A N E X P E R I M E N T A L S T U D Y O F T H E P L A N T - A R T H R O P O D - B I R D F O O D C H A I N IN T H E S O U T H W E S T E R N Y U K O N by N I C H O L A S F R A N C I S G O R I N G F O L K A R D B . S c , The University of York (England), 1986 A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E in T H E F A C U L T Y O F G R A D U A T E S T U D I E S D E P A R T M E N T O F Z O O L O G Y We accept this thesis as conforming to the required standard T H E U N T V E R S I T Y O F BRITISH C O L U M B I A September 1990 (c) Nicholas F . G. Folkard, 1990 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of " Z ^ & v - O Q - M The University of British Columbia Vancouver, Canada DE-6 (2/88) ii ABSTRACT I describe an experimental study of the importance of food limitation and predation at three trophic levels in a terrestrial food web. The study system was the herb layer vegetation - arthropod - insectivorous bird food chain in the boreal forest near Kluane Lake, southwestern Yukon. Since little is known about boreal bird communities, I conducted a descriptive study of the community of passerine and piciform birds at Kluane in addition to the main study. Variable circular plot point counts were used to estimate bird populations in 1987 through 1990. Species' habitat preferences, use of foraging substrates and diets were studied in 1988 and 1989. Population densities, species richness and evenness were all low. Yellow-rumped warblers (Dendroica coronata) and dark-eyed juncos (Junco hyemalis) dominated the community. Common species differed markedly in their habitat preferences, and showed generally low overlaps in their use of foraging substrates. There was little evidence of dietary specialization. There was rather little spatial variation in the community, and species composition and total density remained approximately the same through time. However, there were large fluctuations in some species' populations between 1987 and 1989. The experimental study was conducted at two scales. Chemical fertilizer was applied to two 570m x 570m areas in 1987, 1988 and 1989. I compared arthropod populations, bird populations and bird reproductive performance in these areas with those in two control areas. Two experiments using 5m x 5m plots were performed in 1988 to examine the effects of fertilization on plants and arthropods in more detail, and to study the responses of these trophic levels to the exclusion of passerine birds and mammalian herbivores. All three trophic levels responded positively to iii fertilization, but the results were variable and there were no very large increases in biomass or population size. Dark-eyed juncos nested one week earlier in fertilized areas, which may have enhanced their reproductive success. Passerine exclusion did not increase arthropod biomass, but exclusion of mammalian herbivores increased plant biomass. "Bottom-up" limitation by food appears to dominate this system, but "top-down" limitation also operates at at least one level. More work is needed to fully understand how the system functions. iv CONTENTS ABSTRACT ii LIST OF TABLES v LIST OF FIGURES vi ACKNOWLEDGEMENTS viii CHAPTER 1: GENERAL INTRODUCTION 1 The Collaborative Special Project 2 Study Area 3 CHAPTER 2: COMMUNITY DESCRIPTION 6 Introduction 6 Methods .7 Results 13 Discussion 34 Conclusion 38 CHAPTER 3: EFFECTS OF FERTILIZATION .40 Introduction 40 Methods 41 Results 47 Discussion 78 Conclusion 85 REFERENCES 88 APPENDIX 1: BIRD SPECIES LIST 98 APPENDLX 2: GRID HABITAT CHARACTERISTICS 100 APPENDIX 3: NESTLING JUNCO DIETS 104 LIST OF TABLES TABLE I. Dates of point counts 11 TABLE II. Habitat characteristics measured by M. Nams 12 TABLE III. Species detected fewer than ten times on a grid in one or more years 24-25 TABLE IV. 11 species' preferences among the habitat characteristics shown in Table II 29-30 TABLE V. Percentage of time species spent foraging on different substrates 31 TABLE VI. Pianka's index of niche overlap for the species shown in Table V 32 TABLE VII. Food types seen to be consumed by 11 species 33 TABLE VIII. Measures of reproductive performance of dark-eyed juncos on fertilized and unfertilized grids . . . . 69 vi LIST OF FIGURES FIGURE 1. Map of the study area 5 FIGURE 2. Estimated population densities for 1987-1990 16 FIGURE 3. Numbers of detections of seven common species on the two unfertilized grids . . . , . . . ,18-19 FIGURE 4. Estimated species abundances for each grid in 1988, 1989 and 1990 21-23 FIGURE 5. Biomass indices for flying insects on large grids in 1988 . . .49 FIGURE 6. Biomass indices for ground arthropods on large grids in (a) May and (b) July / August 1988 51-52 FIGURE 7.Biomass indices for arthropods on spruce trees on large grids in September 1988 53 FIGURE 8. Proportion of willow shoots damaged by arthropods on large grids 54 FIGURE 9. Biomass indices for ground arthropods on large grids in 1989 55 FIGURE 10. Biomass indices for arthropods on willows on large grids in 1989 56 FIGURE 11. Biomass indices for arthropods on spruce trees on large grids in 1989 57 FIGURE 12. Numbers of detections of seven common bird species on fertilized and unfertilized grids 61-67 FIGURE 13. Numbers of detections of all bird species combined on fertilized and unfertilized grids 68 vii FIGURE 14. Effects of small plot experiment 1 treatments on the level of arthropod damage to birch leaves 72 FIGURE 15. Effects of small plot experiment 2 treatments on the total wet weight of herb layer vegetation on Microwave plots . . 73 FIGURE 16. Effects of small plot experiment 2 treatments on the total dry weight of plant species other than Festuca, Arctostaphylos and Epilobium on Microwave plots 74 FIGURE 17. Effects of small plot experiment 2 treatments on the biomass of flying Hemiptera on Beaver Pond plots 75 FIGURE 18. Effects of small plot experiment 2 treatments on the biomass of flying Hymenoptera on Microwave plots 76 FIGURE 19. Effects of small plot experiment 2 treatments on the biomass of all flying insects on Microwave plots 77 FIGURE 20. Summary of results of fertilization and exclusion experiments 87 viii ACKNOWLEDGEMENTS I am very grateful to Jamie Smith, my supervisor, for help with all aspects of this thesis. I have learned a great deal about research and writing from him. Judy Myers and Bill Neill gave much good advice, and made many helpful suggestions. Many others in the Ecology Group at U.B.C made useful comments, especially Wes Hochachka, Dave Huggard, Dick Repasky and Tony Sinclair. I thank Dan Doak and Dave Huggard for permission to use their data. I used the accommodation and facilities at the Arctic Institute of North America's Kluane Lake Research Station; many thanks to Andy Williams, the station manager, and Jan Williams and Carole Williams. I am grateful to all at Kluane, particularly Art Blundell, for help with fieldwork, valuable discussions, and entertainment. My work was supported by the National Science and Engineering Research Council through grants to Jamie Smith and the Collaborative Special Project, and by a Commonwealth Scholarship to myself. Finally, thankyou to Suzanne. 1 CHAPTER 1: GENERAL INTRODUCTION One of the most important goals of community ecology is to understand the processes which generate observed patterns in the structure and functioning of communities. Some communities may be organized primarily by environmental factors, such as weather (Andrewartha and Birch 1954). However, most are probably organized mainly by biotic interactions among their component species. Predation and competition for food are perhaps the most important of these biotic interactions. The aim of this study was to examine the importance of these two processes at three trophic levels in a terrestrial food chain: the plant-arthropod-bird chain in the boreal forest near KLuane Lake, southwestern Yukon. There are two extreme views of community organization. The "bottom-up" or "donor control" view is that all trophic levels are limited by food, and that consumers have no significant impact on populations at lower trophic levels (White 1978). The alternative "top-down" view is that predation limits all trophic levels, and that competition for food is unimportant (e.g. Simenstad et al. 1978, Paine 1980, Carpenter et al. 1985, Hanazaito and Yasuno 1989). Between these extremes lies a range of possibilities. All trophic levels may be limited by both predation and food: levels may be tightly coupled by Lotka-Volterra interactions, so that both factors are important throughout webs; alternatively, food limitation may be most important at low trophic levels, and predation at high levels (McQueen et al. 1986). Trophic levels may alternate between predator limitation and food limitation (Hairston et al. 1960); whether plants are limited by the former or the latter may depend on the productivity of the ecosystem (Oksanen et al. 1981, Oksanen 1988). 2 I attempted to determine which of these possibilities applied to the boreal forest community by performing experiments at two spatial scales to examine the effects of fertilization and consumer exclusion on herb layer plants, arthropods and insectivorous birds. I report the results in Chapter 3. In addition, I conducted a detailed descriptive study of the community of insectivorous birds in the study area. Little is known about the bird community near Kluane Lake or anywhere else in the boreal forest. This study was therefore a valuable supplement to the main experiments. I report the results in Chapter 2. THE COLLABORATIVE SPECIAL PROJECT This study was initiated as part of a collaborative special project (C.S.P.) funded by the National Science and Engineering Research Council. The C.S.P. is a long-term, large-scale attempt to analyze the dynamics of the vertebrate community at Kluane (Krebs et al. 1988). My study was a part of this analysis. The components I studied are linked to the rest of the community in several ways. Herb layer plants provide food for mammalian herbivores, including the snowshoe hare, which is believed to be the "keystone species" (Paine 1966) in the community. Arthropods are the numerically dominant herbivores, and may thus compete with mammalian herbivores (Brown et al. 1986, Neuvonen 1988). Small birds are preyed upon by raptors, especially the sharp-shinned hawk (Accipiter striatus). Their eggs and nestlings are eaten by small mammals. 3 STUDYAREA The study was conducted in the Shakwak Trench, a broad valley running east from the south end of Kluane Lake in the southwest Yukon (see Figure 1). The valley sides are formed by the rugged Kluane Mountains to the south and the edge of the rolling Yukon Plateau to the north. Bostock (1948, 1952) describes the physiography of the area in detail. The floor of the Shakwak Trench lies at about 780m above sea level and is covered by boreal forest. White spruce (Picea glauca) is the dominant tree species, but there are also stands of aspen (Populus tremuloides) and balsam poplar (P. balsamifera). Willow, especially Salix glauca, and birch (Betula glandulosa) are the commonest shrubs. The herb layer contains many species of grasses, forbs and creeping shrubs. The forest is very heterogeneous, due mainly to fire. Douglas (1974) describes the many vegetation types of southwestern Yukon. The climate is continental sub-Arctic, with mean daily temperatures of 11.8 degrees C in July and -25.6 degrees C in January. The Shakwak Trench receives only about 220mm of precipitation per year because it lies in the rainshadow of the St. Elias Mountains (Theberge 1980). My study sites were all located on or close to ten 570m x 570m experimental grids marked out on the valley floor by the C.S.P.. These grids were divided into 20 equally-spaced rows, labelled A to T. Each row was marked by stakes at 30m intervals. Stakes were labelled 1 to 20. Most of my work was done on four grids: Flint and Grizzly, which were fertilized in 1987, 1988 and 1989, and Silver and Sulphur, which were untreated control grids. 4 FIGURE 1. Map of the study area. Filled squares show the locations of grids used in the study. Kluane Mountains 138°W 6 CHAPTER 2: COMMUNITY DESCRIPTION INTRODUCTION Accurate community descriptions are essential if we are to understand and conserve ecosystems. Field studies of populations or even individuals are unlikely to succeed if little is known of the communities in which they are conducted, and laboratory and modelling studies are largely pointless if there are no "real" data available for comparison. Similarly, both the conservation of endangered species and the assessment and mitigation of large-scale, long-term environmental changes require detailed knowledge of the communities involved. Birds are among the most intensively studied of all organisms. However, little work has been done on the birds of the boreal forest, an immense biome which covers 30% of the land area of Canada and circles the Earth in a band approximately 1000 km wide. There have been recent studies of boreal bird communities in Fennoscandia (e.g. Solonen 1986, Monkkonen 1990), but very little work has been done in Canada since Erskine's 1977 report. The boreal forest is very different from most temperate and tropical biomes: it is colder; it has longer winters with shorter days, and shorter summers with longer days; it has fewer species of plants and animals; and it has a simpler physical structure (Erskine 1977, Larsen 1980, Van Cleve et al. 1986, Bonan and Shugart 1989, Danks and Foottit 1989, Walter and Breckle 1989). Boreal communities are therefore likely to differ in structure and function from those further south. The most important lesson learned from the intense debate over community structure in the early 1980s (Strong et al. 1984) was that there is no single process behind all community patterns and that a given community can only be understood through direct study. 7 The first aim of this study was to describe the community of passerine and piriform birds close to Kluane Lake as completely as possible. Several annotated lists of the birds of southwestern Yukon have been published (Hoefs 1973 and references therein). However, these are largely compilations of the observations of birdwatchers or other naturalists, and contain little quantitative or biological information. Theberge (1976) assessed population densities and habitat relations in the Kluane Mountains, but his study lasted less than five weeks and covered a very large area. Erskine (1977) draws general conclusions from studies made using a wide variety of techniques in many, mainly southern Canadian locations. I report the results of an intensive four-year quantitative study of bird species' abundances, habitat preferences, uses of foraging substrates and diets. METHODS Species abundances Variable circular-plot point counts (Reynolds et al. 1980) were used to estimate bird population densities in 1987 through 1990. These counts were conducted by Dave Huggard in 1987, by the author in 1988 and 1989 and by Jamie Smith in 1990. Counts were done on eight grids in 1987 and Flint, Grizzly, Silver and Sulphur grids in other years. Four sets of three ten-minute counts were conducted on each grid between May 18th and June 8th 1987. In other years, sets of 11 five minute counts were conducted on each grid on the dates shown in Table I. All counts were conducted in the period between 90 minutes before sunrise and 45 minutes after sunrise, and points were visited in reverse order on alternate counts to reduce biases associated with the time of day. The species and 8 estimated distance from the observer to every bird heard or seen was noted. Positions of birds and differences among songs were used to ensure as far as possible that individual birds were recorded only once per count. I analysed the data using Reynolds et al.'s (1980) modification of Emlen's (1971) method. Densities were estimated only for species detected ten times or more in a given year, since estimates from few detections are unlikely to be accurate (Tomialojc and Verner 1990). The number of detections of each species was summed for each lOm-wide distance band up to 90m from the observer and for the bands 90m-129m, 130m-179m, 180m-249m and 250m-349m. I converted these numbers to densities, and constructed bar charts of density against distance band. From these probability density functions (PDFs) (Burnham et al. 1980), the distance beyond which detectability decreased sharply (d) was estimated. I used Reynolds et al.'s objective method to estimate d, unless it gave an obviously incorrect value (for example, when one or two birds were detected very close to the observer; in these cases, Reynolds et al's method often gave a very low value for d, and I estimated d by visual inspection of the PDF instead). I divided the number of detections inside this distance by the area of a circle with radius d to obtain a pooled density estimate, which I converted to an estimate of true density by dividing by the total number of point counts. Niche characteristics Species' use of resources was examined at three levels: general habitat preference, detailed substrate use and diet. The data from all grids were pooled for all analyses. 9 General habitat preference I noted the grid location of every bird encountered during the 1988 and 1989 behavioural observations (see below) and the 1989 and 1990 point counts. Magi Nams measured the ten habitat variables listed in Table II at every stake on every grid in 1987 and 1988. This enabled me to determine the characteristics of the habitat at each recorded bird location. I used chi-squared goodness-of-fit tests to compare the habitat utilization of the 11 most common species with habitat availability (Neu, Byers and Peek 1974). Rare categories were pooled where appropriate, and tests were performed only when all "expected" values were 5 or more (Bailey 1981) (79 tests out of a possible 110). This condition is conservative (Roscoe and Byars 1971) but was used to reduce Type I errors. Detailed substrate use I spent approximately 100 hours walking over Flint, Grizzly, Silver and Sulphur grids in 1988 and 1989, recording the amount of time every bird encountered spent foraging on each of 12 substrates. Each bird was watched for five minutes or until it moved out of sight. I calculated Levins' (1968) index of niche width for each species and Pianka's (1973) index of niche overlap for six common species. Diet I classified food items taken by birds during behavioural observations into one of 14 categories. In addition, the diets of dark-eyed junco (Junco hyemalis) nestlings were studied in more detail by placing pipecleaner ligatures around the their necks to prevent them from swallowing (Johnson, Best and Heagy 1980). Only nestlings between five and eight days of age 10 were used, since younger birds were too small to ligature safely and older birds left the nest when disturbed. I watched ligatured nestlings continuously from a blind, and food items were retrieved using tweezers as soon as the adult had left. Ligatures were removed after two hours, and no nestlings were used more than twice. 11 TABLE I. Dates of point counts GRID 1988 1989 1990 Flint 23 May 28 May 10 June 11 July 5 May 12 May 24 May 7 June 4 July 26 May 1 June Grizzly 24 May 30 May 20 June 18 July 11 May 26 May 21 June 7 July 28 May 4 June Silver 26 May 31 May 21 June 6 May 29 May 8 June 5. July 27 May 30 May Sulphur 25 May 1 June 15 June 8 May 25 May 16 June 10 July 29 May 2 June 12 TABLE II. Habitat characteristics measured by M. Nams Variable Tree cover Dominant tree species Stand age Shrub cover Dominant shrub species Moss cover Deadfall cover Snags Aspect Slope Classification 0 (absent) to 4 (closed) S (spruce) or A (aspen) M (mature) or I (immature) 0 (absent) to 2 (closed) W (willow), B (birch), or X (other) 0 (absent) or 1 (present) 0 (absent) or 1 (present) 0 (absent) or 1 (present) Secondary compass direction or 0 0 (negligible) to 2 (steep) 13 RESULTS Appendix 1 lists the common names, Latin names and American Ornithologists' Union four-letter codes of all passerine and piriform species observed in the study area between 1987 and 1990 (excluding the common raven (Corvus corax), a large predator and scavenger that was not included in the study). Species abundances Figure 2 shows species abundances in each year, pooled for all grids. It can only be used to make detailed comparisons between 1988 and 1989, because much of the variation among other years is probably due to observers varying in their distance estimation (Scott et al. 1981). This is suggested by the results from the 1988 and 1989 counts, which I conducted; these were very similar, while other counts differed markedly. Despite this problem, there was a consistent pattern in relative abundances of species within years: the seven commonest species were the yellow-rumped warbler, dark-eyed junco, boreal chickadee, Swainson's thrush, grey jay, American robin and ruby-crowned kinglet, and they occurred in approximately the same rank order of density each year. The results from 1987 also conform to this pattern, because the high densities of white-crowned sparrows, Wilson's warblers and blackpoll warblers stem from the different grids sampled in this year. These three species prefer open habitat (the blackpoll warbler inhabits the ecotone between forest and tundra (Erskine 1977)), which was not common on the four grids censused in 1988-1990. Even if these species are included in the analysis, Kendall's coefficient of concordance among rank abundances for all four years is 0.738, and the probability that the similarity among rankings is due to chance is 0.001 (Friedman test). 14 To allow meaningful comparisons among all years, I calculated the mean number of detections per point count for the seven commonest species on the two unfertilized grids. This method avoids the problem of observer variability in estimating distances, but makes the reasonable assumptions that all observers have equally good hearing and are equally competent at identifying songs. It does not allow comparisons among species' densities to be made, since species differ in detectability. Figure 3 shows that: (a) numbers of yellow-rumped warblers increased sharply from 1989 to 1990; (b) numbers of dark-eyed juncos decreased sharply over the same period; and (c) numbers of robins decreased moderately from 1988 to 1989. Changes in other species were less marked. The pooled results conceal variation among grids, apparent in Figure 4, which shows species abundances for each grid in each year, and Table III, which lists species detected fewer than ten times in a year. This variation among grids was probably due largely to species' differing habitat preferences, which are described in the following section. Flint was dominated by immature spruce, but it was also the only grid with a large number of aspens. It had relatively little willow, moss or deadfall and few snags. Sulphur was quite similar, but had denser tree cover and more open shrub cover dominated by willow. Grizzly was dominated by mature spruce, and had many snags, dense shrub cover (mainly willow) and much moss and deadfall. Silver was similar, but had almost no immature spruce and even more snags, willow, moss and deadfall. Appendix 2 shows the detailed habitat characteristics of the four grids. Differences between fertilized and unfertilized grids are discussed in Chapter 3. Overall, however, the similarities among grids are more notable than the differences, which were most marked among the rarer species. The seven common species mentioned 15 FIGURE 2. Estimated population densities for 1987-1990. Only species detected more than ten times in a given year are shown. Much of the variation among years is probably due to observers varying in their distance estimation. Counts were conducted on eight grids in 1987 and four in other years. YRWA = yellow-rumped warbler, DEJU = dark-eyed junco, BOCH = boreal chickadee, SWTH = Swainson's thrush, GRJA = grey jay, AMRO = American robin, RCKI = ruby-crowned kinglet, WCSP = white-crowned sparrow, VATH = varied thrush, WIWA = Wilson's warbler, BPWA = blackpoll warbler. Males per square km Males per sauare km o o o o o o o o o o o o o o o o o o J i i i i i i l i ' i i 1 i ' i Males per square km Males per square km J i ' ' ' 91 17 FIGURE 3. Numbers of detections of seven common species on the two unfertilized grids. Error bars show one standard error on either side of the mean. Abbreviations are explained in the caption to Figure 2. Results are plotted on two panels to allow error bars to be shown. 20 FIGURE 4. Estimated species abundances for each grid in 1988, 1989 and 1990. Only species detected more than ten times in a given year are shown. Abbreviations are explained in the caption to Figure 2. The author conducted three to five counts per grid in 1988 and 1989; Jamie Smith conducted two counts per grid in 1990. Males per sauare km Males per square km Males per square km Males per square km co o 33 II o _ •o o o o o __ N N < 0 0 0 0 4-o IS Males per sauare km ro o o C O o CO o o o o o 1 1 I I I -- GO -r~ VE - V 1 - oo -- CO _ I I 1 Mates per square km Males per square km Males per square km ZZ Males per square km Males per sauare km o o CD o CD o o o o — o 33 I ffi <_ c CO —* I s 33 O 3) O s ci CO TJ 1 '1 - i i i - I — 1 CO -< m JJ ~ - C D " - O _ \ i i r o O o CD O CO o o o r o O o 33 > s c_ C CO > > _: 33 o 33 o _ CO TJ 1 I i I "1 SB m ~ n -r~ _ _ • - —i _ C O -O ' 1 1 1 Males per square km r o 4 -o o CO o co o ro 4 -o o o o 33 > s c X CO $ —I X CD 33 > 33 o _ CO TJ 1 1 1 1 1 1 C O -c: 1 r~ -x -) cz X ) " - C O • - o _ . . 1 1 . . I I I Males per square km r o o o 05 o CO o o o r o o — o 33 > s c CO § —I X o 33 > > _ 33 o 33 O CO TJ 1 1 I 1 1 1 o -JJ 1 N ' - N . i — - i • < . C O O " . 1 1 1 S3 24 TABLE III. Species detected fewer than ten times on a grid in one or more years. Abbreviations are explained in Appendix 1. Numbers are the mean number of detections per 11 point counts. M = species detected more than ten times. Species Flint Grizzly Silver Sulphur 88 89 90 88 89 90 88 89 90 88 89 90 SWTH M M M 2.3 M M M M M M M M AMRO M M M 1.0 0.8 2.5 M 2.3 0.5 M M M BOCH M M 3.0 M M 2.5 M 2.0 1.3 M 1.5 GRJA 1.8 M 4.5 2.7 1.3 2.0 M 1.8 1.5 M 1.5 M RCKI 0.4 M M M 0.3 1.3 1.0 3.0 WCSP 0.5 M 0.5 0.3 1.5 1.0 1.3 1.0 2.0 VATH 0.2 0.5 0.3 M 0.3 0.5 M M NOFL 1.0 0.6 2.5 3.5 0.3 1.0 1.7 2.0 TTWO 0.3 0.5 1.3 0.3 0.5 0.7 0.5 0.3 WIWA 0.5 0.3 1.3 1.8 2.0 CHSP 0.8 0.7 1.3 1.0 0.3 SASP 0.3 0.7 1.5 1.0 COYE 0.3 0.5 0.5 OSFL 1.0 0.3 WWCR 0.3 0.3 0.3 RUBL 2.3 1.0 PISI 1.0 0.3 PIGR 0.3 0.7 25 TABLE III continued. Species BPWA YEWA RWBL BCCH WWPE RBNU TRSP Flint 88 89 90 0.3 Grizzly 88 89 90 0.5 1.3 Silver 88 89 90 0.3 Sulphur 88 89 90 1.0 0.5 0.8 26 above accounted for almost all of the measurable densities, and again occurred in a fairly consistent order of abundance. Niche characteristics General habitat preference Table TV summarizes the habitat preferences of the eleven commonest bird species. Since I performed 79 goodness-of-fit tests, only the two preferences significant at p < 0.00063 (0.05/79, the Bonferroni-adjusted acceptance level) are strictly statistically significant. However, many of the other preferences may be ecologically significant, particularly those with p < 0.01.1 therefore show all preferences significant at p < 0.05 in Table TV. Many of the listed preferences agree well with what is already known about the biology of these species. For example, the yellow-rumped warbler, which showed no preferences, is "one of the most generalized and opportunistic of all our insectivorous birds" (Ehrlich et al. 1988); all of the junco's preferences are consistent with specialization in open habitats; and the three-toed woodpecker, which feeds mainly on wood-boring beetles (Ehrlich et al. 1988), preferred mature stands with much deadfall and many snags. These preferences may also explain much of the variation among grids described above. However, there are several unexpected preferences: the grey jay, which is an extremely generalized and opportunistic species, showed several significant preferences; the northern flicker preferred areas with no deadfall, and showed no preference for snags; and the varied thrush, which is usually found in dense, mature stands, showed no preferences. The unexpected preferences and non-preferences of the latter two species are probably due to 27 the small number of observations made on them. There is no obvious explanation for the grey jay's preferences. Detailed substrate use Table V shows the proportion of time that seven species spent on each substrate, and Levins' niche width indices calculated from these proportions. This subset of species is not the same as that shown in Figure 3: the ruby-crowned kinglet and Swainson's thrush are replaced by the northern flicker and red-breasted nuthatch, which were observed foraging more frequently, despite their lower abundances. The yellow-rumped warbler had the broadest niche, further emphasizing its generalized ecology. The boreal chickadee also had a broad niche. In contrast, specialized species, such as the dark-eyed junco (which spent over 75% of its time on the ground) and the red-breasted nuthatch, had very narrow niches. Pianka's indices of niche overlap for the same seven species are shown in Table VI. A value of one would indicate that two species used exactly the same substrates in exactly the same proportions; two species that used mutually exclusive sets of substrates would have an overlap index of zero. Most of the overlaps are low: 16 (76%) are lower than 0.5, and five (24%) lower than 0.1. There are overlaps greater than 0.5 between the robin on the one hand and the boreal chickadee, dark-eyed junco and red-breasted nuthatch on the other; between the chickadee and the nuthatch; and between the grey jay and the northern flicker. 28 Diet Table VII is a qualitative summary of the food types seen to be consumed by eleven species. Appendix 3 lists the food items retrieved from ligatured junco nestlings. I was not able to calculate indices of diet width or overlap, since I was not able to obtain accurate quantitative results. A few preferences are obvious: for example, boreal chickadees consumed many spiders and spider egg-cases, and yellow-rumped warblers were the only species seen to take insects in flight. However, diets were almost certainly broader than indicated here. Only four of the nine food types known to be taken by juncos were recorded in the field, and most species were observed less often than juncos. It is likely that insectivorous species in general consumed a wide variety of arthropods and did not select strongly on the basis of taxon or size. 29 TABLE TV. 11 species' preferences among the habitat characteristics shown in Table II. P values are from chi-squared goodness-of-fit tests comparing use to occurrence. Letters in parentheses show the statistical power of these tests: H = high, M = moderate, L = low. Species abbreviations are explained in Appendix 1. SPECIES PREFERRED CHARACTERISTICS P VALUE N AMRO None 41 (M) BOCH Absent and scattered tree cover 0.046 99(H) Shrubs other than willow 0.006 DEJU Absent, scattered and open tree cover <0.0005 336 (H) Aspen 0.018 Immature stands 0.004 Shrub s other than willow 0.001 GRJA Immature stands 0.006 50 (M) Shrubs other than willow < 0.0005 Southwest-facing slopes 0.018 NOFL Immature stands No moss No deadfall 0.048 0.002 0.013 18 (L) 30 TABLE TV continued. SPECIES PREFERRED CHARACTERISTICS P VALUE N RCKI Mature stands Moss Snags 0.016 0.009 0.018 22 (L) SWTH Closed tree cover Immature stands 0.048 0.030 126 (H) TTWO Mature stands Deadfall Snags 0.014 0.005 0.026 12 (L) VATH None 12 (L) WIWA Mature stands Closed shrub cover Moss Snags 0.014 0.014 0.001 0.014 13 (L) YRWA None 433 (H) 31 TABLE V. Percentage of time species spent foraging on different substrates. I < 5 = trunk or inner half of crown of spruce less than 5m above the ground, O < 15 = outer half of crown less than 15m above the ground, etc.. WIDTH = Levins' index of niche width, TIME = total observation time (seconds). Only species observed foraging on at least five separate occasions are included. Species abbreviations are explained in Appendix 1. SPECIES AMRO BOCH DEJU GRJA NOFL RBNU YRWA GROUND 43.8 5.0 75.2 16.9 5.5 12.9 ASPEN 2.1 0.5 3.2 19.9 SHRUB 1.6 2.2 0.3 18.9 SNAG 8.2 3.3 57.9 94.5 0.8 1.8 I<5 41.1 35.6 7.7 15.6 60.3 4.1 I< 10 17.6 0.3 30.2 0.4 I< 15 0.5 1.3 1.6 I<20 7.4 0.3 0<5 13.8 1.3 2.1 11.4 0< 10 10.4 4.0 23.6 0< 15 15.1 5.1 3.1 O<20 2.0 2.0 WIDTH 2.607 4.997 1.726 2.556 1.116 2.171 6.086 TIME 730 2822 6146 1184 550 398 5116 32 T A B L E VI . Pianka's index of niche overlap for the species shown in Table V . B O C H D E J U G R J A N O F L R B N U Y R W A A M R O 0.635 0.766 0.357 0.041 0.590 0.311 B O C H 0.230 0.426 0.189 0.855 0.399 D E J U 0.340 0.101 0.093 0.407 G R J A 0.940 0.233 0.226 N O F L 0.012 0.063 R B N U 0.096 33 TABLE VII. Food types seen to be consumed by 11 species. F = observed in the field, L = observed only in samples from ligatured nestlings. Spi = spiders, Egg = spider eggcases, Col = Coleoptera, Dip = Diptera, Aph = aphids, Hop = leafhoppers, Hym = Hymenoptera, Lep = adult Lepidoptera, Cat = larval Lepidoptera or sawflies, Ort = Orthoptera, Fly = insects in flight, Sna = snails and slugs, Spr = spruce seeds, Con = young spruce cones. Bird species abbreviations are explained in Appendix 1. Species Food type Spi Egg Col Dip Aph Hop Hym Lep Cat Ort Fly Sna Spr Con AMRO F BOCH F F F F CHSP F F DEJU L L F L L F F F L GRJA F PIGR F PISI F F TRSP F WWCR F YRWA F F F 34 DISCUSSION Total densities The estimated densities of passerine and piriform birds pooled for all grids were >121.8 males/km2 in 1988 and >155.4 males/km2 in 1989. These totals include only species detected ten times or more in a year. Inclusion of all species would raise the estimates by up to one third. In addition, I counted only singing males; non-territorial, non-singing males may also have been present. My estimates fall within the range of 107.5 to 587.5 males/km2 cited by Theberge (1976) for spruce forests in northern British Columbia and southwestern Mackenzie, and within or just below the range of 133.3 males/km2 to 200 males/km2 reported by James and Rathbun (1981) for pure coniferous forests across the United States and Canada. They are lower than the densities given by Erskine (1977) for most other boreal forest types and by James and Rathbun (1981) for all other forest habitats. This is probably due to low levels of food in coniferous forests in general and northern spruce forests in particular. I will discuss the relationship between food availability and bird density in Chapter 3. Species diversity Both richness and evenness were low. Only 38 species were recorded, and only seven of these (yellow-rumped warbler, dark-eyed junco, boreal chickadee, Swainson's thrush, American robin, grey jay and ruby-crowned kinglet) were detected ten times or more during each of the 1988, 1989 and 1990 point counts. 35 Richness Coniferous forests typically have few bird species (James and Rathbun 1981). This is probably due to their structural and biological simplicity (MacArthur and MacArthur 1961, Barden et al. 1986, Moermond 1986, Arnold 1988, Bogolyubov 1988, Helle and Fuller 1988, Urban and Smith 1989, Verner and Larson 1989). In addition, northern regions have relatively few bird species (Brown and Gibson 1983). This latitudinal trend has been documented in many taxa (Fischer 1960). Many explanations have been proposed (Ehrlich and Roughgarden 1987, Turner et al. 1988, Powell 1989), but there is still no consensus on which is correct. Evenness Domination of boreal spruce forest bird communities by a few species, especially the yellow-rumped warbler and dark-eyed junco, has been widely reported (see Theberge 1976). Gillespie and Kendeigh (1982) and Erskine (1977) suggest that this unevenness is due to the ecological "immaturity" of the habitat, and that older, climax forests generally have more even communities. However, Theberge (1976) found almost no change in the bird community during succession after fire in the Kluane area. Moreover, the level of unevenness I observed was similar to that in many communities. It probably indicates an underlying lognormal distribution. This would suggest that the community was structured by many independent factors, rather than by a single agent such as competition or niche preemption (May 1975). The yellow-rumped warbler's success is probably due to its behavioural plasticity (Kolasa 1989). The dark-eyed junco is much more specialized. However, its preferred habitat (open forest and forest edges and clearings) is common throughout North America. The junco is therefore abundant over 36 most of the continent. Densities of species tend to decrease away from the centres of their ranges (Brown 1984, Gaston 1990), but the junco is so abundant that it is common even in distant areas such as the southwestern Yukon. Species' niches The common species in the study area differed in their habitat preferences, and generally showed little overlap in their use of foraging substrates. However, there was little evidence of dietary specialization, as is found in many guilds (e.g. Grant 1986, Dickman 1988). Diet specialization usually depends on the availability of a wide range of prey sizes. There were very few large arthropods in the study area: large carabids, grasshoppers and moths (larvae and adults) were the only arthropods too large for small passerines to consume. Several more specialized bird species were present in the study area (for example, alder and olive-sided flycatchers, Say's phoebes and western wood pewees, all of which catch flying insects by hawking from a perch). However, I have no data on the diets of these species, since I did not see them foraging. The seven most abundant species included two very different Fringillids, one Parid, one Regulid, one Corvid and two Turdids. The ecological and taxonomic differentiation of the dominant species may allow their coexistence. However, this assumes that interspecific competition is or was important in structuring the community, and that low overlap among species reduces competition. The first assumption is untested, and the second is often false (Holbrook and Schmitt 1989, Mahdi and Law 1989). 37 Variation in community structure Temporal variation The species list in Appendix 1 is very similar to those given by Hoefs (1973) for 1969-1972 and Theberge (1976) for 1973. Relative abundances of species were approximately the same in 1987-1990 as they were during these earlier studies. Total density was approximately the same in 1988 and 1989, the only two years which can be meaningfully compared. This constancy agrees with Erskine's (1977) statement that boreal bird communities are "comparatively unvarying". However, Jarvinen (1979) found that communities of northern European breeding birds were temporally more variable than those further south. Wiens (1989) describes several studies which demonstrate considerable variation in bird communities, and remarks that "species constancy in communities is noteworthy, given the recent stress on non-equilibrium properties of communities". The overall constancy concealed a large increase in the yellow-rumped warbler population between 1989 and 1990, and marked decreases in the populations of dark-eyed juncos (1989-1990) and robins (1988-1989). Jarvinen (1979) and Solonen (1986) describe similar "quasi-stability" in communities of northern European birds. Changes in different directions in different species may indicate competition during the breeding season (Solonen 1986, Monkkonen 1990). However, most migratory passerines spend less than one third of their lives on breeding grounds (Morse 1989). Changes in abundance are therefore likely to be due to factors acting outside the breeding season: either competition (Oksanen 1987) or density-independent processes such as habitat change and environmentally-induced changes in food supply (Wiens 38 1977, Solonen 1986, Menge and Olson 1990). The latter processes may also operate during the breeding season. Spatial variation The bird community in the study area was typical for subarctic western Canada. Of Erskine's (1977) list of characteristic birds of boreal spruce forests west of 105°W, only the golden-crowned kinglet (Regulus satrapa) was not recorded in the Shakwak Trench in 1987-1990. There was rather little variation among grids: the dominant species were present throughout the study area. Most of the variation occurred among the rarer species, and was probably due to the distribution of the habitats preferred by these species (Theberge 1976, Cody 1985). However, the "core-satellite" hypothesis (Hanski 1982) predicts a similar pattern without invoking habitat heterogeneity. Hanski (1982) describes a model incorporating an inverse relationship between distribution (i.e. the number of sites occupied within a region) and extinction rate, and stochastic variation in immigration and extinction rates. He shows that this model can generate a bimodal frequency distribution, with a large peak of rare "satellite" species, present in only a few sites, and a smaller peak of common "core" species, found in most or all sites. CONCLUSION This is the longest and most detailed study of a bird community in boreal North America that I am aware of. However, there is a need for further research. More intensive censusing would allow accurate estimation of populations of rare species populations, as well as more accurate measurement numbers of common species. Longer-term monitoring would 39 clarify population trends. Better information on habitat and food preferences in both breeding and non-breeding areas would provide insight into the way the community is structured. 40 CHAPTER 3: EFFECTS OF FERTILIZATION INTRODUCTION This chapter reports the results of an experimental study of part of a terrestrial food web. There have been several hundred descriptive studies of food webs (Pimm 1982), and a number of hypotheses have been developed to explain the general patterns that have emerged from these studies (Krebs et al. 1988, Lawton 1989, Cohen et al. 1990). However, there have been few experimental studies of food webs, particularly in terrestrial systems. The C.S.P. applied chemical fertilizer to Flint and Grizzly grids in 1987, 1988 and 1989. This is one of the largest-scale experimental manipulations ever performed on a terrestrial ecosystem. I examined the effects of this large-scale manipulation on the plant - insect - passerine bird food chain. I also performed two experiments at a much smaller scale to study the effects of fertilization and its interaction with herbivory and insectivory in more detail. These experiments used 5m x 5m plots as experimental units. Questions I studied the importance of food limitation and predation at each trophic level by addressing the following questions. 1. Does the biomass of herb layer vegetation respond to fertilization? 2. Does the biomass of herb layer vegetation respond to exclusion of mammalian herbivores? 3. Does the biomass of arthropods respond to fertilization? 41 4. Does the biomass of arthropods respond to exclusion of insectivorous birds? 5. Do the populations and reproductive success of insectivorous birds respond to fertilization? Questions 1, 2 and 4 were addressed only on the small plots. Question 5 was addressed only on the large grids. Question 3 was addressed at both scales. Experiments and sampling schemes were designed to detect large responses by several broad categories of organisms. I did not attempt to study subtle responses or individual species in detail. METHODS (a) Methods for large grid experiment Fertilization of large grids Flint and Grizzly grids were fertilized from the air in 1987, 1988 and 1989. In 1987, 250 kg/ha of nitrogen (N2) were applied between 8th and 15th June; in 1988, 175 kg/ha of nitrogen, 25 kg/ha of potassium and 50 kg/ha of phosphorus were applied between 21st and 25th June; and in 1989, 125 kg/ha of nitrogen were applied on May 15th. Nitrogen was applied as ammonium nitrate, potassium as potash fertilizer and phosphorus as phosphate fertilizer. Silver and Sulphur grids were used as unfertilized controls. 42 Arthropod sampling on large grids All arthropod traps were set for three days. I sampled flying insects only in 1988. Sticky traps were set at each stake on G row on each grid on May 23rd 1988 and July 19th-20th 1988. Each trap consisted of a 30cm x 22cm sheet of 3mm Plexiglass. This was wedged into a notch cut in the top of a lm-high 5cm x 5cm stake so that its long axis was vertical. I set the traps by slipping a 38cm x 25cm clear plastic bag over the Plexiglass and covering both sides of the bag with Tree Tanglefoot from an aerosol spray can. Ground arthropods were sampled in two ways. On May 23rd 1988, I placed sticky traps identical to those described above flat on the ground at stakes GI through G20 on each grid. However, this method was ineffective; I therefore used pitfall traps at the same locations on July 19th-20th 1988. Pitfall traps were made from plastic cups 12cm deep and 8cm in diameter at the top. These were sunk into the ground, so that their rims were just below ground level, and filled to a depth of 3cm with water. In 1989, I used 30 pitfall traps per grid, and distributed them across the grids in open, grassy areas. These traps were set four times in 1989: on 25th-31st May, 17th-26th June, 4th-5th July and 25th July-lst August. Dan Doak measured arthropod damage to willow bushes (Salix glauca) on 18th-23rd June 1988, and kindly allowed me to use his data. He randomly selected four large bushes close to each of four stakes (E5, E15, P5 and P15) on each grid. Four groups of 20 shoots on each bush were scored for the presence or absence of damage. A shoot was defined as the new growth at the end of a twig, and damage as missing tissue. I repeated this procedure on 3rd-4th July 1989, but examined four groups of five shoots per bush. I used the proportion of damaged shoots as an index of damage. Willow bushes were 43 also sampled in 1989 by beating their foliage with a strong stick lm in length until no further arthropods were dislodged. I sampled 30 bushes approximately lm high by 1.5m in diameter from open, willow-dominated areas on each grid. Dislodged arthropods were collected and counted on a lm x lm hand-held beating sheet. Willows were sampled four times by this method in 1989: on llth-20th May, 16th-27th June, 6th-9th July and 26th July-10th August. Beating of willows and spruce trees (see below) was done between 1:30pm and 6pm, when arthropods were most active. Arthropods on spruce trees were also sampled by beating foliage with a lm stick. In 1988, the foliage of twenty trees on each grid was beaten to a height of 3m between the 1st and 4th September. A lm x lm tray was placed on the ground to collect dislodged arthropods. In 1989, I attempted to standardize the volume of foliage sampled by cutting two 80cm branch ends from a height of 2m on each of thirty trees on each grid. These branch ends were beaten on a heavy wire grid held over a cotton sheet. All trees sampled in 1989 were in fairly dense, mature stands. I sampled spruce trees four times in 1989: on llth-22nd May, 2nd-13th June, 30th June-3rd July and 26th July-lOth August. I converted arthropod numbers to an index of biomass by cubing the length in mm of each arthropod sampled, and summing these cubes. This method assumes that all arthropods are the same length and density. However, I found it preferable to the equations listed by Sage (1982); these gave absurdly high masses for large arthropods. 44 Bird population measures on large grids Population densities Bird populations were estimated using the point counts described in Chapter 2. Reproductive success I visited all passerine nests found in the study area at approximately four day intervals and recorded or calculated as many of the following variables as possible for each one: date of first egg, clutch size, number hatched and nestling weights. I also erected twenty nestboxes on each of the four study grids early in 1988. I hoped that these would provide easily accessible nests of hole-nesting species for the study of reproductive success and diets, but only two of the boxes were used in 1988 and only four in 1989. (b) Methods for small plot experiments Design of small plot experiments Small plot experiment 1: effects of different levels and types of fertilizer A randomized block design was used to examine the effects of different fertilizer treatments. One block of four 5m x 5m plots was marked out in an open, grassy area in each of the following eight locations: E5 and G15 on Sulphur, a control grid; BI and L l on Gravel Pit grid, where supplemental hare food was provided; K5 and M9 on Beaver Pond grid, which was surrounded by an electric fence to exclude large mammalian predators; and 45 J5 and E9 on "Rudy 1", a 200m x 200m grid surrounded by plastic fencing to exclude snowshoe hares. Stakes on this last grid were only 10m apart. Three treatments were used: NPK, equivalent to the application of fertilizer used on the large grids in 1988; 5NPK, five times NPK; and N, equivalent to the 1987 large grid application. The fourth plot in each block was used as a control. I applied fertilizer on 29th June 1988. Small plot experiment 2: effects of fertilization and passerine and herbivore exclusion I designed a factorial experiment to determine how the effects of fertilization were altered by the exclusion of passerines and mammalian herbivores. The experiment crossed two levels of fertilization, 2NPK and control, with three kinds of exclusion: passerine exclusion, using fine mesh fishing net staked out over plots; herbivore exclusion, using 2m fences of 1" mesh chicken wire; and control. I replicated the experiment four times, randomizing treatments spatially within each replicate. One replicate was set up in dense birch scrub outside the predator exclusion fence around Beaver Pond grid, one in the same habitat inside the fence, and two in willow scrub at opposite sides of the disused Microwave grid. The Microwave replicates used 5m x 5m fences erected in 1976. There was insufficient fishing net to cover all eight passerine exclusion plots. I therefore left the passerine exclusion plots on Microwave open. Fertilizer was applied on 28th June 1988. 46 Vegetation sampling on small plots Herb layer vegetation was sampled on 9th-15th August 1988 (small plot experiment 1) and 16th-24th August 1988 (small plot experiment 2). I chose six sample sites within each plot by throwing a 25cm x 25cm wire quadrat over my shoulder. Areas with less than 75% cover were not sampled. All living herbaceous and creeping woody vegetation inside each quadrat was clipped at ground level. I placed samples in plastic bags to prevent dehydration, and determined their total wet weight in the evening of the day they were sampled. Samples were then placed in a drying oven at 40°C for 60 hours, before being sorted and reweighed. I measured the dry weights of: Festuca altaica, the commonest grass and commonest species overall; Arctostaphylos uva-ursi, the commonest woody species; Epilobium angustifolium, usually the commonest forb; and all other species combined. I calculated the mean total wet weight and mean dry weight of each component for each plot, and measured the total species richness of each plot. Arthropod sampling on small plots I sampled flying and ground arthropods with the sticky traps and pitfall traps described above. One of each kind of trap was set on each plot on 18th-19th August 1988 (small plot experiment 1) and 20th-31st August 1988 (small plot experiment 2). I used the method described above to calculate indices of total arthropod biomass for each plot. I also calculated indices of biomass for Diptera, Hemiptera and Hymenoptera from sticky traps, and for spiders from pitfall traps. I calculated indices of arthropod damage to birch and willow shrubs by examining twenty shoots of each species on each plot on 11th-12th August 1988. Shoots were scored as 3 (high), 2 (medium), 1 (low) or 0 (none) for the 47 amount of leaf tissue missing. In addition, birch shoots were scored for the level of woolly aphid infestation, and the number of caterpillars on each willow shoot was counted. These numbers were summed to give total scores for each plot. Analysis of small plot data I ranked all data from small plots, because large or extremely large values from single plots, or even single samples, strongly affected the results of parametric tests. ANOVAs were performed on the ranked data. RESULTS (a) Results of large grid experiment Arthropods on large grids Figures 5, 6 and 7 show the biomass indices of flying insects, ground arthropods and arthropods on spruce trees in 1988. Biomasses of flying insects increased approximately five-fold from late May to late July on all four grids (Figure 5). This increase was significant (p < 0.00025, one-tailed t test). Biomasses of ground arthropods were much higher on fertilized grids, especially Flint, than on unfertilized grids in May, but not in late July/early August (Figure 6). The difference in May was not significant (p = 0.056). However, this test has very little power, since there were only two replicates; the difference was so great that it was probably biologically significant. Arthropod biomasses on spruce trees were more than twice as high on Flint than on any other grid in early September (Figure 7). Grizzly had the second 48 highest biomass index in September, but fertilized grids did not have significantly higher biomasses than unfertilized grids (p = 0.148). The levels of arthropod damage to willows are shown in Figure 8. The proportion of damaged shoots increased approximately three-fold from 1988 to 1989. The spring of 1989 was much warmer and dryer than that of 1988. In addition, willows were sampled two weeks later in 1989 than in 1988. The tendency for fertilized willows to show less leaf damage than fertilized willows in 1989 was not significant (p = 0.088). Figures 9, 10 and 11 show biomass indices for ground arthropods, arthropods on willows and arthropods on spruce trees in 1989. Ground arthropods showed no consistent seasonal trend and no significant differences between treatments (Figure 9). Arthropod biomasses on willows (Figure 10) increased sharply from May to late June, reaching their highest levels on unfertilized grids. However, they decreased on these grids after late June, while biomasses on fertilized grids remained steady. Spruce arthropods showed a similar sharp increase from May to late June (Figure 11). Biomasses were higher on Flint than on other grids during this period. After late June, biomass continued to increase rapidly on Grizzly and much more slowly on Flint and Sulphur, and decreased slightly on Silver. Fertilized grids had significantly higher biomasses of arthropods on spruce foliage than unfertilized grids by the last sampling period (p = 0.002). Summary of results for arthropods on large grids Flying, willow and spruce arthropods increased greatly from May to late July / August on all grids. There was no such trend in ground arthropods. Arthropod damage to willow leaves increased from 1988 to 1989. There was a greater biomass of ground arthropods on fertilized grids than 49 FIGURE 5. Biomass indices for flying insects on large grids in 1988. Points are offset slightly to allow error bars to be displayed. 20 30 40 50 60 70 80 90 Days after May 1st O SULPHUR • SILVER • GRIZZLY • FLINT 50 FIGURE 6. Biomass indices for ground arthropods on large grids in (a) May and (b) July / August 1988. Sticky traps were used in May, pitfall traps in July / August. Filled bars in Figures 6 and 7 indicate fertilized grids. 51 600 CO TD I Q . CD Q . X CD "O _fZ CO CO cO E o i n 500 400 300 200 100 0 mmmm (a) FLINT GRIZZLY SILVER SULPHUR 52 900 800 700 -Q. 600 o3 500 C L | 4 0 0 CO § 300 o CO 200 100 -0 FLINT GRIZZLY SILVER SULPHUR 53 FIGURE 7. Biomass indices for arthropods on spruce trees on large grids in September 1988. 2100 1800 CD 1500 CD CD c - 1200 X CD T D C CO CO CO E o CO 900 600 300 0 - V-* ' . 8 1 vxv< F L I N T GRIZZLY SILVER SULPHUR 54 FIGURE 8. Proportions of willow shoots damaged by arthropods on large grids. 60 19 JUNE 88 3 JULY 89 O SULPHUR • SILVER • GRIZZLY • FLINT 55 FIGURE 9. Biomass indices for ground arthropods on large grids in 1989. Note that the vertical axes on Figures 9,10 and 11 have logaritiimic scales. 5Q 1 I I I I I I I I 20 30 40 50 60 70 80 90 100 Days after May 1st O SULPHUR • SILVER • GRIZZLY • FLINT 56 FIGURE 10. Biomass indices for arthropods on willows on large grids in 1989. 2 I i i 0 50 100 Days after May 1st O SULPHUR • SILVER • GRIZZLY • FLINT 57 FIGURE 11. Biomass indices for arthropods on spruce trees on large grids in 1989. 0 50 100 Days after May 1st . O SULPHUR • SILVER • GRIZZLY • FLINT 58 unfertilized grids in May 1988, and a greater biomass of arthropods on fertilized spruce trees than unfertilized spruce trees in late July / August 1989. Differences were less clear between fertilized and unfertilized grids in: spruce arthropods in early September 1988; willow damage in early July 1989 (less on fertilized grids); willow arthropods in late July/August 1989; and spruce arthropods between early June and mid-July 1989. Birds on large grids Populations Figures 12 (a) to (g) show the population trends of the seven commonest species on fertilized and unfertilized grids. Five species were detected more often on unfertilized grids in 1988 and on fertilized grids by 1990. The exceptions, the boreal chickadee and ruby-crowned kinglet, were detected more often on fertilized grids throughout. Treatments differed significantly only in the number of detections of yellow-rumped warblers in 1988 (more on unfertilized grids, p = 0.0475) and boreal chickadees in 1989 (more on fertilized grids, p = 0.0305). Neither of these differences are significant at the Bonferroni-adjusted acceptance level of p = 0.00263. Populations on fertilized grids had higher annual growth rates than populations on unfertilized grids significantly more often than expected by chance (p < 0.01, chi-squared test). Figure 13 shows the numbers of detections of all species combined. The pattern is similar to that for most individual species. There were more detections on unfertilized grids in 1988, slightly more on fertilized grids in 1989 and considerably more on fertilized grids in 1990. None of the differences were significant. 59 Reproductive success One Swainson's thrush nest, six boreal chickadee nests and 19 dark-eyed junco nests were found in 1988 and 1989. The only chickadee nest found in an unfertilized area (Sulphur grid) had seven eggs, the first of which was laid on May 10th. The mean clutch size on fertilized grids was 6.25, and first eggs were laid an average of 25.7 days after May 1st. All nests except one were preyed upon by red squirrels (Tamiasciurus hudsonicus) Table VIII shows the mean date of first egg, clutch size, number of young hatched and nestling growth rate for junco nests on fertilized and unfertilized grids. First eggs were laid significantly earlier on fertilized grids (p = 0.0025). No clutches were begun in the ten days after June 5th, which suggests that the clutches started after this were second attempts. However, even if these clutches are excluded from the analysis, there is still a significant difference between treatments: clutches were started an average of 7.42 days earlier on fertilized grids than on unfertilized grids (p = 0.029). Fertilization did not have a significant effect on clutch size (p = 0.087) or nestling growth rate (p = 0.0805). Only one nest survived intact until hatching on the fertilized grids. Therefore, I could not test whether fertilization affected the number of young hatched. Summary of results for birds on large grids There is weak evidence that bird populations increased in response to fertilization (see Figures 12 and 13). Dark-eyed juncos nested earlier on fertilized grids. 60 FIGURE 12. Numbers of detections of seven common bird species on fertilized grids (filled circles) and unfertilized grids (open circles). 62 0 i 1 1 1— 1 9 8 8 . 1 9 8 9 . 1 9 9 0 . Y E A R Number of detections per 1 1 point counts o cn o cn 64 1 5 CO ~_ Z5 O o "c "o C L CD C L CO C . o ~o _CD CD c5 CD 1 0 5 -0 1 9 8 8 . 1 9 8 9 . 1 9 9 0 Y E A R Number of detections per 1 1 point counts O N) GO 4^ CJl O) I I I I I I I 67 1 9 8 8 . 1 9 8 9 . 1 9 9 0 . Y E A R 68 FIGURE 13. Numbers of detections of all bird species combined on fertilized and unfertilized grids. CO "c Z5 o o ~c "o CD-CD C L CO a o CD 8 0 7 0 -6 0 -5 0 4 0 -CJ B 3 0 CD TD 2 0 -1 0 -0 1 9 8 8 . 1 9 8 9 . 1 9 9 0 . Y E A R i 69 TABLE VIII. Measures of reproductive performance of dark-eyed juncos on fertilized and unfertilized grids (mean +/- standard error). Sample sizes are shown in parentheses. Clutch sizes from 1988 and 1989 were pooled, since there was no significant difference between years. Other measures are from 1989 only. Fertilized Unfertilized Date of first egg 20.25+/- 37.30+/-(days after May 1st) 2.50 (4) 13.34 (10) Date of first egg, 20.25+/- 27.67+/-excluding probable 2.50 (4) 2.09 (6) second attempts Clutch size 4.00 +/- 4.54 +/-0.36 (5) 0.14 (13) Number of young 1.00 (1) 3.86 +/-hatched 0.34 (7) Nestling growth rate 1.82 +/-(g/day) 0.08 (3) 2.09 +/-0.13 (9) 70 (b) Results of small plot experiments. Small plot experiment 1: effects of different levels and types of fertilizer Vegetation Fertilization had no significant or consistent effects on herb layer vegetation. Arthropods Figure 14 shows that birch leaves were significantly less damaged on 5NPK plots than on N plots (p < 0.05, Tukey's Honestly Significant Difference test; Bonferroni-adjusted acceptance level = 0.005). 5NPK plots also had fewer flying Hemiptera than other plots, but this difference is not significant. There were no other significant effects or noteworthy trends. Small plot experiment 2: effects of fertilization and passerine and herbivore exclusion Vegetation There were no significant effects of fertilization or exclusion on the Beaver Pond plots, although unfertilized open plots had markedly lower total wet weights and dry weights of Festuca. Figure 15 shows that fertilization increased total wet weight on the Microwave plots (p = 0.001), and Figure 16 shows that fenced plots at Microwave had increased the total dry weights of species other than Festuca, Arctostaphylos and Epilobium (p < 0.0005). 71 Arthropods Fertilized birch leaves were slightly less damaged than unfertilized ones on the Beaver Pond plots. Unfertilized, open plots had the lowest biomasses of ground arthropods and ground spiders, and the least damage to willow leaves. None of these differences was significant. Fenced plots had the lowest biomasses, and open plots the highest biomasses of flying insects. This tendency was significant only for Hemiptera (p = 0.043) (see Figure 17). There was an increased biomass of flying Hymenoptera on fertilized plots at Microwave (p = 0.036) (Figure 18), but a reduced number of caterpillars on willow leaves on these plots (not significant). Fertilized, fenced plots had the fewest woolly aphids on birch leaves (not significant). Fenced plots had greater biomasses of Diptera (p = 0.016) and all flying insects combined (p = 0.005) (Figure 19), and the most damage to birch leaves (not significant). Summary of results for small plots Fertilization increased the biomass of herb layer vegetation only on the Microwave plots. Fertilized plots had increased biomasses of flying Hymenoptera, but reduced numbers of willow caterpillars, at Microwave. Fertilized plots also had reduced biomasses of flying Hemiptera at Beaver Pond, and reduced damage to birch at Beaver Pond and in experiment 1. Fenced plots at Microwave had increased biomasses of uncommon plant species and flying insects, and increased damage to birch leaves. Fenced plots at Beaver Pond had a reduced biomass of flying insects. Unfertilized, open plots at Beaver Pond had the least vegetation and the lowest biomasses of ground arthropods, as well as the least damage to willows. Fertilized, fenced plots at Microwave had the fewest woolly aphids on birch. 72 FIGURE 14. Effects of small plot experiment 1 treatments on the level of arthropod damage to birch leaves. A rank of 1 indicates that a given treatment had the lowest level of damage, a rank of 4 that it had the highest. 4 3 -< cc 2 -1 -0 73 FIGURE 15. Effects of small plot experiment 2 treatments on the total wet weight of herb layer vegetation on Microwave plots. A rank of 6 indicates that a given treatment had the highest wet weight. F = fenced plots, U = unfenced plots. < CC 7 6 5 -4 -3 -2 -1 -0 0 o 74 FIGURE 16. Effects of small plot experiment 2 treatments on the total dry weight of plant species other than Festuca, Arctostaphylos and Epilobium on the Microwave plots. < 7 6 -5 -4 -2 -1 -0 75 FIGURE 17. Effects of small plot experiment 2 treatments on the biomass of flying Hemiptera on the Beaver Pond plots. N = netted plots. 6 5 -4 -< cc 3 -0 76 FIGURE 18. Effects of small plot experiment 2 treatments on the biomass of flying Hymenoptera on the Microwave plots. < rr 7 6 5 -4 -3 -2 -0 77 FIGURE 19. Effects of small plot experiment 2 treatments on the total biomass of flying insects on the Microwave plots. 7 6 5 -4 < •C 2 -1 -0 78 DISCUSSION Question 1. Does the biomass of herb laver vegetation respond to fertilization? Vegetation biomass increased in response to fertilization on the Microwave small plots. This is consistent with the results of studies on the large grids: David Hik (pers. comm.) found a two- to three-fold increase in Festuca biomass on fertilized grids; Jamie Smith (pers. comm.) found that the biomasses of Epilobium, Mertensia paniculata and Solidago multiradiata approximately doubled oh fertilized grids, although Lupinus arcticus (a nitrogen fixer) and Anemone parviflora showed no such response. In addition, spruce trees have shown a dramatic growth response on fertilized grids (Tony Sinclair, pers. comm.). Studies elsewhere have demonstrated macronutrient limitation of boreal forest vegetation (Chapin 1980, Chapin et al. 1987, DeAngelis et al. 1989). In contrast, vegetation on the other small plots showed no response to fertilization. The response of vegetation on the large grids was also very variable (Jamie Smith, pers. comm.). This was probably due to great spatial variation in soil moisture content. Plant growth may have been limited primarily by macronutrients on Microwave, but by water on dryer plots. Light and/or micronutrients may also have limited growth elsewhere. In addition, the design of the small plot experiments may have precluded detection of subtle responses to fertilization. Fertilization in late June may have been too late. Several studies have shown that nutrients are most limiting early in spring (Goldberg and Miller 1990 and references therein). Sampling may have been done too soon after fertilization. Plants in poor soils are typically slow to respond to nutrient additions, and perennial plants may 79 respond only after a lag of a year (Grime 1977, Chapin and Shaver 1985, Shipley and Keddy 1988, Campbell and Grime 1989, McGraw and Chapin 1989, Polley and Detling 1989, Huenneke et al. 1990, but see Jackson et al. 1990). Finally, six samples per plot may have been too few. Summary The results from the Microwave small plots indicate that fertilization increases vegetation biomass. Experiments on the large grids support this conclusion, and show that the effect may be dramatic. Vegetation on other small plots, and on parts of the large grids, did not respond to fertilization. This suggests that vegetation was strongly limited by macronutrients only when other factors (such as water) were not limiting. Question 2. Does the biomass of herb layer vegetation respond to exclusion of mammalian herbivores? Fenced plots on Microwave had significantly greater biomasses of species other than Festuca, Arctostaphylos and Epilobium. In 1990, they had particularly high biomasses of Mertensia (Smith, pers. comm.). This suggests that grazing and/or trampling limited the standing crops of some herb layer species (Naiman 1988 discusses the effects of herbivores on boreal vegetation). Again, this effect may have been important only on Microwave because other factors were not limiting there. Alternatively, the Microwave plots may have shown a significant response because fences there were erected in 1976. The 12-year period of large herbivore exclusion may simply have allowed the response of herb layer vegetation to reach a detectable level. 80 However, it also resulted in a much taller and denser growth of willows (pers. obs.), which probably benefitted shade-tolerant species of herb layer plants. Unfertilized, open plots at Beaver Pond had lower wet weights of all plants combined and dry weights of Festuca than other plots. This is consistent with the results discussed above. It suggests an interaction between fertilization and exclusion, which was also noted on the Microwave plots in 1990 (Smith, pers. comm.). Summary Vegetation biomass increased when herbivores were excluded. However, this effect was large only on the Microwave plots, which had been fenced for 12 years. This suggests that vegetation is limited, but not severely, by mammalian herbivores. Question 3. Does arthropod biomass respond to fertilization? By far the most important changes in arthropod biomass were the seasonal increases in flying, willow and spruce arthropods on all of the large grids. Such seasonality is typical of most insect populations (Wolda 1988), especially those in the North (Danks and Foottit 1989). Seasonally varying environmental parameters affect insect populations directly and through their effects on plant growth. There were more ground arthropods on fertilized grids than unfertilized grids in May 1988. This contrasts with the results of Rushton et al. (1989) and Snodgrass and Studelbacher (1989), who found that fertilization had no effect and a negative effect on ground arthropods respectively. The great majority of ground arthropods at Kluane were 81 predatory carabid beetles and spiders. Therefore, fertilization did not directly affect their food supply. They may have benefitted from increases in their prey (Wise 1975, Juliano 1986) and/or from improvements to their habitat. The latter suggestion is supported by the results of small plot experiment 2: unfertilized, open plots at Beaver Pond, which had the fewest ground arthropods, also had the least herb layer vegetation. There were more arthropods on willows on fertilized grids than unfertilized grids in late summer 1989. The rapid decline in arthropod biomass on unfertilized willows after late June may have resulted from decreases in the palatability and nutrient quality of old leaves (Bach 1990). Fertilization may have delayed leaf senescence, and allowed arthropod biomasses to remain high. Many studies have demonstrated that populations of willow-feeding arthropods are affected by host quality (Fritz and Price 1988, Preszler and Price 1988, Waring and Price 1988, Denno et al. 1990, Fritz and Nobel 1990). In contrast to the above result, fertilization reduced the number of caterpillars on willows on the Microwave plots. There were more flying Hymenoptera on these plots. These may have been parasitic species responsible for increased caterpillar mortality. Alternatively, sawfly larvae (included in the general term "caterpillars") may have developed more rapidly on fertilized willows, and metamorphosed into winged adults before the sampling period in mid-August. The reduction of development time in response to increased food is described by Slansky and Rodriguez (1987), Gruber and Dixon (1988) and Knisley and Juliano (1988). A third possibility is that flying Hymenoptera were attracted by an increase in the number of flowers on fertilized plots. 82 Fertilization reduced damage to birch leaves in both small plot experiments. Fertilized, fenced plots on Microwave had fewer birch woolly aphids than other plots. The increase in nutrients may have allowed birches to synthesize more defensive secondary compounds, although other studies have shown that this is typically a response to defoliation in birch (Haukioja and Niemela 1979, Tuomi et al. 1988, Hanhimaki 1989). The theory that plants well-supplied with nutrients are less vulnerable to herbivory is discussed by White (1984), Heinrichs (1988) and Larsson (1989). Arthropods on spruce trees on the large grids benefitted from fertilization in late summer 1988 and 1989. Again, fertilization may have delayed or reduced age-related declines in leaf palatability and/or quality. In 1989, arthropods on fertilized spruces continued to increase after late June. Those on unfertilized spruces did not increase after this date, but neither did they decrease like arthropods on unfertilized willows. Spruce needles would not be expected to deteriorate as rapidly as willow leaves, since the latter are deciduous. Neither spruce nor willow arthropods responded to fertilization in spring. This suggests that either the arthropods or their host plants were not nutrient-limited at this time of year. Summary The results show that fertilization results in increased biomasses of some arthropod groups at certain times of year. 83 Question 4. Does arthropod biomass respond to exclusion of insectivorous birds? Arthropods showed no response to passerine exclusion, which suggests that they were not limited by bird predation. However, I am not confident that this was the case, due to the design problems mentioned above. Many studies have found that predation does affect arthropod numbers (Askenmo et al. 1977, Holmes et al. 1979, Loyn et al. 1981, Gunnarsson 1983, DeGraf 1987, Bernays 1989, Crawford and Jennings 1989, Higashiura 1989, Smith and Proctor 1989, Whelan et al. 1989). The higher biomasses of flying insects on the fenced plots at Microwave were probably due to the shelter from the wind provided by tall, dense shrubs on these plots. There is no obvious explanation for the decrease in flying insects on fenced plots near Beaver Pond. Summary Arthropods did not respond to exclusion of passerine birds. However, this may be due to problems with experimental design. Question 5. Do the populations and reproductive success of insectivorous birds respond to fertilization? There was some evidence that bird populations increased in response to fertilization. The increases were small, which contrasts with the large increases in survival or reproduction of passerine birds demonstrated by many food supplementation studies (Krebs 1971, Clamens 1987, Arcese and Smith 1988, Brittingham and Temple 1988, Desrochers et al. 1988, Knight 1988, Kraft 1988, review by Boutin 1990, Simons and Martin 1990). This is 84 probably because most of these authors supplied extra seeds in winter or early spring. Oksanen (1987) emphasizes that this is usually the "crunch" time of year for migratory passerines, although Martin (1987) suggests that some birds are food-limited during the summer. In my study, the increase in arthropod abundance between May and July far outweighed the increase in bird numbers due to production of young. Oksanen's view is therefore probably correct. Fertilization had little effect on arthropod populations in early spring, so the small response by birds is not surprising. Dark-eyed juncos nested earlier on fertilized grids than on unfertilized grids. This species nests on the ground in open areas, and conceals its nests in tussocks of grass. If fertilization resulted in an earlier and lusher growth of grass, juncos would have been able to begin nesting earlier. In addition, ground arthropods were the only group to show a response to fertilization in spring. Juncos feed mainly on the ground, so their food supply probably increased earlier than that of most other species. Early nesting is a common response to food supplementation (Arcese and Smith 1988), and is often positively correlated with high reproductive success (Perrins and Birkhead 1983, Hochachka 1990 and references therein). It could be particularly valuable in northern areas, where there is only a short time after breeding for birds to moult and become strong enough to migrate. The failure of juncos to become more abundant in response to fertilization, despite their earlier nesting dates, suggests that a longer-term study will be necessary to detect population-level effects of fertilization, or that other factors are important. The extremely high rates of nest predation may have resulted from my frequent visits (Willis 1973, Lenington 1979, Fetterolf 1983, Westmoreland and Best 1985). However, passerine birds, especially species that nest on the ground, often suffer very high nest predation rates (Wilcove 1985, O'Connor 85 1986, Martin 1988a, Pieman 1988, McLaughlin and Montgomerie 1989, Trevelyan and Read 1989, Weathers and Sullivan 1989, Yahner et al. 1989, Leitsma et al. 1990, Moller 1990). Martin (1988b) suggests that high predation on eggs and nestlings may be an important factor in the organization of bird communities. It is possible that nest predation limited the populations of some species in my study. Summary Bird populations showed small increases in response to fertilization. Dark-eyed juncos nested earlier on fertilized grids. This may have been due to an earlier growth of grass on these grids, since juncos conceal their nests in ground vegetation. CONCLUSION Figure 20 summarizes the results. It shows that vegetation was limited by both nutrients and mammalian herbivores. The response to fertilization was greater and more rapid than the response to herbivore exclusion. This suggests that "bottom-up" limitation is more important than "top-down" limitation in this system. However, herbivorous arthropods were not excluded by fences. The total effect of herbivory may therefore be considerably greater than the results show. Arthropods were limited by food, but not by predation. The latter result may have been due to limitations of the experiment. Birds may actually have some impact on arthropod populations. However, it appears that food is the more important limiting factor at this level. 86 Birds were weakly limited by arthropods. However, arthropod numbers were supplemented only in summer. Birds may be much more food-limited in early spring. I did not study the effects of predators on insectivorous bird populations. Overall, limitation from below appears to dominate this system. All three trophic levels are limited from below. This rules out the pure "top-down" view. However, plants, and possibly other levels, are also limited from above. The extreme "bottom-up" view is therefore not applicable. The results are also inconsistent with any of the hypotheses predicting alternation of top-down and bottom-up limitation. Herbivory appears to affect vegetation more than predation affects arthropods, contrary to McQueen et al.'s (1986) predictions. Of the ideas described in Chapter 1, only the suggestion that all trophic levels are regulated from both above and below is not contradicted by my results (note that my experiments examined limitation, not regulation). However, my study is far from complete. Work on the links indicated by open arrows in Figure 20 is needed to complete even this crude picture of the system. A clearer indication of how the system functions will be gained by conducting more intensive, quantitative research on each link, and by splitting the categories used here into smaller, more homogeneous groups. Finally, long term research will allow responses to perturbations to become large and easily measurable, and will clarify how the dynamics of the system vary through time. 87 FIGURE 20. Summary of results of fertilization and exclusion experiments. Only interactions between trophic levels are shown. A -> B means "A limits B". "Y" indicates that there was some evidence of limitation, "N" that there was no such evidence. Open arrows with question-marks indicate links that I did not study. NEST/ADULT PREDATORS A V PASSERINE BIRDS 1!" ARTHROPODS PLANTS | Y (1) NUTRIENTS MAMMALIAN PREDATORS A V HERBIVOROUS MAMMALS 88 REFERENCES Andrewartha, H. G., and L. C. Birch. 1954. The distribution and abundance of animals. University of Chicago Press, Chicago, Illinois, USA. Arcese, P., and J. N. M. Smith. 1988. Effects of population density and supplemental food on reproduction in song sparrows. Journal of Animal Ecology 57:119-136. Arnold, G. W. 1988. The effects of habitat structure and floristics on the densities of bird species in wandoo woodland. Australian Wildlife Research 15:499-510. Askenmo, C , A. von Bromssen, J. Ekman, and C. Jansson. 1977. Impact of some wintering birds on spider abundance in spruce. Oikos 28:90-94. Bach, C. E. 1990. Plant successional stage and insect herbivory: flea beetles on sand-dune willow. Ecology 71:598-609. Bailey, N. T. J. 1981. Statistical methods in biology. Second edition. Hodder and Stoughton, London, U.K. Barden L. S., R. D. Brown, J. F. Matthews, and R. E. Cross. 1986. Vegetation structure and bird species diversity in a North Carolina Piedmont forest. Journal of the Elisha Mitchell Scientific Society 102:35-41. Bernays, E. A. 1989. Host range in phytophagous insects: the potential role of generalist predators. Evolutionary Ecology 3_:299-311. Bogolyubov, A. S. 1988. Dependence of spatial structure and numbers in bird communities on surface area of phytoelements in forest biocenoses. Soviet Journal of Ecology 19:289-293. Bonan, G. B., and H. H. Shugart. 1989. Environmental factors and ecological processes in boreal forests. Annual Review of Ecology and Systematics 20:1-28. Bostock, H. S. 1948. Physiography of the Canadian Cordillera, with special reference to the area north of the fifty-fifth parallel. Geological Survey of Canada Memoranda 247. . 1952. Geology of the northwest Shakwak Valley, Yukon Territory. Geological Survey of Canada Memoranda 267. Boutin, S. 1990. Food supplementation experiments with terrestrial vertebrates: patterns, problems, and the future. Canadian Journal of Zoology 68:203-220. Brittingham, M. C , and S. A. Temple. 1988. Impacts of supplemental feeding on survival rates of black-capped chickadees. Ecology 69:581-589. Brown, J. H. 1984. On the relationship between abundance and distribution of species. American Naturalist 124:255-279. 89 Brown, J. H., and A. C. Gibson. 1983. Biogeography. C. V. Mosby, St Louis, Missouri, USA. Brown, J. H., D. W. Davidson, J. C. Munger, and R. S. Inouye. 1986. Experimental community ecology: the desert granivore system. Pages 41-62 in J. Diamond and T. J. Case, editors. Community ecology. Harper and Row, New York, New York, USA. Burnham, K. P., D. R. Anderson and J. L. Laake. 1980. Estimation of density from line transect sampling of biological populations. Wildlife Monographs 72. Campbell, B. D., and J. P. Grime. 1989. A comparative study of plant responsiveness to the duration of episodes of mineral nutrient enrichment. New Phytologist 12:261-267. Carpenter, S. R., J. F. Kitchell, and J. R. Hodgson. 1985. Cascading trophic interactions and lake productivity. BioScience 35:634-639. Chapin, F. S. III. 1980. The mineral nutrition of wild plants. Annual Reviews of Ecology and Systematics 11:233-260. Chapin, F. S. Ill, A. J. Bloom, C. B. Field, and R. H. Waring. 1987. Plant responses to multiple environmental factors. BioScience 37:49-57. Chapin, F. S. Ill, and G. R. Shaver. 1985. Individualistic growth response of tundra plant species to environmental manipulations in the field. Ecology 66:564-576. Clamens, A. 1987. [The role of food in determining demographic parameters in tits Paridae in holm-oak woodland: preliminary results.] Alauda 5_:254-266. French, English summary. Cody, M. L., editor. 1985. Habitat selection in birds. Academic Press, New York, New York, USA. Cohen, J. E., T. Luczak, C. M. Newman, and Z.-M. Zhou. 1990. Stochastic structure and nonlinear dynamics of food webs: qualitative stability in a Lotka-Volterra cascade model. Proceedings of the Royal Society of London (B) 240:607-627. Crawford, H. S., Jr., and D. T. Jennings. 1989. Predation by birds on spruce budworm Choristoneura fumifera: functional, numerical and total responses. Ecology 70:152-163. Danks, H. V., and R. G. Foottit. 1989. Insects of the boreal zone of Canada. Canadian Entomologist 121:625-690. DeAngelis, D. L., P. J. Mulholland, A. V. Palumbo, A. D. Steinman, M. A. Huston, and J. W. Elwood. 1989. Nutrient dynamics and food-web stability. Annual Reviews of Ecology and Systematics 20:71-96. DeGraf, R. M. 1987. Breeding birds and gypsy moth defoliation: short-term responses of species and guilds. Wildlife Society Bulletin 15:217-221. 90 Denno, R. F M S. Larsson, and K. L. Olmstead. 1990. Role of enemy-free space and plant quality in host plant selection by willow beetles. Ecology 7_1:124-137. Desrochers, A., S. J. Harmon, and K. E. Nordin. 1988. Winter survival and territory acquisition in a northern population of black-capped chickadees. Auk 105:727-736. Dickman, C. R. 1988. Body size, prey size, and community structure in insectivorous mammals. Ecology 69:569-580. Douglas, G. W. 1974. Montane zone vegetation of the Alsek River region, southwestern Yukon. Canadian Journal of Botany. 52:2505-2532. Ehrlich, P. R., and J. Roughgarden. 1987. The science of ecology. Macmillan, New York, New York, USA. Ehrlich, P. R., D. S. Dobkin, and D. Wheye. 1988. The birder's handbook. Simon and Schuster, New York, New York, USA. Emlen, J. T. 1971. Population densities of birds derived from transect counts. Auk 88:323-342. Erskine, A.J. 1977. Birds in boreal Canada. Canadian Wildlife Service Report Series Number 11. Fetterolf, P. M. 1983. Effects of investigator activity on ring-billed gull behaviour and reproductive performance. Wilson Bulletin 95:23-41. Fischer, A. G. 1960. Latitudinal variations in organic diversity. Evolution 14:64-81. Fritz, R. S., and J. Nobel. 1990. Host plant variation in mortality of the leaf-folding sawfly on the arroyo willow. Ecological Entomology 15:25-35. Fritz, R. S., and P. W. Price. 1988. Genetic variation among plants and insect community structure: willows and sawflies. Ecology 69:845-856. Gaston, K. J. 1990. Patterns in the geographical ranges of species. Biological Reviews 65:105-129. Gillespie, W. L., and S. C. Kendeigh. 1982. Breeding bird populations in northern Manitoba. Canadian Field-Naturalist 9J3:272-281. Goldberg, D. E., and T. E. Miller. 1990. Effects of different resource additions on species diversity in an annual plant community. Ecology 71:213-225. Gottfried, B. M., and C. F. Thompson. 1978. Experimental analysis of nest predation in an old-field habitat. Auk 95:304-312. Grant, P. R. 1986. Ecology and evolution of Darwin's finches. Princeton University Press, Princeton, New Jersey, USA. 91 Grime, J. P. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. American Naturalist 111:1169-1194. Gruber, K., and A. F. G. Dixon. 1988. The effect of nutrient stress on development and reproduction in an aphid. Entomologica Experimentalis et Applicata 47:23-30. Gunnarsson, B. 1983. Winter mortality of spruce-living spiders: effect of spider interactions and bird predation. Oikos 40:226-233. Hairston, N. G., F. E. Smith, and L. B. Slobodkin. 1960. Community structure, population control, and competition. American Naturalist 24:421-425. Hanazaito, J., and M. Yasuno. 1989. Zooplankton community structure driven by vertebrate and invertebrate predators. Oecologia 81:450-458. Hanhimaki, S. 1989. Induced resistance in mountain birch: defence against leaf-chewing insect guild and herbivore competition. Oecologia 81:242-248. Hanski, I. 1982. Dynamics of regional distribution: the core and satellite species hypothesis. Oikos 38:210-221. Heinrichs, E. A., editor. 1988. Plant stress - insect interactions. John Wiley and Sons, New York, New York, USA. Helle, P., and R. J. Fuller. 1988. Migrant passerine birds in European forest successions in relation to vegetation height and geographical position. Journal of Animal Ecology 57:565-580. Higashiura, Y. 1989. Survival of eggs in the gypsy moth Lymantria dispar. I. Predation by birds. Journal of Animal Ecology 58: 403-412. Hochachka, W. M. 1990. Seasonal decline in reproductive performance of song sparrows. Ecology 21:1279-1288. Hoefs, M. 1973. Birds of the Kluane Game Sanctuary, Yukon Territory, and adjacent areas. Canadian Field-Naturalist 87:345-355. Holbrook, S. J., and R. J. Schmitt. 1989. Resource overlap, prey dynamics, and the strength of competition. Ecology 70:1943-1953. Holmes, R. T., J. C. Schultz, and P. Nothnagle. 1979. Bird predation on forest insects: an exclosure experiment. Science 206:462-463. Huenneke, L. F., S. P. Hamburg, R. Koide, H. A. Mooney, and P. M. Vitousek. 1990. Effects of soil resources on plant invasion and community structure in Californian serpentine grassland. Ecology 71:478-491. Jackson, R. B., J. H. Manwaring, and M. M. Caldwell. 1990. Rapid physiological adjustment of roots to localized soil enrichment. Nature 344:58-60. 92 James, F. C , and S. Rathbun. 1981. Rarefaction, relative abundance, and diversity of avian communities. Auk 98:785-800. Jarvinen, O. 1979. Geographical gradients of stability in European land bird communities. Oecologia 38:51-69. Johnson, E. J., L. B. Best, and P. A. Heagy. 1980. Food sampling biases associated with the "ligature method". Condor 82:186-192. Juliano, S. A. 1986. Food limitation of reproduction and survival for populations of Brachinus (Coleoptera: Carabidae). Ecology 67:1036-1045. Knight, R. L. 1988. Effects of supplemental food on the breeding biology of the black-billed magpie. Condor 90_:956-958. Knisley, C. B., and S. A. Juliano. 1988. Survival, development, and size of larval tiger beetles: effects of food and water. Ecology 69:1983-1992. Kolasa, J. 1989. Ecological systems in hierarchical perspective: breaks in community structure and other consequences. Ecology 70:36-47. Kraft, M. 1988. [Studies of the effects of additional food provision on the breeding density and territorial behaviour of woodland birds.] Verhandlungen der Ornithologischen Gesellschaften Bayern 24:555-590. German, English summary. Krebs, C. J., R. Boonstra, S. Boutin, M. Dale, S. Hannon, A.R.E. Sinclair, and J. N. M. Smith. 1988. Kluane boreal forest ecosystem project: a collaborative special project. Unpublished report to N.S.E.R.C. Krebs, J. R. 1971. Territory and breeding density in the great tit, Parus major L. Ecology 52:2-22. Larsen, J. A. 1980. The boreal ecosystem. Academic Press, New York, New York, USA. Larsson, S. 1989. Stressful times for the plant stress - insect performance hypothesis. Oikos 56:277-283. Lawton, J. H. 1989. Food webs. Pages 43-78 in J. M. Cherrett, editor. Ecological concepts: the contribution of ecology to an understanding of the natural world. 29th Symposium of the British Ecological Society. Blackwell, Oxford, U.K. Leitsma, L. R., R. T. Holmes, and T. W. Sherry. 1990. Effects of removal of red squirrels, Tamiasciurus hudsonicus, and eastern chipmunks, Tamias striatus, on nest predation in a northern hardwood forest: an artificial nest experiment. Oikos 57:375-380. Lenington, S. 1979. Predators and blackbirds: the "uncertainty principle" in field biology. Auk 96:190-192. 93 Levins, R. 1968. Evolution in changing environments. Princeton University Press, Princeton, New Jersey, USA. Loyn, R. H., R. G. Runnalls, G. Y. Forward, and J. Tyers. 1981. Territorial bell miners and other birds affecting populations of insect prey. Science 221:1411-1413. MacArthur, R. TL, and J. MacArthur. 1961. On bird species diversity. Ecology 42:594-598. Martin, T. E. 1987. Food as a limit in breeding birds: a. life-history perspective. Annual Review of Ecology and Systematics 18:453-487. , 1988a. Nest placement: implications for selected life history traits, with special reference to clutch size. American Naturalist 132:900-910. 1 1988b. Processes organizing open-nesting bird assemblages: competition or nest predation? Evolutionary Ecology 2:37-50. May, R. M. 1975. Patterns of species abundance and diversity. Pages 81-120 in M. L. Cody and J. M. Diamond, editors. Ecology and evolution of communities. Harvard University Press, Cambridge, Massachusetts, USA. McGraw, J. B., and F. S. Chapin III. 1989. Competitive ability and adaptation to fertile and infertile soils in two Eriophorum species. Ecology 70:736-749. McLaughlin, R. L., and R. D. Montgomerie. 1989. Brood dispersal and multiple central place foraging by Lapland longspur parents. Behavioural Ecology and Sociobiology 25:207-215. McQueen, D. J., J. R. Post, and E. L. Mills. 1986. Trophic relationships in freshwater pelagic ecosystems. Canadian Journal of Fisheries and Aquatic Science 43:1571-1581. Menge, B. A., and A. M. Olson. 1990. Role of scale and environmental factors in regulation of community structure. Trends in Ecology and Evolution 5:52-57. Moermond, T. C. 1986. A mechanistic approach to the structure of animal communities: Anolis lizards and birds. American Zoologist 26:23-37. Moller, A. P. 1990. Nest predation selects for small nest size in the blackbird. Oikos 57:237-240. Monkkonen, M. 1990. Removal of territory holders causes an influx of small-sized intruders in passerine bird communities in northern Finland. Oikos 57:281-288. Morse, D. H. 1989. American warblers: an ecological and behavioural perspective. Harvard University Press, Cambridge, Massachussetts, USA. 94 Naiman, R. J. 1988. Animal influences on ecosystem dynamics. BioScience 38:750-752. Neu C. W., C. R. Byers, and J. M. Peek. 1974. A technique for analysis of utilization-availability data. Journal of Wildlife Management 38:541-545. Neuvonen, S. 1988. Interactions between geometrid and microtine cycles in northern Fennoscandia. Oikos 51:393-397. Nolan, V., Jr. 1978. The ecology and behaviour of the prairie warbler, Dendroica discolor. Ornithological Monographs 26. O'Connor, R. J. 1986. Patterns of nestling mortality in British passerines. Pages 1457-1466 in Ouellet, H., editor. Acta XIX Congressus International Ornithologici Volume 2. Oksanen, L. 1987. Interspecific competition and the structure of bird guilds in boreal Europe: the importance of doing fieldwork in the right season. Trends in Ecology and Evolution 2:376-379. . 1988. Ecosystem organization: mutualism and cybernetics or plain Darwinian struggle fr existence? American Naturalist 131:424-444. Oksanen, L., S. D. Fretwell, J. Arruda, and P. Niemela. 1981. Exploitation ecosystems in gradients of primary productivity. American Naturalist 118:240-261. Paine, R. T. 1966. Food web complexity and species diversity. American Naturalist 100:65-75. . 1980. Food webs: linkage, interaction strength and community infrastructure. Journal of Animal Ecology 49:667-685. Perrins, C M . , and T. R. Birkhead. 1983. Avian ecology. Blackie, Glasgow, U. K. Pianka, E. R. 1973. The structure of lizard communities. Annual Review of Ecology and Systematics 4:53-74. Pieman, J. 1988. Experimental study of predation on eggs of ground-nesting birds: effects of habitat and nest distribution. Condor 90:124-131. Pimm, S. L. 1982. Food webs. Chapman and Hall, New York, New York, USA. Polley, H. W., and J. K. Detling. 1989. Defoliation, nitrogen, and competition: effects on plant growth and nitrogen nutrition. Ecology 70:721-727. Powell, G. V. N. 1989. On the possible contribution of mixed species flocks to species richness in neotropical avifaunas. Behavioural Ecology and Sociobiology 24:387-393. 95 Prezzler, R. W., and P. W. Price. 1988. Host quality and sawfly populations: a new approach to life table analysis. Ecology 69:2012-2020. Reynolds, R. T., J. M. Scott, and R. A. Nussbaum. 1980. A variable circular-plot method for estimating bird numbers. Condor 82:309-313. Roscoe, J. T., and J. A. Byars. 1971. An investigation of the restraints with respect to sample size commonly imposed on the use of the chi-square statistic. Journal of the American Statistical Association 66:755-759. Rushton, S. P., M. L. Luff, and M. D. Eyre. 1989. Effects of pasture improvement and management on the ground beetle and spider communities of upland grasslands. Journal of Applied Ecology 26j 489-504. Sage, R. D. 1982. Wet and dry-weight estimates of insects and spiders based on length. American Midland Naturalist 108:407-411. Scott, J. M., F. L. Ramsey, and C. B. Kepler. 1981. Distance estimation as a variable in estimating bird numbers. Pages 334-340 in C. J. Ralph and J. M. Scott, editors. Estimating the numbers of terrestrial birds. Studies in avian biology 6. Shipley, B., and P. A. Keddy. 1988. The relation between relative growth rate and sensitivity to nutrient stress in 28 species of emergent macrophytes. Journal of Ecology 76:1101-1110. Simenstad, C. A., J. A. Estes, and K. W. Kenyon. 1978. Aleuts, sea otters and alternate stable state communities. Science 200:403-411. Simons, L. S., and T. E. Martin. 1990. Food limitation of avian reproduction: an experiment with the cactus wren. Ecology 71:869-876. Slansky, F., Jr., and J. G. Rodriguez, editors. 1987. Nutritional ecology of insects, mites, spiders and related invertebrates. Wiley Interscience, New York, New York, USA. Smith, H. R., and N. S. Proctor. 1989. No home to return to. Discovery 21:3-7. Snodgrass, G. L., and E. A. Stadelbacher. 1989. Effect of different grass and legume combinations on spider (Araneae) and ground beetle (Coleoptera: Carabidae) populations in roadside habitats in the Mississippi Delta. Environmental Entomology 18:575-581. Solonen, T. 1986. Stability of a bird community in southern Finland. Ornis Scandinavica 17:106-116 Strong, D. R., D. Simberloff, L. G. Abele, and A. B. Thistle, editors. 1984. Ecological communities: conceptual issues and the evidence. Princeton University Press, Princeton, New Jersey, USA. Theberge, J. B. 1976. Bird populations in the Kluane Mountains, southwest Yukon, with special reference to vegetation and fire. Canadian Journal of Zoology 54:1346-1356. 96 , editor. 1980. Kluane, pinnacle of the Yukon. Doubleday Canada, Toronto, Ontario, Canada. Tomialojc, L., and J. Verner. 1990. Do point counting and spot mapping produce equivalent estimates of bird densities? Auk 107:447-450. Trevelyan, R., and A. F. Read. 1989. Nest predators and the evolution of avian reproductive strategies: a comparison of Australian and New Zealand birds. Oecologia 81:274-278. Tuomi, J., P. Niemela, M. Rousi, S. Siren, and T. Vuorisalo. 1988. Induced accumulation of foliage phenols in mountain birch: branch response to defoliation? American Naturalist 132:602-608. Turner, J. R. G., J. J. Lennon, and J. A. Lawrenson. 1988. British bird species distributions and the energy theory. Nature 335:539-541. Urban, D. L., and T. M. Smith. 1989. Microhabitat pattern and the structure of forest bird communities. American Naturalist 133:811-829. Van Cleve, K., F. S. Chapin III, P. W. Flanagan, L.A. Viereck, and C. T. Dyrness, editors. Forest ecosystems in the Alaskan taiga: a synthesis of structure and function. Ecological studies: analysis and synthesis 57. Springer-Verlag, New York, New York, USA. Verner, J., and T. A. Larson. 1989. Richness of breeding bird species in mixed-conifer forests of the Sierra Nevada, California. Auk 106:447-463. Walter, H., and S. W. Breckle. 1989. Ecological systems of the geobiosphere. Volume 4. Temperate and polar zonobiomes other than northern Eurasia. Springer-Verlag, New York, New York, USA. Waring, G. L., and P. W. Price. 1988. Consequences of host plant chemical and physical variability to an associated herbivore. Ecological Research 3j205-216. Weathers, W. A., and K. A. Sullivan. 1989. Nest attentiveness and egg temperature in the yellow-eyed junco. Condor 91:628-633. Westmoreland, D., and L. B. Best. 1985. The effect of disturbance on mourning dove nesting success. Auk 102:774-780. Whelan, C. J., R. T. Holmes, and H. R. Smith. 1989. Bird predation on gypsy moth (Lepidoptera: Lymanbiidae) larvae: an aviary study. Environmental Entomology 18:43-50. White, T. C. R. 1978. The importance of a relative shortage of food in animal ecology. Oecologia 3:71-86. . 1984. The abundance of invertebrate herbivores in relation to the availability of nitrogen in stressed food plants. Oecologia 63:90-105. 97 Wiens, J. A. 1977. On competition and variable environments. American Scientist 65:590-597. . 1989. The ecology of bird communities. Cambridge University Press, Cambridge, U.K. Wilcove, D. S. 1985. Nest predation in forest tracts and the decline of forest songbirds. Ecology 66:1211-1214. Willis, E. O. 1973. Survival rates for visited and unvisited nests of bicolored antbirds. Auk 90:263-267. Wise, D. H. 1975. Food limitation of the spider Linyphia marginata: experimental field studies. Ecology 56:637-646. Wolda, H. 1988. Insect seasonality: why? Annual Reviews of Ecology and Systematics 19:1-18. Yahner, R. H., T. E. Morrell, and J. S. Rachael. 1989. Effects of edge contrast on depredation of artificial avian nests. Journal of Wildlife Management 53:1135-1138. 98 APPENDIX 1: PASSERINE AND PICIFORM BIRDS OBSERVED IN THE STUDY AREA BETWEEN 1987 AND 1990. Common name Latin name Code Northern flicker Colaptes auratus NOFL Three-toed woodpecker Picoides tridactylus TTWO Black-backed woodpecker Picoides arcticus BBWO Olive-sided flycatcher Contopus borealis OSFL Western wood pewee Contopus sordidulus WWPE Say's phoebe Sayornis saya 0 SAPH Alder flycatcher Empidonax alnorum ALFL Grey jay Perisoreus canadensis GRJA Black-capped chickadee Parus atricapillus BCCH Boreal chickadee Parus hudsonicus BOCH Brown creeper Certhia americana BRCR Red-breasted nuthatch Sitta canadensis RBNU Ruby-crowned kinglet Regulus calendula RCKI Townsend's solitaire Myadestes townsendi TOSO Swains on's thrush Catharus ustulatus SWTH Grey-cheeked thrush Catharus minimus GCTH Hermit thrush Catharus guttatus HETH Varied thrush Ixorius naevius VATH American robin Turdus migratorius AMRO Water pipit Anthus spinoletta WAPI Bohemian waxwing Bombycilla garrulus BOWX Orange-crowned warbler Vermivora celata OCWA 99 Yellow-rumped warbler Dendroica coronata YRWA Blackpoll warbler Dendroica striata BPWA Yellow warbler Dendroica petechia YEWA Wilson's warbler Wilsonia pusilla WIWA Common yellowthroat Geothlypis trichas COYE Savannah sparrow Passerculus sandwichensis SASP American tree sparrow Spizella arborea TRSP Chipping sparrow Spizella passerina CHSP Dark-eyed junco Junco hyemalis DEJU White-crowned sparrow Zonotrichia leucophrys WCSP Rusty blackbird Euphagus carolinus RUBL Brown-headed cowbird Molothrus ater BHCO Pine siskin Carduelis pinus PISI White-winged crossbill Loxia leucoptera WWCR Pine grosbeak Pinicola enucleator PIGR Common redpoll Carduelis flammea CORE 100 APPENDIX 2. HABITAT CHARACTERISTICS OF THE FOUR MAIN STUDY GRIDS. Data were collected by M. Nams. Vertical axes show the proportion of stakes on a grid with the indicated habitat characteristic. The last two figures show slope aspect and steepness. 101 GRID T r e e c o v e r 1.0 0.8 0.6 0.4 H CLOSED 02 m DENSE 0 OPEN • SCATTER • ABSENT 0 0 GRID T r e e s p e c i e s 1.0 0.8 0.6 0.4 -02 • IMMATURE H MATURE oo I' 0 NO-SNAGS _l SNAGS GRID GRID S t a n d m a t u r i t y 102 1.0 0.8 0.6 0.4 0.2 0.0 G R I D S h r u b c o v e r 1.0 0.8 0.6 0.4 02 • A B S E N T Q O P E N S C L O S E D oo I" G R I D • O T H E R 0 B I R C H B W I L L O W S h r u b s p e c i e s 1.0 0.8 0.6 0.4 0.2 0.0 I I I I G R I D M o s s c o v e r 1.0 0.8 0.6 0.4 02 a N O _ M O S S B M O S S o.O • G R I D 0 N O N E B DEADFALL D e a d f a l l 103 • STEEP 0 MODERATE _* NEG GRID 104 APPENDIX 3: FOOD ITEMS RETRIEVED FROM LIGATURED DARK- EYED JUNCO (JUNCO HYEMALIS) NESTLINGS Nest 1 Nestling 1 Snail: 1 x 5mm (shell diameter) Diptera: species A 1 x 10mm; sp.B 1 x 10mm Nestling 2 Spider: 1 x 6mm Diptera: sp.A 1 x 10mm, 1 x 9mm; sp.C 1 x 15mm; other 1 x 3mm Nestling 3 Slug: 1 x 10mm Spiders: 2 x 7mm Diptera: sp.A 2 x 11mm, 1 x 10mm; sp.C 1 x 8mm; others 1 x 5mm, 1 x 4mm, 3 x 3mm Nest 2 Nestling 1 Spider: 1 x 6mm Beetles: 2 x 8mm Aphids: 2 x 5mm, 2 x 3mm, 3 x 2mm, 3 x 1mm 105 Nestling 2 Spider: 1 x 5mm Beetles: 1 x 8mm, 1 x 7mm, 5 x 6mm Aphids: 2 x 4mm, 8 x 3mm, 5 x 2mm Wasp: 1 x 11mm Caterpillars (Lepidoptera): 1 x 14mm, 1 x 12mm, 1 x 9mm, 1 x 8mm Nest 3 Nestling 1 Snail: 1 x 4mm Spider: 1 x 10mm Diptera: sp.B 1 x 11mm; sp.C 1 x 21mm; others 1 x 6mm, 1 x 5mm, 1 x 4mm, 1 x 3mm Dipteran species A was a robber fly (family Asilidae), species B was a muscoid fly (division Schizophora), and species C was a crane fly (family Tipulidae). Only one species of beetle and one species of caterpillar were retrieved.