 1999. Breeding bird assemblages inhabiting riparian buffer strips in Newfoundland, Canada. Journal of Wildlife Management 63:167179.  Willson, M.F. 1974. Avian community organization and habitat structure. Ecology 55:1017-1029.  180  Willson, M.F., and T . A . Comet. 1996. Bird communities of northern forests: ecological correlates of diversity and abundance in the understory. Condor 98:350-362.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0090748/manifest

Comment

Related Items