UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Soil collembola under different conifer species on Vancouver Island, British Columbia Baumbrough, B. 1999

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

Item Metadata

Download

Media
831-ubc_1999-388522.pdf [ 6.69MB ]
Metadata
JSON: 831-1.0088992.json
JSON-LD: 831-1.0088992-ld.json
RDF/XML (Pretty): 831-1.0088992-rdf.xml
RDF/JSON: 831-1.0088992-rdf.json
Turtle: 831-1.0088992-turtle.txt
N-Triples: 831-1.0088992-rdf-ntriples.txt
Original Record: 831-1.0088992-source.json
Full Text
831-1.0088992-fulltext.txt
Citation
831-1.0088992.ris

Full Text

SOIL COLLEMBOLA UNDER DIFFERENT CONIFER SPECIES ON SOUTHERN VANCOUVER ISLAND, BRITISH COLUMBIA by B. BAUMBROUGH B. Sc., The University of British Columbia, 1990 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Soil Science) We accept this thesis as conforming |R5N{he required standard/ THE UNIVERSITY OF BRITISH COLUMBIA 1998 © B. Baumbrough, 1998 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 The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT Three separate but related studies examined collembolan populations under single-species forest stands and how these forest covers might influence feeding in Collembola. In November 1993, a preliminary investigation examined the effects of single-species conifer stands on the density and species diversity of soil collembola on southern Vancouver Island, British Columbia. The top 3 cm of 4.4 cm diameter soil cores, sampled from experimental plots of western redcedar, Douglas-fir, western hemlock, and Sitka spruce, were extracted for collembola using a high gradient extractor. Average linkage cluster analysis of Renkonen's percentage similarity and Morisita's similarity indices did not indicate distinct collembolan communities under each conifer species. However, the density and species richness of collembola under the four conifer species were sufficiently different to imply that further investigation was warranted. A second study was conducted in November 1995 and in May 1996. Soil cores were collected and extracted from two replicate plots of the four conifer species at three different sites located on southern Vancouver Island. Results of this study reflected those ofthe preliminary investigation. Distinct collembolan communities among each ofthe conifer species were not clearly demonstrated using Morisita's similarity index and average linkage clustering. However, significant differences in collembolan density and species richness among conifer species were found using the GLIMMEX procedure in SAS. Collembolan density and species richness was lowest under the western redcedar plots and highest under the Sitka spruce plots. The Modified Simpson's and the Modified Shannon-Wiener indices demonstrated that the structure of the collembolan communities varied only slightly under the different conifer species and was most often characterized by one or two dominant collembolan species and several rare species. Percent moisture content, bulk density and pH were determined from each i i soil core extracted for collembola. Statistical analysis of these data revealed no significant differences among conifer species. In the third investigation, the feeding attributes of collembola species, sampled in the previous investigation from Sitka spruce, Douglas-fir, and western redcedar, were assessed by analysis of the gut contents of mounted specimens. Statistical analysis of the results from five common species found no significant differences in the feeding habits within a collembolan species and between collembolan species among the three conifer species. In November 1993, a preliminary investigation examined the effects of single species conifer stands on the abundance and species diversity of soil collembola on southern Vancouver Island, British Columbia. The top 3 cm of 4.4 cm diameter soil cores, sampled from experimental plots of western redcedar, Douglas-fir, western hemlock, and Sitka spruce, were extracted for collembola using a high gradient extractor. Average linkage cluster analysis of Renkonen's percentage similarity and Morisita's similarity indices did not indicate distinct collembolan communities under each conifer species. However, the density and species richness of collembola under the four conifer species were sufficiently different to imply that further investigation was warranted. A second study was conducted in November 1995 and in May 1996. Soil cores were collected and extracted from two replicate plots of the four conifer species at three different sites located on southern Vancouver Island. Results of this study reflected those of the preliminary investigation. Distinct collembolan communities among each of the conifer species were not clearly demonstrated using Morisita's similarity index and average linkage clustering. However, significant differences in collembolan abundance and species richness among conifer species were found using the GLIMMEX procedure in SAS. Collembolan density and species richness was lowest under the western redcedar plots and highest under the Sitka spruce plots. iii The Modified Simpson's and the Modified Shannon-Wiener indices demonstrated that the structure ofthe collembolan communities varied only slightly under the different conifer species and was most often characterized by one or two dominant collembolan species and several rare species. Percent moisture content, bulk density and pH were determined from each soil core extracted for collembola. Multivariate analysis of this data revealed no significant differences among conifer species. In the third investigation, the feeding attributes of collembola species, sampled in the previous investigation from Sitka spruce, Douglas-fir, and western redcedar, were assessed by analysis ofthe gut contents of mounted specimens. Statistical analysis of the results from five common species found no significant differences in the feeding habits within a collembolan species and between collembolan species among the three conifer species. IV TABLE OF CONTENTS Abstract ii Table of Contents v List of Tables vj List of Figures vii Acknowledgements viii CHAPTER 1: Genera! Introduction 1 CHAPTER 2: Preliminary assessment ofthe density and species diversity of soil Collembola under different conifer species 2.0 Introduction 9 2.1 Study site 10 2.2 Methods 11 2.2.1 Data Analyses 12 2.3 Results 12 2.3.1 Similarity indices and cluster analysis 18 2.4 Discussion 22 CHAPTER 3: Assessment of the density and species diversity of soil Collembola under different conifer species 3.0 Introduction 26 3.1 Study sites 29 3.2 Methods 31 3.2.1 Data Analysis 33 3.3 Results 34 3.3.1 Abundance data 3 4 3.3.2 Species diversity 34 3.3.3 Time of sampling 58 3.3.4 Soil moisture, bulk density, and pH 58 3.4 Discussion 60 CHAPTER 4: Gut content analysis of Collembola collected from different conifer species. 4.0 Introduction 69 4.1 Study Sites 71 4.2 Methods 71 4.2.1 Data analysis 72 4.3 Results 72 4.3.1 Feeding attributes of individual species 77 4.3.2 Tree species effect 85 4.3.3 Time of sampling 86 4.3.4 Parasites 86 4.4 Discussion 91 CHAPTER 5: General Conclusions 98 Literature Cited 100 V LIST OF TABLES Table 2.1. Forest floor collembolan species found in soil core samples taken from 35 year 13 old plantations on south Vancouver Island. Collembola sampled from four conifer species each planted at three different spacings. Table 2.2. Comparison of collembolan mean density (individuals m"2) and diversity 20 under four forest covers. Table 3.1. Analysis of collembolan density data using Log Linear model. 37 Table 3.2. Least square means of the log of the expected count and standard error of 37 collembolan density data. Table 3.3. P-values from pair-wise comparisons of collembolan density under conifer 37 species. Table 3.4. Collembolan species identified from three installations of EP571 in November 38 1995 and May 1996. Table 3.5. Analysis of collembolan species richness data using Log Linear model. 45 Table 3.6. Least square means of the log of the expected count and standard error of 45 collembolan species richness data. Table 3.7. P-values from pair-wise comparisons of collembolan species richness under 45 conifer species. Table 3.8 Comparison of Richness and Evenness of collembolan species collected 46 November 1995. Table 3.9 Comparison of Richness and Evenness of collembolan species collected May 47 1996. Table 3.10. Mean density and percentage of the total abundance of common and/or 53 dominant collembolan species collected in November 1995. Table 3.11. Mean density and percentage of the total abundance of common and/or 55 dominant collembolan species collected in May 1996. Table 3.12. Analysis of moisture content, bulk density, and pH data. 59 Table 3.13. Least square means of moisture content, bulk density, and pH. 59 Table 4.1. Percentage of individuals of all species which contain various gut components. 73 Table 4.2. Analysis of gut content data to test for differences among five 83 collembolan species. Table 4.3. Analysis of collembolan gut content data to test for differences among conifer 87 species. /. notabilis, I. uniens, M. achromata, N. minimus, and T. yosii analysed separately. Table 4.4. Comparison of the quantity of fungal hyphae in gut contents of five species of 88 collembolans collected in November 1995 from three different conifer species. Percentage of individuals with gut contents at least 50% hyaline or darkly pigmented hyphae. Table 4.5. Comparison of the quantity of fungal hyphae in gut contents of five species of 89 collembolans collected in May 1996 from three different conifer species. Percentage of individuals with gut contents at least 50% hyaline or darkly pigmented hyphae. VI LIST OF FIGURES Figure 2.1. Tree diagram resulting from average linkage cluster analysis of Renkonen's 21 index of percent similarity of Collembola species data. Figure 3.1 Map of location of sampling sites. 31 Figure 3.2 Relative mean density (individuals m"2) of Collembola collected from four 36 conifer species at three different installations and two different sampling times. Figure 3.3 Species-area curve for November 1995 sampling. 42 Figure 3.4 Species-area curve for May 1996 sampling. 43 Figure 3.5. Relative species richness (number of species) of Collembola from three 44 different installations and two different sampling times. Figure 3.6. Rank of collembolan species versus density for three conifer species and 48 three installations sampled in November 1995. Figure 3.7. Rank of collembolan species versus density for three conifer species and 50 three installations sampled in May 1996. Figure 3.8 Tree diagram resulting from average linkage cluster analysis of Collembola 57 species data. Figure 4.1. Double cuticle of Folsomia ozeana. 74 Figure 4.2. Fungal spores from gut of Isotoma (Desoria) uniens. 76 Figure 4.3. Relative percentage of individuals of Isotoma (Desoria) notabilis with empty 78 guts and with guts containing at least 50% of various materials. Figure 4.4. Relative percentage of individuals of Isotoma (Desoria) uniens with empty 79 guts and with guts containing at least 50% of various materials. of at least 50% of various materials. Figure 4.5. Relative percentage of individuals of Micrisotoma achromata with empty 80 guts and with guts containing at least 50% of various materials. Figure 4.6. Relative percentage of individuals of Neelus (Megalothorax) minimus with 81 empty guts and with guts containing at least 50% of various materials. Figure 4.7. Relative percentage of individuals of Tullbergia (Tullbergia) yosii with 82 empty guts and with guts containing at least 50% of various materials. Figure 4.8. Gut contents of Willemia denisi. 84 Figure 4.9. "Bucky-ball" structures from Tullbergia (Tullbergia) yosii. 90 vii ACKNOWLEDGEMENTS I would like to extend thanks to my committee members: Dr. Shannon Berch, my thesis supervisor, for her support throughout the years and the place to crash during the final weeks. Dr. Tim Ballard, for his unforgettable impressions of soil fauna and the great fun I experienced as a T.A. in Soils 2 0 0 . Dr. Valin Marshall, for patiently sharing his tremendous knowledge of soil fauna. Dr. John McLean, for his help with statistics and his provocative questions regarding Collembola, nematode feeding and spaghetti noodles. I would also like to thank: Dr. Jan Addison for her technical support. Dr. Derek Harrison for his assistance with my many and varied computer woes. Chio Woon for his help in the lab. Sandy Traichel and Dr. Marcia Monreal for their unwavering friendship. Jude, Mark, Hugh and Kate, for making my seemingly endless days in the "Abysmal Pit of Doom" more than bearable and for Trooper, the best sailboat I have ever had the pleasure to live aboard. Mom and Dad, thanks for hanging in there with me. This work was supported by a grant from Forest Renewal British Columbia. GENERAL INTRODUCTION The Collembola, also known as springtails, are small, wingless arthropods that occur in virtually every terrestrial habitat. They have a distinct head, one pair of antennae, and six abdominal segments. Their exact taxonomic position is the subject of some debate. Currently considered to be a Class of the Phylum Arthropoda, they have been most commonly placed in the class Insecta subclass Apterygota. Two appendages, the collophore and furcula, distinguish Collembola from other small arthropods. The collophore, or ventral tube, consists of eversible sacs derived from a pair of appendages on the first abdominal segment. It plays an important role in fluid balance and may also be used for adherence to slippery surfaces. The furcula, or springing organ, has evolved from the basal fusion of a pair of appendages on the fourth abdominal segment. In species of Collembola confined to the soil, the furcula is reduced or completely absent, thus allowing for easier movement between soil particles. In epedaphic species, a well-developed furcula provides a mechanism to escape from predators. Individuals can use the appendage to propel their bodies a distance many times their own body length in a fraction of a second. Collembola are small in size. Adults range from approximately 0.5 to 8 mm in length. Most soil forms are 1 to 3 mm long as adults. Many species are pigmented and elaborate patterning may occur, particularly in the Entomobryidae and Sminthuridae. Pigment is located in the epidermal cells and can be present in almost any color though blue is most common. The smallest and least-pigmented species are euedaphic forms that dwell permanently in the spaces between soil or sand particles. Reproduction in soil Collembola is achieved through indirect sperm transfer. Sperm are held within a spermatophore, which is deposited on the substrate to be picked up by females. Collembola do not undergo any striking metamorphosis. Development is direct with the 1 juveniles differing from adults only in size, the proportions of organs, pigmentation and the absence of a genital aperture. Numerous studies have been published on the life histories of different collembolan species. Life span, fecundity (number of eggs laid per clutch and number of clutches laid per female), spermatophore production, and number of juvenile and adult instars will vary among species. Temperature, as well as food quality and availability will influence these factors (Booth and Anderson 1979; Choudhuri et al. 1979; Blancquaert et al. 1981; Zettel 1982; van Straalen and Joosse 1985; Walsh and Bolger 1990; Lavy and Verhoef 1996). In temperate regions, the life cycle is usually completed within two to twelve months (Christiansen 1964). Collembolan longevity is uncertain due to the laboratory conditions from which estimates are made, but it appears that most forms can live from four to five months and many can live more than one year (Christiansen 1964). Collembola molt throughout life, continuing even after sexual maturity has been reached, with instars ranging from four to more than 50 (Christiansen 1990). Eggs are laid either singly or in clumps of various sizes throughout the soil and leaf litter. The number of eggs produced in the course of a lifetime varies tremendously from 60 up to 800 (Christiansen 1964). Parthenogenesis has been observed in some collembolan species. The most well-known parthenogenic species is Folsomia Candida Willem, a species that is easily maintained in cultures and therefore often used for laboratory experiments. Many vertebrates including birds, juvenile toads, and small lizards prey upon Collembola. Their main predators however, are other arthropods, including harvestmen (Opiliones), hunting spiders, pseudoscorpions, ants and mites (Hopkin 1997). Many beetles are specialised to prey specifically on Collembola, having evolved trap mechanisms to catch Collembola before they jump, or to prevent their escape (Bauer 1982, 1985). 2 Collembola are found in a variety of habitats. Some are common on the seashore, including the well-known Anurida maritima (Guerin-Meneville), while others, such as Podura aquatica (Linnaeus), live almost permanently on the surface of fresh water. One species, Hypogastrura nivicola (Fitch), referred to as the snow-flea, is sometimes found in very large numbers on the surface of snow. Other species, mainly of the Smmthuridae and Entomobryidae, live on vegetation above ground, most extensively in regions where warm, moist conditions prevail. Some species are found even in extreme climates such as deserts or polar regions (Hopkin 1997). Most collembolan species are restricted to the soil environment where they, along with the mites, make up the majority of the microarthropod fauna. Collembola are extremely abundant in soil and leaf litter. In most terrestrial ecosystems, their densities range from 104 to 105 m"2 (Petersen and Luxton 1982). Knowledge of the Collembola in British Columbia's forest soils is limited to only a few studies. Density estimates from these studies range from 7,000 to 115,000 m"2 (Marshall 1993). Despite their large numbers, the small size of Collembola results in a low contribution (1 to 5%) to the total soil animal biomass and respiration in temperate ecosystems. This increases to about 10% in some arctic sites and to as much as 33% of total soil fauna respiration in ecosystems in the early stages of succession (Petersen 1994). Approximately 6,500 species of Collembola have been described world-wide (Hopkin 1997). The exact number of collembolan species is difficult to ascertain for a number of reasons. Many species have yet to be discovered or described. There may be some synonymy in particular families, which results in a single species being described under more than one name. Also, there is debate among taxonomists as to the degree of morphological difference required between specimens to designate separate species. The collembolan fauna of Canada have never been systematically treated or catalogued. The scattered distributional records and 3 the few taxonomic treatments available have been summarized by Christiansen and Bellinger (1980-1981). Skidmore (1995) lists 412 species of Canadian and Alaskan Collembola distributed in 16 families and 83 genera, 148 of which are known to British Columbia. The estimate for collembolan species known to British Columbia is undoubtedly low, as very few studies of collembolan species diversity have been conducted in this province. Collembola consume a wide variety of food materials. These include decayed and undecayed vegetation, pollen grains, algae, moss, bacteria, arthropod faeces, fungal spores and fungal hyphae (Bodvarsson 1970; Kevan and Kevan 1970; Marshall 1978; Vegter 1983; Walter 1987; Ponge 1991; Hodkinson et al. 1994; Chen et al. 1996; Davidson and Broady 1996). The diet of soil forms consists mainly of decayed plant material, and/or microflora (Christiansen 1990). Most Collembola occasionally consume material derived from animals, including other Collembola. Laboratory experiments have demonstrated extensive feeding on entomopathogenic nematodes by the collembolan species Folsomia Candida and Sinella coeca (Schott) (Gilmore and Potter 1993). Several species have been found on dead animals, either scavenging or grazing the fungal hyphae growing on the corpse (Payne et al. 1968). Among the soil forms, only members of the genus Friesea and Cephalotoma grandiceps (Reuter) have been shown to be carnivorous. Species of Friesea prey on the eggs of other Collembola, tardigrades, Protura, and rotifers, whereas C. grandiceps consumes other Collembola (Christiansen 1964; Hopkin 1997). The foods included in the diet of Collembola are extensive. However, the question of what any given species of Collembola will eat is more difficult to answer. It appears that, in many cases, the diet of an animal is the reflection of what is available in its environment (Bodvarsson 1970; Gilmore and Raffensperger 1970; McMillan and Healey 1971). On the other hand, specialist feeding has been demonstrated in both field and laboratory experiments (Mills and 4 Sinha 1971; Visser and Whittaker 1977; Aitchison 1983; Moore et al. 1987; Klironomos et al. 1992; Hasegawa and Takeda 1995). The problems associated with the assessment of collembolan feeding activity are discussed further in Chapter four. The feeding activity of Collembola may have considerable impact on the processes of decomposition and nutrient cycling. Dead vegetation is consumed and subsequently excreted, partially decomposed, as faecal pellets. This increases the surface area and suitability of the material for microbial and fungal attack (Seastedt 1984; Teuben and Verhoef 1992). Fungal grazing by Collembola often stimulates fungal growth and respiration, though this appears to be dependent on the grazing pressure, the condition of the fungus, and also the collembolan species (Hanlon and Anderson 1979; Hanlon 1981; Bengtsson and Rundgren 1983; Walsh and Bolger 1990; Faber and Verhoef 1991). Selective grazing may alter the competitive relationships and distributions of fungal species, thereby influencing decomposition rates (Parkinson et al. 1979; Newell 1984; Visser 1985). Collembola may also influence mycorrhizal infections. However, their impact appears to depend on experimental conditions, as well as the species involved and their population densities (Finlay 1985; Setala 1995; Hopkin 1997). The Collembola, in light of their relatively high densities in the forest soils of British Columbia and their role in decomposition processes and nutrient cycling, combined with the feasibility of species identification, provide an excellent faunal group to explore the effects of certain forest management practices on soil biodiversity. Our current knowledge of the species richness, species composition, and density of Collembola in British Columbia's forest soils is quite poor. Our understanding of the effects of certain forest management practices on their populations is also quite limited. 5 The approximately 35-year-old plantations of Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco), western hemlock (Tsuga heterophylla [Raf] Sarg.), Sitka spruce (Picea sitchensis [Bong.] Carr), and western redcedar (Thuja plicata Dorm.) of a Ministry of Forests Experimental Project (EP571), established to determine growth parameters, provide an experimental setup for the examination of the effects of single species stands of conifers on collembolan populations. The following study is comprised of three separate, but related, investigations. The study began as a preliminary investigation ofthe effects of conifer species on the density and species diversity of Collembola. Results of the preliminary study implied an effect of conifer species on collembolan populations. This led to a more in-depth investigation into the hypothesis that single species conifer stands support distinct collembolan communities. Collembolan density and species diversity were determined from extracted soil cores collected from replicate plots of western redcedar, Douglas-fir, western hemlock, and Sitka spruce in three different locations on southern Vancouver Island, British Columbia. The resulting data were analyzed to assess differences in collembolan density and richness among conifer species. Several ecological factors, including water content, soil pH, base saturation, Ca, Mg, Mn, and K, and organic matter accumulation have been correlated with collembolan species (Hagvar and Abrahamsen 1984; Pozo 1986; Blair et al. 1994; Geissen et al. 1997). This study measured the moisture content, bulk density, and pH of the soil cores from which the Collembola were extracted. Statistical analysis was used to determine if these factors differed significantly under the different conifer species. Forest floor morphology ofthe soil cores was examined for differences among the conifer species. 6 The study described above found significant differences in collembolan density and diversity under the different conifer species. It was hypothesized that differences in collembolan populations under the different conifer species might be due, in part, to differences in the abundance and diversity of fungi under the different conifer species. Thus the feeding attributes of collembolan species sampled from plots of Sitka spruce, Douglas-fir, and western redcedar in the previous investigation were assessed by analysis of the gut contents of mounted specimens. The results from five common species were analyzed to determine if there were differences in the feeding habits within a collembolan species and between collembolan species among the three conifer species. 7 CHAPTER 2: PRELIMINARY ASSESSMENT OF THE DENSITY AND SPECIES DIVERSITY OF SOIL COLLEMBOLA UNDER DIFFERENT CONIFER SPECIES 8 2.0. INTRODUCTION An integral component of the soil biological community is the soil fauna. Of the soil invertebrates, the Collembola, together with the mites, numerically dominate the soil microarthropod fauna. Several studies indicate that the grazing activities of Collembola significantly affect the rate and process of decomposition (Parkinson et al. 1979, Hanlon and Anderson 1979, Ineson et al. 1982, Faber and Verhoef 1991). Collembola are relatively well-known taxonomically, with an estimated 80% of North American species described and a comprehensive taxonomic key available (Christiansen and Bellinger 1980-81). Though relatively well known taxonomically, our understanding of the biology of these invertebrates is limited. Factors determining the diversity, distribution, and abundance of Collembola remain unclear (Wood 1966, 1967; Curry 1978; Addison 1980; Hagvar 1982; Berg and Pawluk 1984). Perhaps a combination of several factors, including vegetative cover, soil type, moisture, temperature, nutrient status, and others, determine the faunal abundance and diversity of any particular habitat. The relative importance of each of these factors may vary from species to species. A Ministry of Forests experimental project (EP571), located on the west coast of Vancouver Island, provided an excellent opportunity to explore the effects of forest cover on the diversity of soil fauna. Conifer species differ in many ways including foliage quality at litterfall, types and quantities of allelochemicals produced, form and pattern of root development in the soil profile, and the symbiotic fungi with which they are associated (Binkley 1995). By potentially influencing the diversity and abundance of the soil biota, these qualities may, in turn, impact the essential processes of nutrient cycling. The objective of this study was to conduct a preliminary examination ofthe effects of single species conifer stands on collembolan species richness and density and from the results, determine if further investigation would be warranted. The study was conducted in the approximately 35-year old plantations of Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco), western hemlock (Tsuga heterophylla [Raf.] Sarg.), Sitka spruce (Picea sitchensis [Bong.] Carr), and western redcedar {Thuja plicata Donn.) ofthe EP571. 2.1 STUDY SITE Sampling was conducted at the Upper Klanawa Mainline site of the Ministry of Forests Experimental Project (EP571) near Franklin River. The study site is located in the Submontane Very Wet Maritime Coastal Western Hemlock (CWHvml) variant on the windward side of the Vancouver Island Mountains (Klinka et al. 1994). The Coastal Western Hemlock variant is characterized by typically cool summers and mild winters. Mean annual temperature is about 8.2 °C and is above 10°C for approximately 5 months of the year. The mean annual precipitation is 2,787 mm (Meidinger and Pojar 1991). The soils are Ferro-Humic or Humo-Ferric Podzols with soil moisture and nutrient regimes ranging from moist and rich to moist and very rich (Klinka et al. 1994). The study area initially supported old-growth stands of western hemlock, western redcedar, amabilis fir (Abies amabilis [Dougl.] Forbes), and occasional Douglas-fir and Sitka spruce (Omule and Krumlik 1987). It was logged between 1958 and 1960 and slashburned in 1961. In 1962, 81 seedlings were planted in each plot. A plot consisted of 9 rows of 9 trees planted to one of three different square spacings, 9 ft (2.7 m), 12ft (3.7 m), or 15ft (4.6 m) designated as high, medium, and low density, respectively. The study site contains two replicate plots of western redcedar, Douglas-fir, western hemlock, and Sitka spruce planted at each of the three different spacings. 2.2 METHODS In November 1993, three 4.4 cm diameter cores were removed from one plot of each of the four conifer species at each of the three initial spacing levels. The top 3 cm of the cores were extracted for mesofauna using a high gradient extractor (Lussenhop 1971). Soil fauna were extracted into dilute picric acid and then transferred to 75% aqueous alcohol for storage. Soil mesofauna were sorted and counted to the taxonomic level of Order or Family under low magnification. Collembola were further identified to the species level and enumerated. The procedure used to prepare Collembola for species identification was based on methods outlined by Christiansen and Bellinger (1980-1981) and an adaptation of Rusek's (1974) methods provided by V.G. Marshall (pers. comm.). After boiling in 95% ethanol for approximately 5 minutes to destroy the fat bodies, specimens were cleared by immersion in lactic acid. Strongly pigmented individuals were often slightly heated in the lactic acid to promote clearing. Each specimen was mounted dorso-ventrally on a microscope slide using a permanent mounting medium known as PVLG, a polyvinyl alcohol/lactic acid/glycerol combination (Morton 1990). Slides were cured at 40°C for 12 hours or until mountant was dry. Keys used for identification included Christiansen and Bellinger (1980-81) and Fjellberg (1985). Although the species classification system follows Christiansen and Bellinger (1980-1981), to facilitate ease of reading, the use of subgeneric names has been limited to Table 2.1. 2.2.1 Data Analyses Renkonen's percentage similarity index and Morisita's similarity index (Morisita 1959; Wolda 1981; Krebs 1989) were used to assess the similarity of collembolan species diversity among the different conifer stands. The conifer stands collembolan faunas most similar to each other were then grouped together using the average linkage clustering method PROC CLUSTER procedure in SAS (SAS Institute Inc. 1989-1996). 2.3 RESULTS A total of 40 collembolan species were identified from the samples collected for this study and are listed alphabetically, including families and subfamilies (Table 2.1). Some species could not be identified positively and in such cases, the species name is preceded by a question mark in Table 2.1 and in the text. Species that differed slightly from already described taxa are given the designation "nr." before the species name. Ofthe Collembola identified, Anurophorus binoculatus is a new record for Canada (J.A. Addison, pers. comm.; Battigelli and Marshall 1993; Setala and Marshall 1994; Rusek and Marshall 1995; Skidmore 1995). The western redcedar plots had the lowest average number of individuals (approximately 10,000 m~2) and lowest number of species (22). The Sitka spruce plots had the greatest average density of collembolans (about 32,000 individuals m2) and a total of 30 species. The Douglas-fir forest-cover also had a high approximate density (30,000 m"2) and diversity (31) of collembolans that was similar to the Sitka spruce stands while western hemlock supported a moderate density of about 20,000 individuals m"2and a total of 29 species (Table 2.2). 12 CO o * g i 5 o 2 GO co c o <u o Pi P4 ps5 ND ON OO at O C CO s o ON 3 O to CO CD R CO o c 1 O D-O o C O I co a o 5s U 03 c (-1 <u o o 6 o H S CO I* in NO ON U CO •2 J ca •s I ss I CJ O s 0 0 60 !~ <U to R CD cj > CD I R I 9, CD CJ O Pi Pi Pi Pi Pi Pi ON ON 3 ~to^  S R O CD CJ o s u u pi pi Pi u Pi Pi Pi Pi u CO o NO C N OS NO «—1 ON Yos: cott, CO nftj CS s R S • s t " '3 •« - R a c^ a 3 .h CO to a a bo too o o Hyp Hyp CJ .g -4—» C o o CN CJ CO o bD O Q • CJ O s CO 61 CJ O co e CJ C CJ 0 0 T3 f t o ej 6 X! bO St CJ a -a o Pi u u CJ Pi CJ CJ s CD CJ to I o oo ON cj bD .g PQ CO 5 CJ Pi Pi Pi Pi 2 to o la. CJ Pi Pi Pi Pi Pi < Pi 1985 llberg, u ST < « 1 ia) sp. CO 3 -2 s 5 Itchet itchet 2 to I co ON s a is I a s 2 I co I CJ CN co ON K 3^ Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi CJ CCj "S *§ oo ON SD ON o •§ .bo «£, ? a K o a cu 8. 5 a K J3 o oo ON bO CJ PQ •a u c CJ CO CJ co ON 2 o o cj 2 a 2 e r/) 3^ "ob o a 1/1 03 oo g -4-* r/) ol W o (D l i o o a ID e 0 0 O U a '•a o h3 00 o I-T3 s '•a Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi o Pi V Pi CJ Pi u U u u Pi Pi . cj CJ CJ Pi Pi Pi Pi Pi CJ Pi - Pi u CJ CJ Pi Pi Pi Pi Pi Pi Pi o u Pi Pi Pi Pi f4 o oo ON 13 U 3 .g C o o tN 9 e o H O oo a cj o •5 o I a K "§ a CD a ON s u a u 00 NO ON I i o fa a •*^» CD to Ik cj a -cr m ON o >-a e a CD 8 o Jo ON o 3 H C u 00 CJ CO fci o Jo t--ON 00 I i o S •S o 5 o NO ON 0 0 :o3 o oo a O 8. a & o •*^» o u oo .g CQ a <D CO J3 u to K .CD s a it o s a S o CD C3 CJ o K O S o o to s CD ON o cs co ON u 00 .g -s: CJ C3 5$ s o o .to CJ Cd CJ o U O s 6 0 Pi o u o u P4 Pi J3 "5b o u 6 6 0 CJ o CJ < Pi Pi Pi Pi Pi o 9-I t/2 o Its s '•a CJ u CJ CJ Pi Pi Pi Pi Pi Pi CO CJ o  e cj J 3 SB O 6 0 CJ CJ CJ Pi CJ CJ Pi Pi Pi Pi T 3 U i C N <D i—< •8 H o  os e CJ e t CJ OO O as ••o PQ to a 2! o CJ CJ -S I SO Os O c CO S ca cj go cj > CJ CJ 3 3 is. •3 CJ I CJ I O 0 0 OS 1-H CJ 6 0 P3 cj CQ C cj CO CJ SO Os c CJ Si 3 •3 CJ I CO a •2 CJ I o   OS o IS o cj If I CJ so 0 0 6 0 CJ X) to 6 S M 3 3 ¥ .CJ •b "o ? g © If o co a IS. •3 CJ SO o  6 0 I i CJ I co a .CJ •b c? g I cj OS o g o If I CJ T 3 cj 73 c o o cN u co CO "Ho o Q co ed 13 CJ o CJ O o 6 CD J 3 St o cj 0 -a o 4) B 6 0 St o nd CJ s -a st o T3 CJ B 6 0 u Pi Pi u O SO CJ CO £ o to e to so os CJ CO .CJ CO a o CJ R a sS CO Pi Pi u Pi Pi so as CO 8 CD CD Pi Pi Pi Pi Pi Pi Pi SO so OS c u CO J3 u to CD 1 so OS 00 o CO "3 to a . - ^ '5 CD CD Os 00 ••o o 0 0 a I a CJ a & o s CD The Sitka spruce and western hemlock forests had the highest number (3) of unique species. The Douglas-fir and western redcedar plots both supported one unique collembolan species. Of the collembolan species identified, 15 were widespread, being found under at least one of the plots of all four conifer species. Onychiurus flavescens and Neelus minimus were found under all spacings of all conifer species. Hypogastrura virga, Isotoma notabilis, and Sensiphorura marshalli were found under all conifers and all but one planting density. These five species accounted for 49% of the total density of Collembola found under western hemlock (combined data from all three planting densities), 51% under Sitka spruce, 50% under Douglas-fir, and 52% under western redcedar. Isotoma uniens was found under all species and planting densities with the exception of western hemlock and Douglas-fir both planted at a density of 4.6 square metres. All of the remaining widespread species, Paranura colorata, Onychiurus nr. reluctus, Onychiurus similis, Onychiurus nr. sibiricus, Anurophorus binoculatus, Folsomia nr. Stella, Folsomia macroseta, Isotoma ekmani, and Micrisotoma achromata, though represented at least once under each of the conifer species, were missing from three or more planting densities. Eight collembolan species identified in this study were found under only one conifer species. These were Arrhopalites amarus, Tomocerus ? dubius, and Tomocerus flavescens under Sitka spruce, Onychiurus sibiricus, Onychiurus lusus, and Neanura frigida under western hemlock, Microgastrura minutissima under Douglas-fir, and Tullbergia vancouverica under western redcedar. In almost every case, these species were represented by a single specimen at only one planting density. 18 2.3.1 Similarity indices and cluster analysis Renkonen's index measures the percentage similarity between two samples and ranges from 0 (no similarity) to 100 (complete similarity). Values calculated using collembolan species data ranged from 26.99 to 69.54. Average cluster analysis using this index suggested that no correlation existed between Collembola community and conifer species (Figure 2.1). Morisita's index ranges from 0 (no similarity) to 1 (complete similarity). The indices calculated for collembolan species ranged from 0.41 to 0.94. Results of cluster analysis were similar to those obtained using Renkonen's index. 19 Table 2.2. Comparison of collembolan mean density (individuals m"2) and diversity under four different forest covers. Conifer species No. of No. of species No. of individuals (range/plot) unique species Western hemlock planting density: 2.7 x 2.7 m 31126 8-11 2 3.7 x 3.7 m 21264 13 1 4.6 x 4.6 m 8329 3-7 0 average 20240 total spp. 29 Sitka spruce planting density: 2.7 x 2.7 m 24549 6-10 1 3.7 x 3.7 m 42311 9-14 1 4.6 x 4.6 m 28276 9-11 1 average 31712 total spp. 30 Douglas-fir planting density: 2.7 x 2.7 m 21700 8-11 1 3.7 x 3.7 m 24112 8-12 0 4.6 x 4.6 m 44281 9-12 0 average 30031 total spp. 31 Western redcedar planting density: 2.7 x 2.7 m 7890 3-7 0 3.7 x 3.7 m 10519 3-11 0 4.6 x 4.6 m 10520 4-10 1 average 9643 total spp. 22 20 1.2 1.1 0.9 AVERA3E 0.8 • S T / M E 0.7 BETWffiM 0.6 CLUSTERS 0.5 0.4 0.3 0.2 0.1 0 Figure 2.1. Tree diagram resulting from average linkage cluster analysis of Renkonen's index of percent similarity of Collembola species data. HW = western hemlock; FD = Douglas-fir; SS = Sitka spruce; CW = western redcedar. Numbers 9, 12, and 15 represent conifer planting spacings: 9, 12, and 15 feet, respectively, or 2.7, 3.7 and 4.6 metres, respectively. W OJ ^ 9 1 X LL I PLOT 2.4 DISCUSSION The results of the cluster analysis using the Renkonen's and Morisita's similarity indices for collembolan species data suggest that forest cover has little effect on collembolan populations. However, the consistently lower overall density and number of species, found under western redcedar suggests this conifer species may, in some way, be influencing collembolan populations. It should also be noted that there were no "abundant" species found under western redcedar. Unfortunately, an analysis of variance, to determine if collembolan populations were significantly different under western redcedar, could not be run with the limited data set. Results from previous studies of collembolan species relations to habitat or vegetation type are inconclusive. While several studies found no direct relationship between plant species and collembolan species (Wood 1967; Curry 1978; Addison 1980), there remains some indication that several factors, including vegetation type, soil type, moisture, temperature, locality, and plant growth forms, contribute to the over-all composition of collembolan fauna (Wood 1966; Addison 1980; Hagvar 1982; Berg and Pawluk 1984). Only a small number of the collembolan species identified in this study were widespread and common to all plots, but they accounted for the near majority of the total abundance. This finding is not atypical of other collembolan population studies (Milne 1962; Hale 1966; Usher 1970; Niijima 1971; Hagvar 1982; Setala and Marshall 1994). Hale (1966) found, in a population study of moorland Collembola, that the rare species were the better indicators of variations between habitats than the more common species. Unfortunately, these rare species tend to occur in insufficient numbers to allow for detailed analysis of their distribution. The widespread collembolan species from this study are well-known and common species for British Columbia (Battigelli and Marshall 1993; Setala and Marshall 1994; Rusek and Marshall 1995). Several investigations on the horizontal distribution of Collembola have shown that they tend to be aggregated in the soil (Poole 1961; Milne 1962; Hale 1966; Joosse 1970). This aggregation phenomenon was observed in this study. For example, in one of the western hemlock plots, one of the three cores contained a substantially higher number of Collembola (100 individuals) relative to the remaining two cores (15 and 27 individuals). Similar results were observed in one of three cores from a Douglas-fir plot. Field and experimental studies have shown several factors to be correlated with this non-random distribution. These include depth and moisture content of the organic layer (Poole 1961, 1962), species-specific humidity preferences (Joosse 1970), presence of food (Christiansen 1970; Barra and Christiansen 1975; Usher and Hider 1975), and the production of pheromones (Verhoef et al. 1977). One ofthe obvious limitations of this study is the one-time nature of the sampling. Several studies indicate that collembolan populations vary seasonally (Hale 1966, Usher 1970; Niijima 1971). The findings of Hale (1966) suggest that the timing of population peaks is species-specific. While some species exhibit only one annual population peak, others have two peaks, either in the spring and autumn or, in some cases, the summer and early winter. Where data exist, these peaks appear to be correlated with periods of egg laying. 23 The factors determining the distribution of microarthropods remain largely unknown and some authors argue they are unlikely to be reflected by descriptions of soil or vegetation type. Nonetheless, despite its limited scope, results of this preliminary study suggest the hypothesis that forest cover influences collembolan populations. Collembolan density and diversity was noticeably less under western redcedar compared to the other forest types sampled. It was therefore decided to expand on the preliminary study so as to include more than one sampling site and more than one sampling time. The results of this subsequent study are discussed in the following chapter. 24 CHAPTER 3: ASSESSMENT OF THE DENSITY AND SPECIES DIVERSITY OF SOIL COLLEMBOLA UNDER DIFFERENT CONIFER SPECIES 25 3.0 INTRODUCTION Collembola are a dominant element of the soil fauna in coniferous forests. Their function in the soil ecosystem is not yet fully understood, though it appears their grazing activities impact the rate and process of decomposition (Parkinson et al. 1979; Visser 1985; Lussenhop 1992). Knowledge of the factors affecting the distribution, density and species diversity of forest soil Collembola in British Columbia is limited. Vlug and Borden (1973) examined the effects of logging and slash burning on soil Collembola populations in a coastal British Columbia coniferous forest of western redcedar (Thuja plicata Donn), yellow cedar (Chamaecyparis nootkatensis [D. Don] Spach), and western hemlock (Tsuga heterophylla [Raf.] Sarg.). The authors sampled from adjacent areas, each treated differently - forested, logged, and slash burned. They found that collembolan populations near the soil surface were significantly reduced in the recently logged area and further reduced in the slash burned area. In the lower depths of the soil, population densities decreased with the severity of treatment in all groups except for the Istomidae, in which an increase in population density was observed at depths of about 2.5 to 10 cm. The decrease may be due to the combined result of mortality at the surface layers and a downward migration towards more favorable conditions. No significant correlation was found between collembolan population density and soil moisture, pH, or temperature. This suggests that other unmeasured factors, such as food supply, may be responsible for regulating population density at these sites. A Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco) plantation on southern Vancouver Island was the site of an investigation into the short-term effects of urea on the distribution of soil fauna (Marshall 1974). The plantation was thinned and urea was applied at 0, 224, and 448 kg N/ha just prior to sampling. Monthly sampling over the period of a year revealed significant seasonal changes in the proportion of Collembola at different depths. Urea 26 application appeared to accentuate the seasonal downward distribution. The author surmised that, at this site, temperature was more important than soil moisture in influencing collembolan distribution, as there was no marked downward movement during the dry summer periods. Seasonal fluctuations in population density appeared to be affected by urea application. While Collembola in the control plot exhibited a population peak in August, this summer peak was delayed to September in the treatment plots. Fertilizer treatment did not significantly affect the mean annual density of Collembola. In a more recent study located on northern Vancouver Island, Battigelli et al. (1994) considered the soil fauna communities of two distinct but adjacent forest types, old growth western redcedar (Thuja plicata Donn)-western hemlock (Tsuga heterophylla [Raf] Sarg.) and mature western hemlock-amabilis fir (Abies amabilis [Dougl.] Forbes). In this study, sampling took place on five occasions over a two-year period. As in the Marshall (1974) study, seasonal changes in the vertical distribution of Collembola were observed, with the percentage of individuals in the LF horizon lower in May and August compared with the March, July, and October samplings. Seasonal fluctuations in population density showed a population peak in August. Though statistical analyses were not run on the data, there appeared to be differences in collembolan density between the two forest types. The above studies provide no information towards the understanding of the factors affecting collembolan community structure, as they did not identify Collembola to the species level. However, Rusek and Marshall (1995) investigated the long-term changes in collembolan communities in three coniferous forests on southern Vancouver Island, including old-growth western hemlock and Douglas-fir forests and a Douglas-fir plantation. Changes in the collembolan communities of these permanent forest plots, following a time period of 17 or 19 years, were described on the basis of several community parameters including, for example, 27 density, dominance, constancy, species number, and evenness. Ordination and cluster analyses were used to compare the data from each site and different collecting dates. Ordination analysis demonstrated clear differences in the collembolan community assemblages among all three sites. The climax forests of western hemlock and Douglas-fir supported similar communities over time, whereas the communities found in the Douglas-fir plantation site were quite different between sampling times. The authors suggest that the extensive long-term changes observed in the plantation site were attributable to succession. After 44 years of plantation growth, this forest had yet to establish a well-developed community. Though collembolan community changes in the climax forests were less drastic, the Douglas-fir stand showed a substantial change in the dominant species composition. The majority of these dominant species belonged to the ecomorphological group of epigeic species. The authors suggest that, since this forest is virtually surrounded by municipalities of Victoria, wind-carried pollutants may be affecting certain collembolan species, thus providing a possible explanation for the observed changes in dominant species composition. In the previous chapter, sampling of soil Collembola was conducted in experimental plots comprised of single species plantations of western redcedar, Douglas-fir, western hemlock, and Sitka spruce (Picea sitchensis [Bong.] Carr) located on the west coast of Vancouver Island. Sampling was limited to one site and one sampling time as the study was intended as a preliminary investigation. Although Renkonen and Morisita similarity indices did not indicate a correlation between collembolan communities and conifer species, the density and species diversity of Collembola was notably less under western redcedar relative to the other conifer species, suggesting that conifer species may have an effect on collembolan populations. 28 The main objective of the present study was to further examine the effects of conifer species on soil Collembola. Collembolan density and species diversity were determined from extracted soil cores collected from replicate plots of western redcedar, Douglas-fir, western hemlock, and Sitka spruce and the resulting data analyzed for significant differences among conifer species. Seasonal fluctuations in collembolan populations are well-documented (Usher 1970; Niijima 1971). Therefore, each plot was sampled in the spring and fall in order to compensate for different life cycle patterns and to obtain a better sense of the species composition of each habitat. Only the top 3 cm of each core was extracted for soil fauna. This may result in the exclusion of certain collembolan species found lower in the soil profile, yet should successfully sample the majority of species (Marshall 1974). Several ecological factors, including water content, soil pH, base saturation, Ca, Mg, Mn, and K, and organic matter accumulation have been correlated with collembolan species (Hagvar and Abrahamsen 1984; Pozo 1986; Blair et al. 1994; Geissen et al. 1997). This study measured the moisture content, bulk density, and pH of the soil cores from which the Collembola were extracted. Statistical analysis was used to determine if these factors differed significantly under the different conifer species. 3.1 STUDY SITES Sampling was conducted on the west coast of Vancouver Island, British Columbia. The study area consists of three installations (sites). The UK Mainline and Branch 167 sites are located near Franklin River (latitude 48° 50-54'N; longitude 124°46-54'W) and a third site, Fairy Lake, is located near Port Renfrew (latitude 48°33-36'N; longitude 124° 19-21'W) (Figure 3.1). The study area is located in the Submontane Very Wet Maritime Coastal Western Hemlock (CWHvml) variant on the windward side of the Vancouver Island Mountains (Klinka 29 et al. 1994). Typically cool summers and mild winters characterize the Coastal Western Hemlock variant. Mean annual temperature is about 8.2°C and is above 10°C for approximately five months of the year. The mean annual precipitation is 2,787 mm (Meidinger and Pojar 1991). The soils are Ferro-Humic or Humo-Ferric Podzols. Soil moisture and nutrient regimes vary among sites. The soils ofthe Fairy Lake installation range from slightly dry and poor to fresh and medium. Fresh and medium soils are found at the Branch 167 installation while UK Mainline soils range from moist and rich to moist and very rich (Klinka et al. 1994). The study area initially supported old-growth stands of western hemlock (Tsuga heterophylla [Raf] Sarg.), western redcedar (Thuja plicata Donn.), amabilis fir (Abies amabilis [Dougl.] Forbes), and occasional Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco) and Sitka spruce (Picea sitchensis [Bong.] Carr) (Omule and Krumlik 1987). The site was logged between 1958 and 1960 and slashburned in 1961. In 1962, plots of 81 seedlings were established. A plot consisted of 9 rows of 9 trees planted to one of three different square spacings: 9 ft (2.7 m), 12 ft (3.7 m), or 15 ft (4.6 m). Each installation contains two replicate plots of western redcedar, Douglas-fir, western hemlock, and Sitka spruce planted at each ofthe three different spacings. 30 Figure 3.1. Dots represent locations of sampling sites near Lake Cowichan and Port Renfrew. Reprinted from Ecosystems of British Columbia, Meidinger and Pojar 1991. 3.2 METHODS Sampling was carried out in November 1995 and in May 1996. At each sampling time, four 4.4 cm diameter cores were removed from each of two replicate plots of western redcedar, western hemlock, Douglas-fir, and Sitka spruce planted at a spacing of 2.7 square meters. The top 3 cm of each core sample was extracted for mesofauna over a period of seven days using a high gradient extractor (Lussenhop 1971). Soil fauna were extracted into dilute picric acid and then transferred into 75% alcohol for storage. Extraction efficiency of the high gradient extractor used in this study was determined by using a heptane flotation method (Walter et al. 1987) followed by hand sorting of the soil from the heptane extraction as a double check. The extraction efficiency for forest floor samples was found to be 92% while the efficiency for mineral soil was slightly higher at 95% (V.G. Marshall, pers. comm.). 31 The Collembola from all extracted samples were identified to the taxonomic level of Family and counted. Counts of the number of individuals collected from each core were used to calculate the average density (individuals m"2) of Collembola for the various conifer species. Due to time limitations, the number of samples from which Collembola were identified to species was reduced to two ofthe four cores collected from each plot of western redcedar, Douglas-fir and Sitka spruce. Western hemlock was the conifer species chosen for exclusion for two reasons. One, there were difficulties with these plots at the Fairy Lake site as one plot could not be found at the desired spacing (the next closest spacing was sampled). Secondly, a concurrent study of the macrofiingi, a potential source of data useful to this study, included only western redcedar, Douglas-fir, and Sitka spruce plots. Identification of Collembola to Family was carried out at low magnification, the specimens remaining in alcohol. Preparation for species identification included clearing the specimens in lactic acid. Some specimens were slightly warmed in lactic acid to improve clearing. Cleared specimens were then mounted on microscope slides using PVLG (Morton 1990) as a permanent mounting medium and the slides cured at 40°C for 12 hours or until mountant was dry. Keys used for identification included Christiansen and Bellinger (1980-81) and Fjellberg (1985). Species names used in this study follow Christiansen and Bellinger (1980-1981) though the use of subgeneric names is limited to Table 3.4. Sampling for the determination of certain soil properties was limited to the 3 cm core samples extracted for soil fauna. The wet weight of each core was determined before extraction of soil fauna. Following extraction, cores were dried at 70°C for at least 24 hours and weighed so that percent moisture content and bulk density could be calculated. Percent moisture content was calculated as the difference between the wet and dry weights of the soil sample divided by the dry weight of the sample, multiplied by 100%. Bulk density was calculated as the weight of 32 the oven-dried soil divided by the volume of the core sample. Soil pH of each core was measured in a water suspension according to the method outlined in Kalra and Maynard (1991). 3.2.1 Data Analysis Significant differences in collembolan density and species richness among conifer species were assessed using analysis based on a Log Linear model with a Poisson error structure. The GLIMMEX procedure in SAS was used to complete the analysis (SAS Institute Inc. 1989-1996). The accepted level of significance was set at a p-value of 0.05, except in the case of pair-wise comparisons where a Bonferroni corrected cc-level was used. This method of data analysis was chosen because the data were not normally distributed, an important assumption of the more widely used ANOVA method. Morisita's similarity index (Morisita 1959; Wolda 1981; Krebs 1989) was used to assess the similarity of collembolan communities among different conifer stands. Conifer stands most similar to each other were then grouped together using the average linkage clustering method PROC CLUSTER procedure in SAS (SAS Institute Inc. 1989-1996). Dominant collembolan species were determined for each community (i.e. each tree species at all three sites) by expressing the density of a collembolan species as a percentage of the mean density of all Collembola of that community. Two indices, the Modified Simpson's and the Modified Shannon-Wiener, were used to assess the species evenness among conifer species (Routledge 1979). Significant differences in moisture content, bulk density and soil pH among conifer species were assessed using PROC GLM in SAS (SAS Institute Inc. 1989-1996). 33 3.3 RESULTS 3.3.1 Abundance data Collembolan densities ranged from as low as 2,700 Collembola m"2 to as high as 55,000 Collembola m"2 (Figure 3.2). The western redcedar plots consistently supported the lowest density (ranging from 2,700 to 13,000 m"2) while the Sitka spruce plots supported the greatest collembolan density (18,000-55,000 m"2). The remaining two conifer species supported intermediate densities ranging from 3,000-21,000 m"2 (Douglas-fir) to 6,800-35,000 m"2 (western hemlock). Collembolan density was significantly different among conifer species at alpha level of p=0.05. (Tables 3.1 and 3.2). Pair-wise comparisons of conifer species found collembolan density under western redcedar (Cw) to be significantly lower than western hemlock (Hw) and Sitka-spruce (Ss). In addition, density of Collembola under Douglas-fir (Fd) was significantly lower than Sitka spruce (Table 3.3). 3.3.2 Species diversity A total of 62 collembolan species were identified from all samples (Table 3.4). Since the collembolan communities of this study contained a relatively large proportion of rare species, and because time constraints imposed a low amount of sampling, it is likely that further sampling would have revealed more species. Species-area curves, which plot the cumulative number of species against the sample number, provide a sense of the sampling required for obtaining most of the species at a site. The majority of species-area curves for the conifer species in this study (Figures 3.3 and 3.4) did not level off, suggesting that increased sampling would have unveiled more species. 34 Four species, Harlomillsia oculata, Odontella Stella, Onychiurus lusus, and Arrhopalites diversus are new records for British Columbia and Canada. Three species, Hypogastrura palustris, Isotomiella minor, and Arrhopalites amarus are new records for British Columbia (Battigelli and Marshall 1993; Setala and Marshall 1994; Rusek and Marshall 1995; Skidmore 1995) (Table 3.4). Figure 3.5 illustrates the variation in species richness (number of species) under the different conifer stands. At each installation, the Sitka spruce stands supported the greatest number of species (ranging from an average of 11 to 17) and the western redcedar stands (with one exception) supported the lowest richness, the average number of species ranging from 5 to 9. The difference in species richness between these two conifer species was significant at p=0.01 (Tables 3.5, 3.6, and 3.7). The spruce plots supported the highest number of unique collembolan species in both November (15) and May (12). Douglas-fir supported far fewer unique species (1 and 4, respectively) while western redcedar plots supported 3 unique species in November and 5 unique species in May (Table 3.4). The Modified Simpson's and the Modified Shannon-Wiener indices were used to assess the evenness of species distribution under each plot in November (Table 3.8) and in May (Table 3.9). When all species are equally abundant, the evenness measurement is equivalent to the species count (richness). On the other hand, a distribution dominated by a few abundant species will result in a low evenness measurement. It can be seen that the structure of the collembolan communities varied only slightly under the different conifer species. A high level of dominance most often characterized the communities with one species comprising between 26-53% of the total abundance, an additional few species found at an intermediate abundance, and many species collected as a single specimen. In a few cases, the dominance of one species would be less acute and two to five collembolan species would account for 43-52% of the total abundance (Figures 3.6 and 3.7). 35 60000 -, 50000 40000 A S § 30000 co ^ 20000 10000 • CwNov • Fd Nov • HwNov 0 Ss Nov B Cw May | n Fd May M Hw May | ES Ss May Fairy Lake Branch 167 Uk Mainline Installation Figure 3.2. Relative mean density (individuals m"2) of Collembola collected from four conifer species at three different installations and two different sampling times. Cw = western redcedar; Fd = Douglas-fir; Hw = western hemlock; Ss = Sitka spruce. 36 Table 3.1. Analysis of collembolan density data using Log Linear model. Source NumDF DenDF P value Date 1 2 0.2534 Tree species 3 6 0.0014 Date*species 3 6 0.6066 Table 3.2. Least square means ofthe log of the expected count and standard error of collembolan density data. Level LS Mean SE Date 1 3.169a 0.273 Date 2 3.410a 0.270 Species Cw 2.611c 0.293 Species Fd 3.069bc 0.283 Species Hw 3.574ab 0.276 Species Ss 3.905a 0.273 * Values followed by the same letter are not significantly different (P < 0.01). Table 3.3. P-values from pair-wise comparisons of collembolan density under conifer species. Cw Fd Hw Ss Cw 0.057 .002 .0004 Fd - .024 .002 Hw - - .068 Ss - - -37 Table 3.4. Collembolan species identified from three installations of EP571 in November 1995 and May 1996. Fairy Branch UK Lake 167 Main Family ENTOMOBRYIDAE Cw Fd Ss Cw Fd Ss Cw Fd Ss Subfamily Entomobryinae Entomobrya (Entomobrya) confusa nov R Christiansen 1958 may Sinella nr. sexoculata (Schott, 1896) nov R may Subfamily Oncopodurinae ** Harlomillsia oculata (Mills, 1937) nov R R C may C Subfamily Tomocerinae Tomocerus (Pogonognathellus) dubius nov Christiansen, 1965 may R R R R Tomocerus (Pogonognathellus) nov R R R flavescens Tullberg, 1871 may R R R Tomocerus (Tomolonus) reductus nov R R R (Mills, 1949; may R R R Family HYPOGASTRURIDAE Subfamily Hypogastrurinae Friesea sp. A nov R may Friesea cera nov R Christiansen and Bellinger, 1974 may Hypogastrura (Ceratophysella) sp. A nov may R C R Hypogastrura (Ceratophysella) nov R palustris (Martynova, 1978) may R R Hypogastrura (Mitchellania) horrida nov R Yosii, 1960 may Hypogastrura (Mitchellania) krafti nov R R R (Scott, 1962; fypogastrura ( Christiansen and Bellinger, 1980 fypogastrura (Mi Fjellberg, 1985 may R H t  (Mitchellania) virga nov R R may R R R H  ( itchellania) wallmoi nov R R may 38 Table 3.4 continued. Willemia denisi Mills, 1932 Wdlemia sp. A Subfamily Neanurinae Anurida (Micranurida) pygmaea (Borner, 1901) Anurida (Micranurida) spirillifera (Hammer, 1953) Neanura (Christobella) ornata (Folsom, 1902) Neanura setosa (Canby, 1926) Odontella (Odontella) biloba Christiansen and Bellinger, 1980 Odontella (Odontella) shasta Christiansen and Bellinger, 1980 Odontella (Odontella) nr. shasta Christiansen and Bellinger, 1980 Odontella substriata Wray, 1952 Paranura colorata Mills, 1934 Fairy Lake Branch 167 UK Main Cw Fd Ss Cw Fd Ss Cw Fd Ss nov may nov may may R nov R may nov may nov may nov may nov may nov may R nov ** Odontella (Odontella) Stella Christiansen and Bellinger, 1980 may nov may nov R may R Pseudachorutes lunatus Folsom, 1916 nov may Pseudachorutes (Pseudachorutes) sp. A nov may Family ISOTOMIDAE Anurophorus (Pseudanurophorus) nov binoculatus (Kseneman, 1934) may Cryptopygus sp. A nov may Cryptopygus sp. B nov may R R R nov R R R R R C R R R R A R R R R R R R R C R R R R R R C R R R R R C R R R R R R R R R R 39 Table 3.4 continued. Fairy Lake Branch 167 UK Main Folsomia Candida Willem, 1902 Folsomia macroseta Ford, 1962 Folsomia ozeana Yosii, 1954 Folsomia nr. Stella Christiansen and Tucker, 1977 Isotoma (Desoria) ekmani Fjellberg, 1977 Isotoma (Desoria) notabilis Schaffer, 1896 Isotoma (Desoria) uniens Christiansen and Bellinger, 1980 Isotoma (Pseudisotoma) monochaeta Kos, 1942 * Isotomiella minor Schaffer, 1896 Micrisotoma achromata Bellinger, 1952 Family NEELIDAE Neelus (Megalothorax) minimus (Willem, 1900) Family ONYCHIURIDAE Lophognathella choreutes Borner, 1908 Onychiurus (Onychiurus) eisi Rusek, 1976 Onychiurus (Onychiurus) flavescens Kinoshita, 1916 -** Onychiurus (Onychiurus) lusus Christiansen and Bellinger, 1980 Onychiurus (Onychiurus) nr. reluctus Christiansen, 1961 Cw Fd Ss Cw Fd Ss Cw Fd Ss nov may nov may nov R may C nov may nov R may R nov C may R nov R may nov may nov may nov R may R nov R may nov R may R nov may nov may nov may nov may R R R R R A R R C R R C C R C C C C R C C R C R C R C R C C R R R C R R R R R R R C R R R R R R R C C C C R R R R R C R R R C R C R R R R C C R R R R R R R R R R R R R R R R R C R R R R R C R R C C R R R R C R C R R C C R C R R C R R R R 40 Table 3.4 continued. Fairy Branch UK Lake 167 Main Cw Fd Ss Cw Fd Ss Cw Fd Ss Onychiurus (Protaphorura) cocklei nov R (Folsom, 1908) may R Onychiurus (Protaphorura) sibiricus nov R (Tullberg, 1876) may R Onychiurus (Protaphorura) nr. sibiricus nov R R R (Tullberg, 1876) may R Onychiurus (Protaphorura) similis nov R C (Folsom, 1917) may C R Sensiphorura marshalli Rusek, 1976 nov R R R R R R may R C C C C C Tullbergia (Tullbergia) macrochaeta nov R R C (Rusek, 1976) may Tullbergia (Tullbergia) ruseki nov Christiansen and Bellinger, 1980 may C R Tullbergia (Tullbergia) vancouverica nov R R (Rusek, 1976) may R Tullbergia (Tullbergia) yosii n o v R C R R R R R C (Rusek, 1967) may R R C R Family SMINTHURTDAE * Arrhopalites amarus Christiansen 1966 nov may R R R Arrhopalites clarus Christiansen 1966 nov R may ** Arrhopalites diversus Mills, 1934 nov may R Dicyrtoma (Ptenothrix) maculosa nov (Schott, 1891) minthurinus (Sn quadrimaculatus (Ryder, 1879) minthurides (Sphaen (Krausbauer, 1898) may R R R S i t i  minthurinus) nov may R S ridia) pumilis nov may R R_ 'R' = rare: average number of specimens in four cores was < 2; 'C = common: average number of specimens in four cores was 2-10; 'A' = abundant: average number of specimens in four cores was > 10 * denotes new record for British Columbia; ** denotes new record for Canada 41 Figure 3.3. Species-area curves for November 1995 sampling: Cw = western redcedar; Fd = Douglas-fir; Hw = western hemlock; Ss = Sitka spruce; UK = UK Mainline; FL = Fairy Lake; BR = Branch 167. 42 25 -r 1 2 3 4 Number of samples Figure 3.4. Species-area curves for May 1996 sampling: Cw = western redcedar; Fd = Douglas-fir; Hw = western hemlock; Ss = Sitka spruce; UK = UK Mainline; FL = Fairy Lake; BR = Branch 167. .8 -> Fairy Lake Branch 167 Installation • CwNov • Fd Nov 0 Ss Nov B Cw May ID Fd May H Ss May UK Mainline Figure 3.5. Relative species richness (number of species) of Collembola from three conifer species at three different installations and two different sampling times. Cw = western redcedar; Fd = Douglas-fir; Ss = Sitka spruce. 44 Table 3.5. Analysis of collembolan species richness data using Log Linear model. Source Num DF Den DF P value Date I 2 0.2776 Tree species 2 4 0.0191 Date* species 2 4 0.4915 Table 3.6. Least square means of the log of the expected count and standard error of collembolan species richness data. Level LS Mean SE Date 1 2.020a 0.134 Date 2 1.761a 0.139 Species Cw 1.527b 0.141 Species Fd 1.900ab 0.135 Species Ss 2.243a 0.124 * Values followed by the same letter are not significantly different (P < 0.01). Table 3.7. P-values from pair-wise comparisons of collembolan species richness under conifer species. Cw Fd Ss Cw - 0.071 0.008 Fd - - 0.066 Ss - - -45 Table 3.8. Comparison of Richness and Evenness of collembolan species collected November 1995. Plot Richness Modified Simpson's Modified Shannon-Index Wiener Index Fairy Lake cw23 12 7.6 9.1 cw32 5 3.9 4.3 fdl8 14 6.5 8.8 fd25 9 5.2 6.6 ss36 16 8.5 11.2 ss38 17 1.9 3.7 UK Main cw84 10 4.1 6.2 cwlOO 5 3.5 4.1 fd98 9 4.2 5.9 fdlOl 7 3.3 4.4 ss99 13 9.3 11 ssl02 8 5.9 6.5 Branch 167 cwll6 6 5.4 5.7 cwl23 4 3.0 3.5 fdl26 6 3.5 4.4 fdl27 1 1 1 ssll9 12 6.9 8.6 ssl24 11 6.1 7.9 Table 3.9. Comparison of Richness and Evenness of collembolan species collected May 1996. Plot Richness Modified Simpson's Modified Shannon-Index Wiener Index Fairy Lake cw23 5 3.9 4.3 cw32 8 3.7 5.4 fdl8 16 8.8 11.2 fd25 11 5.7 7.3 ss36 15 4.7 7.7 ss38 16 6.7 9.6 UK Main cw84 9 6.8 7.7 cwlOO 10 5.3 7.0 fd98 12 8.3 9.8 fdlOl 12 7.0 8.6 ss99 22 10.2 13.5 ssl02 12 5.4 7.7 Branch 167 cwll6 7 4.8 5.7 cwl23 5 2.2 3.1 fdl26 8 5.1 6.2 fdl27 12 3.7 6.0 ssll9 19 10.0 12.6 ssl24 17 5.9 9.5 Western redcedar 7000 -, 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Rank Douglas-fir 12000 n 1 2 3 4 5 6 7 8 9 10 11 12 .13 14 15 16 Rank Figure 3.6. Rank of collembolan species versus density for three conifer species and three installations sampled in November 1995. 48 Sitka spruce 35000 -, 1 3 5 7 9 11 13 15 17 19 21 23 Rank Figure 3.6 continued. 49 Western redcedar 4500 0 -| 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Rank Douglas-fir 10000 n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Rank Figure 3.7. Rank of collembolan species versus density for three conifer species and three installations sampled in May 1996. 50 Sitka spruce 20000 -, 1 3 5 7 9 11 13 15 17 19 21 23 Rank Figure 3.7 continued. 51 Though the structure of the community changed little, the dominant collembolan species varied greatly among all plots sampled in November (Table 3.10) and May (Table 3.11). Tables 3.10 and 3.11 show that common collembolan species (those found frequently under all or most conifer species) were not necessarily the dominant species (collected in large numbers relative to all other collembolan species). The results of average linkage cluster analysis of similarity values using Morisita's similarity index illustrate the variation in communities among conifer species (Figure 3.8). Morisita's similarity index ranges from 0 (no similarity) to 1 (complete similarity). Index values calculated using collembolan species data ranged from 0.00 to 0.92. Though no clear pattern of clustering emerged, there was some grouping according to conifer species. This clustering can most often be attributed to the sharing of one or two dominant species, which occurred most often among spruce plots. Five spruce plots, representing all three installations and both sampling times, were grouped together. In addition to this, there were two examples of a spruce plot, sampled in November, being most similar to the same plot sampled in May. This means that only 3 of the total 12 spruce plots sampled were not most similar to another spruce plot. The Douglas-fir plots were less similar to each other though some grouping of plots did occur. Virtually no grouping of western redcedar plots occurred. In November, 44 species were collected. Of these, only Isotoma notabilis, Isotoma uniens, Microsotoma achromata, Neelus minimus, and Tullbergia yosii emerged as common species, collected from 11 of the 18 plots. In May, a greater number of species was identified (50) but only three, Isotoma notabilis, Isotoma uniens, and Microsotoma achromata, were widespread, each collected from at least 13 plots. These widespread species collectively accounted for a large percentage of the collembolan population in November (40%) and slightly 52 O a CCt 00 1-8 i -B u CO PQ £ 3 cct "3 § s PQ £.8 cj cc) fa H-l PQ £ 3 cct fa >-) o PH oo S3 1) Q 2? CO C Q Q Q CO P3 u Q c Q to 8 Q © 2 oo O N ^ © o Os CN C O co in O oo co co co co N O O N co N O C O in N O to 3 a CJ o •5 o N O O N CN O N CN rn O N CN co o o o oo O in N O o o o © o o O N O O O O CN co r--0 0 O N ^ ,—i m r--CN N O N O O N m O N CN C O o _ O N CN C O O N C N C O N O O O O O OO O N C N C O in C O CJ CO to Cj cj C3 N O oo O N C3 s <3 CJ oo O N s cj o c C3 CO oo m in o oo r--oo O N 1973 1315 3618 o 1974 C N CN C O 5262 O N CN C O oo O N 2960 O N CN C O 3288 CN CN CN o oo r-0 0 O N 2302 r~ oo O N 1316 o 4933 co m in o oo oo m N O 1--0 0 O N 3288 0 0 O N o 3618 C N 'Cf m CN N O o © 2631 O N CN C O 1315 0 0 O N © o r- oo C N N O o o 9069 1645 11509 2631 o © O N O N o 1645 1644 0 0 >n N O 1973 o 1973 m o OO 0 0 o 0 0 oo m N O o O N CN C O O N CN C O o O N CN C O O N 0 0 o CN o © 2960 1316 1644 O N cs C O 1644 © notabilis uniens '. achromata minimus biloba flavescens 3 O O . ed co co a "Ho , o Q '"2 PQ £ 3 ed ed cd I " SO .£3 £ 3 PQ &3 (2 J CO O 4) 00 c Q CO c Q QN CO Q c CJ Q G u Q £3^  QN ~ o ~ co so CN CO ^i o r--os oo I/O SO CD a CJ o V. CJ Q I CJ CD Is. CO so CN 00 OS CN os CN O oo co Os r--oo Os o ~ 00 m SO so CN 1/0 SO CO CN CO OS 8 CO too X ! I-I 1 — 1 PQ £ 3 ca as IX, HJ XI Ul ' - , PQ 5*3 cd PQ £ 3 o OH co NO 0s-CO C u Q 8 Q CO g Q co C Q CO g Q CO c u Q x" QN CO C u Q 0s-co g p g p >n CN in i—i CO m co oo in vo CN as CN CO 00 in vo in CO oo os as CN CO a CJ O •3 CO SO CN so CO SO CN OS CN CO OS CN CO 00 m SO SO 00 m so CJ oo OS Os CN CO 00 in SO so CN CO t--o OS © s m CN "cf r--co Os co VO CN SO CO as os in so 00 CN o SO OS CN PQ d, I cj CO K CD 8 o OS in m OS CN co 00 m so o o o o o o o o o o o o CJ 2 $3 © CN so CN m co Os SO o SO OS CN CN o CO CN so 00 oo CN co CN oo m SO CN OS OS CN CN CO CO o o o o o o o o o o o o in CO © ^ Os CN CO CN O SO Os CN OH CO 3 s cj O s is. o K u o PH oo 03 co I * '•2 o e u +-» 00 T 3 C o c o u 1 § S i-. 1 — 1 PQ £ 3 g L 2 PQ | 3 £ 3 fa >5 g PQ £ 3 03 C3 fa H4 00 'o u PH 00 N ? o NO o CO 00 CO o o © © © © o 7892 00 m NO o 1645 3947 1973 6905 1645 4275 © © © © ON o o ON © © © © CO ON CN CO 8879 3947 o 1973 ON CN CO ON CN CO 5590 © © © OO >n NO 4604 1316 o ON l O i n © CN CN © © i n c o -8221 7234 6248 3289 ON CN CO 1645 1315 00 ON © 2960 © 10852 2302 00 ON o o c o o *• 00 © © CN ON CN CO 2960 5261 o 00 ON 1315 o 2302 ON CN c o 4275 © © ON CN c o 00 i n NO NO ON CO CO o o CN o CN © © © © o NO f -o o ON 6905 2302 o o ON CN CO o ON CN c o © © © © © r~ o o ON CO NO r-- o ON CO •—1 m © © r-- © i n CN r-» o o ON 9207 2630 3617 3289 4604 1645 © © 2630 ON CN CO © 1644 00 i n NO CN CO I—1 o ON © NO t-- CO CN © © © CN ON CN c o 2302 1316 o 1645 00 i n NO © 2960 1316 4275 © © © ON CN CO O m o o o o CN © © CN © © © O 1315 ON CN CO o 00 i n NO o r-00 ON © ON CN CO © r-00 ON © © © O NO o CN NO o © © CO CO ON © © © O 00 m NO O 1315 00 m NO o © © ON CN CO ON CN CO o o ON © © © CO a CJ o « -*5 oo K •SJ '5 s CJ s cj C3 S Co CJ cs 3 •S 5 e CJ CJ to CJ 3 <3 CO 3 8! a? CJ fo ^3 UIC2IM0 Lzm C 2 I * 3 w g g s s w g g s s 911*3 I 0 I P J 9 G S S 96PI 8 I P J UI20ISS 2 0 i s s m 9 6 P J UIOOIAiD 66ss ui6iiss iu66ss 6 1 i s s 9 G S S m 9 2 I P J 001^° 9 2 I P J m S 2 P J Q2PJ t2iss C2MO CNI O) o co o d CD CO d CD O u LO d a CD CD X> d CD o CO d CM d co t-l CD 1 3 -»-» O "a, NO ON OS -a CD -4—» O CD cc5 CO CD ' o CD Q< co "o x> o U C+H o co w CD 5. & c» CO CO B 8 co t i £ <z » II CD OO cO ~ !^  « O cC 3 o Q II C+H CD CD CD II 00 CD U 3 OH o o co CD CD CO CD l-i Pi CD ? ^ S ON £ CD ed u rs > cd o •9 a S3 13 CD CD ,CD O •H S3 T3 o CD O CD co fcl CD ^ & t3 S .J2 CO O CD CD 8 1 Ui o CD —3 C3 t4—i 57 less in May (30%), yet they were rarely the dominant species of any community. Isotoma notabilis was the only widespread species frequently occurring in large numbers. Species that were not widespread but often abundant when present included Anurophorus binoculatus, Folsomia ozeana, Isotomiella minor, and Sensiphorura marshalli. Several species in both November (12) and May (10) were represented by a single specimen. 3.3.3 Time of sampling Samples were taken in both the spring and fall in hopes of catching collembolan species with different life cycle patterns. Though the number of species did not differ significantly between sampling times (Table 3.5), the May sampling yielded 16 additional species. Furthermore, 11 of the species that had been collected in November were absent in May. Time of sampling had no significant effect on collembolan density (Table 3.1) though for all conifer species, density was highest in May. The percentage increase from November to May varied from 17% (under hemlock) to as high as 66% (under Douglas-fir). 3.3.4 Soil moisture, bulk density, and pH No significant differences in moisture content, bulk density, and pH were found under the different conifer species (Tables 3.12 and 3.13). 58 Table 3.12. Analysis of moisture content, bulk density, and pH data. Source NumDF DenDF P value Treatment * Time 6 10 0.3170 Treatment 6 10 0.1915 Time 2 1 0.5875 Table 3.13. Least square means of moisture content, bulk density, and pH. Treatment Moisture content Bulk density pH 1 (Cw) 206.74 041 4^ 59 2(Fd) 231.78 0.33 4.30 3(Hw) 269.61 0.27 4.13 4(Ss) 238.49 0.27 4.51 3.4 DISCUSSION In general, the mean densities (between 2,700 collembola m"2 to 55,000 collembola m"2) found in this study fall within the range of those found in other temperate forests (Petersen and Luxton 1982; Tomlin and Miller 1987; Teuben and Smidt 1992). The mean densities of Collembola collected in British Columbia appear to vary greatly among different forest ecosystems. Marshall (1993) shows that the mean number of individuals collected from an Interior Douglas-fir forest near Kamloops (87,000 m"2) was substantially greater than the density found under a Coastal Douglas-fir forest on southern Vancouver Island (19,000 Collembola m2) and lower than that found under a Coastal Western Hemlock (Cedar-hemlock phase) forest on northern Vancouver Island (115,000 Collembola m"2). Additionally, a Subalpine forest near Sparwood supported a markedly lower density (7,000 individuals m"2) than any of these other three forest ecosystems. The total of 62 species identified in this study is comparable to the numbers found in other coniferous forests, though values of collembolan species richness, as with density values, also tend to vary widely among different conifer forests. Hagvar (1982) compiled a list of 60 collembolan species collected from seven different coniferous forest habitats in Norway. In British Columbia, Setala and Marshall (1994) sampled stumps in Douglas-fir forests on southeastern Vancouver Island and collected 72 collembolan species. Rusek and Marshall (1995) identified a substantially greater number of species (119) in their study of old-growth hemlock, old-growth Douglas-fir forests and a young Douglas-fir plantation on southern Vancouver Island. The factors responsible for the considerable variation in collembolan density and diversity observed among coniferous forests remain unknown. 60 Isotoma notabilis, Isotoma aniens, Microsotoma achromata, Neelus minimus, and Tullbergia yosii were the most common species in this study. Of these, I. notabilis and N. minimus have appeared as common or dominant species in other studies of coniferous forests in Canada and Europe (Poole 1961; Hagvar 1982; Rusek and Marshall 1995). In a study of soil collembolan communities from three forest sites on southern Vancouver Island, T. yosii was a common species while M. achromata and I. uniens were, though collected, not common to all sites (Rusek and Marshall 1995). Anurophorus binoculatus, Folsomia ozeana, Isotomiella minor, and Sensiphorura marshalli were not widespread species but were collected from at least one conifer species in such large numbers as to be the dominant species of that community. Of these, I. minor is a new record for British Columbia and A. binoculatus has only very recently been found elsewhere in British Columbia (J.A. Addison, pers. comm.). Rusek and Marshall (1995) also collected F. ozeana and S. marshalli in relatively high numbers in the old-growth western hemlock forest on southwestern Vancouver Island. The structure of the collembolan communities observed in this study, characterized by one to a few numerically dominant species and a large number of rare species, is typical of other population studies of Collembola (Usher 1970; Niijima 1971; Hagvar 1982; Setala and Marshall 1994; Rusek and Marshall 1995; van Straalen 1997). It is interesting to speculate why so many species were collected as only one or two specimens. It is possible that the extraction technique employed in this study was inefficient for these species or perhaps these species are highly aggregated. Aggregation in Collembola is well documented and the level of aggregation appears to vary among species (Poole 1961; Milne 1962; Joosse 1970). 61 Though there was no significant difference in the number of species collected between sampling times, there were several species that occurred at only one sampling time. The majority of these were rare with only one to a few specimens collected, however, a couple of notable exceptions, Isotomiella minor and Cryptopygus sp. B, were both very abundant only in May. Population peaks have been demonstrated for several collembolan species and the number and the timing of the peaks appear to vary among species (Milne 1962; Usher 1970). A seasonal fluctuation in Isotomiella minor has been found in other studies with a maximum occurring in August and a minimum in November and December (Usher 1970; Niijima 1971). This study found collembolan density significantly different among conifer species. Collembolan density can not always be correlated with vegetation type (Wood 1966; Curry 1978; Al-Assiuty et al. 1993), though evidence of a tree species effect has been found in recent studies. Blair et al. (1994) found significant effects of forest types on collembolan densities in four forest types in northeastern U.S. and Pinto et al.(1997) found significant differences in the density of soil and litter Collembola among stands of four different tree species in Portugal. Tree species can have many differing effects on forest soils (Binkley 1995) and it seems likely that a complex of interactions, involving vegetation as well as soil chemical, physical, and biological characteristics, may be influencing collembolan populations. The four forest types studied by Blair et al. (1994), which included a mixed hardwood forest, a red pine plantation, a beech forest, and a mature hemlock forest, differed greatly with respect to forest floor properties. The hemlock stand had the highest organic matter and moisture content, the lowest pH (3.13) and supported the highest number of Collembola. The red pine plantation had a lower organic matter content, higher pH (3.81), and much lower densities of Collembola. The beech forest had the second highest organic matter content, second lowest pH (3.53), and the second highest collembolan densities. Finally, the mixed hardwood forest had a low organic 62 matter content, the highest pH (4.81), and the lowest density of Collembola. Since no significant differences in moisture content, pH, and bulk density among conifer species were found in the present study, these factors cannot help to explain the observed tree species effect on collembolan density. Pinto et al. (1997) found that organic matter accumulation reduced collembolan densities in eucalyptus stands when compared with poplar, acacia, and alder stands. In that case, it appeared that the chemistry of the eucalyptus leaves (low nitrogen content and high polyphenolic contents) was an important factor in determining collembolan abundance. The nitrogen content of western redcedar litter is low relative to the other conifer species of this study (Prescott and Preston 1994), and the results of the above study would suggest that this may be an important contributing factor in the low collembolan density found under this tree species. The effect of conifer species on collembolan species richness is not easily explained. Hagvar (1982) found, in a study of several different habitats within Norwegian coniferous forests, that the number of collembolan species generally increased with increasing fertility. Conversely, Setala et al. (1995), did not find a positive correlation between collembolan distribution and the nitrogen content of stump wood. The present study found no clear correlation between species richness and soil fertility. In the November sampling, the greatest species richness was found, under all conifers, at the Fairy Lake site. Generally, the nutrient regime of the soils of this site was the poorest of all three sites studied (Klinka et al. 1994). In May, the site on which the greatest species richness was found varied for each conifer species. Furthermore, Prescott et a/.(1995) measured the concentrations of readily available nitrogen and the rates of N mineralization of the four conifer species on these sites and found the relative amounts to be site-specific. The forest floor of the western redcedar plots had the highest 63 values compared with all conifer species at the UK Mainline site yet the lowest values of the four species at the other two sites. The high concentration of available N at the UK Mainline site apparently had little influence on the relative number of collembolan species. Instead, species richness was still the lowest under western redcedar of all three species at this site, as it was at the other two sites. Regardless ofthe nutrient regime of the site, the effect of tree species on collembolan species richness was consistent and statistically significant. Western redcedar supported the lowest number of species at all sites and all sampling times (except in November 1995 at Branch 167) and Sitka spruce consistently supported the highest. The effect of conifer species on collembolan species may be attributable to differences in the litter chemistry. Prescott and Preston (1994) found differences in the nutrient, lignin, and tannin concentrations of needle litter of western hemlock, Douglas-fir, and western redcedar growing in adjacent plantations located in coastal British Columbia. Differences in the concentrations of nutrients in forest floor material were also found. Though not evaluated, it seems possible that similar differences in the litter could exist among the conifer stands in this study, which in turn, may lead to changes in the density of some collembolan species (Hagvar and Abrahamsen 1984; Geissen et al. 1997). Though studies have revealed a relationship between select collembolan species and certain soil chemical parameters, one should use caution in the interpretation of these results. An observed correlation between a species and certain soil characteristics is not necessarily evidence of a causal relationship. Also, as several soil chemical, physical and biological properties are interrelated, the effect of any one soil property on the distribution of a collembolan species is difficult to ascertain. As Hagvar and Abrahamsen (1984) point out, since Collembola live in the air-filled pores of the soil and have a hydrophobic body surface, relationships to soil chemical properties are most likely of an indirect nature. A correlation 64 between soil chemical properties and collembolan populations may well be the reflection of an effect of soil chemical characteristics on soil microflora. The results of a preliminary study of the fungal fruiting bodies found in the western redcedar, Douglas-fir, and Sitka spruce plots of the UK Mainline site suggest that conifer species may have an effect on the macrofungi of these plots (Strub 1996). Cluster analysis showed that the macrofungal communities under the three different forest covers were quite distinct from each other. Western redcedar plots supported the fewest species and the lowest percent abundance of macrofungi, which coincides with a lower abundance and species richness of Collembola found on the same plots. It should be noted that this study recorded only the presence of fungal fruiting bodies. It is not known whether other fimgi were present under western redcedar but not fruiting. It seems plausible that a difference among plots in the abundance and diversity of fruiting bodies reflects a difference in the fungal biomass in the soil (Newell 1984). The differences in macrofungal abundance and the distinct macrofungal communities found under each of the conifer species may provide part of the explanation for the differences in the collembolan communities observed under these same plots. Fungal hyphae are an important food source for many Collembola, particularly the larger species living in the upper layers of the soil (Knight and Angel 1967; Bodvarsson 1970; Gilmore and Raffensperger 1970; Moore et al. 1988; Chen et al. 1996). Thus, several factors related to the fungi, including quality as a food resource, fungal feeding preferences, and competition for limited resources may influence collembolan populations under different conifer species. The nutritional quality of fungal hyphae for fungivorous collembolans may vary not only with the fungal species but also with the nutrient content of the resource utilized by the fungus. In turn, the quality of the food source can impact the growth and fecundity of a collembolan species (Booth and Anderson 1979). It may therefore be of some consequence to 65 certain collembolan species that western redcedar litter contains the lowest nitrogen content of the conifer species studied. Food preferences of Collembola have been observed in both laboratory and field studies (Visser and Whittaker 1977; Newell 1984; Shaw 1985, 1988; Walsh and Bolger 1990; Thimm and Larink 1995; Chen et al. 1996) with rates of growth and reproduction affected by the different foods (Mills and Sinha 1971; Walsh and Bolger 1990; Klironomos et al. 1992; Chen et al. 1995). Findings of an experimental food enhancement study conducted by Chen and Wise (1997) strongly suggest that both the quantity and quality of food in the field have a significant impact on collembolan densities. The authors found that, provided with an increased food supply, densities of all Collembola groups increased. The preliminary results of Strub (1996), which implied a conifer species effect on the abundance and diversity of fungi, may point to a direct effect of food availability and quality on collembolan populations. It will be necessary to wait for results from the more in-depth study on the macrofiingi, not yet complete, to see if the trends in fungal distribution hold true. Western redcedar is unusual as a conifer species in forming vesicular-arbuscular mycorrhizae (Berch et al. 1991, 1992). In a laboratory feeding study, Shaw (1985) found that five collembolan species could distinguish and graze selectively on vesicular-arbuscular mycorrhiza (VAM). All but one species showed a significant preference for roots colonized by VAM when given the choice between colonized and noncolonized root material. Klironomos and Kendrick (1996) however, found that all three collembolan species of their study, when given the choice between VAM and conidial fungi, preferred a conidial fungus. The authors speculated that the results of their study might reflect an evolution by the VAM fungi of strategies to reduce intensive grazing by soil fauna. If VAM are less palatable to Collembola, 66 the formation of these mycorrhizae in western redcedar may contribute to a tree species effect on collembolan density and diversity. This study found that collembolan density and species richness differed significantly among conifer species. In particular, western redcedar stands consistently supported the lowest density, significantly different from the western hemlock and Sitka spruce stands, while the spruce stands consistently supported the highest collembolan abundance. Western redcedar stands also supported a significantly lower number of collembolan species than the spruce stands. It is speculated that the tree-species effect on collembolan populations was related to differences in fungal diversity and abundance. To further explore this possibility, the feeding attributes of five widespread collembolan species from this study were determined by gut content analysis. The results of that investigation are presented in the following chapter. 67 CHAPTER 4: GUT CONTENT ANALYSIS OF COLLEMBOLA COLLECTED FROM DIFFERENT CONIFER SPECIES 68 4.0 INTRODUCTION Collembola are numerically abundant microarthropods in litter and soil. The factors determining their density and community structure are not yet fully understood. Among the parameters which may influence soil collembolan distribution are temperature, soil type, moisture content, pH, the leaf litter layer, and characteristics of the fungal community (Hagvar and Abrahamsen 1984; Blair et al. 1994; Klironomos and Kendrick 1995; Geissen et al. 1997). The previous study (see chapter 3) found significant differences in both the density and richness of Collembola under different conifer species. The same study found no significant differences in soil moisture content, pH, and bulk density among conifer species. Consequently, these factors cannot help to explain the observed tree species effect on collembolan populations. On the other hand, the preliminary findings of Strub (1996) suggested that the abundance and diversity of macrofungi was different among the Sitka spruce, Douglas-fir and western redcedar plantations in at least one site of the study area. It was thus hypothesized that differences in collembolan populations may be partially attributable to differences in the fungal communities under the different conifer stands. Though many studies have attempted to elucidate collembolan feeding activity and choice of foods, determination of the natural diet of Collembola is difficult. While several laboratory experiments have demonstrated that a particular species will show a preference for a specific fungal species (Mills and Sinha 1971; Visser and Whittaker 1977; Aitchison 1983; Moore et al. 1987; Klironomos et al. 1992), such a preference is more difficult to demonstrate under field conditions. Another method of assessing feeding activity is through the examination of the gut contents or faecal pellets of Collembola collected from the field (McMillan and Healey 1971; Anderson and Healey 1972; McMillan 1975; Takeda and 69 Ichimura 1983; Davidson and Broady 1996). There are several limitations in this technique, however. The material in the gut may be difficult to recognize if it has been altered by 'chewing' or partial digestion. Identification of fungi to the species level is impossible using this approach, though the plant origin of pollen grains is more easily determined. Also, presence of a particular food type does not necessarily mean that the animal will derive any nutritional benefit from it. It may have consumed the material incidentally. Gut content analyses also tend to over-emphasize durable, non-digestible materials while under-estimating those materials more easily digested. Despite the problems inherent in gut content analysis, because gut contents are easily visible in animals cleared and mounted in preparation for species identification, this technique was easily incorporated into the population study presented in the previous chapter, while also providing valuable information on the natural diets of Collembola. The objective of this study was to investigate, using gut content analysis, the feeding attributes of collembolan species living under adjacent single species stands of western redcedar (Thuja plicata Dorm.), Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco), and Sitka spruce (Picea sitchensis [Bong.] Carr). The gut contents of five common collembolan species, Isotoma notabilis Schaffer, 1896, Isotoma uniens Christiansen and Bellinger, 1980, Micrisotoma achromata Bellinger, 1952, Neelus minimus (Willem, 1900), and Tullbergia yosii (Rusek, 1967) were statistically analyzed to determine if there were differences in the feeding habits within a collembolan species and between collembolan species among the three tree species. 7 0 4.1 STUDY SITES The study area from which the Collembola were sampled was described in detail in Chapter 3. 4.2 METHODS Approximately 1,900 Collembola were collected and prepared for identification to species by the methods described in Chapter 3. Following identification, the gut contents of all specimens were examined microscopically. The number of specimens with no visible gut contents was recorded. For the remainder of specimens, the proportion of the gut containing food was approximated. In order to survey the gut contents quantitatively, various categories were defined and the percentage of the gut comprised of each category was estimated with the aid of a micrometer fitted into the eyepiece ofthe microscope. The categories included organic matter, mineral matter, fungal hyphae, fungal spores, and animal parts. Fungal hyphae were divided into darkly pigmented hyphae and hyaline hyphae to determine if there were differences in collembolan feeding patterns between the two types of fungal hyphae. To determine if time of year or conifer species had an effect on the material found in collembolan guts, several categories were created and the percentage of total individuals whose guts were described by the following categories was calculated: 1) 50% or more of gut content comprised of organic matter; 2) 50% or more of gut content comprised of mineral material; 3) 50% or more of gut content comprised of darkly pigmented or hyaline fungal hyphae; 4) gut containing any amount of animal parts; 5) gut containing no more than a few fungal spores; 71 6) gut content 100% fungal spores; 7) gut without visible content. 4.2.1 Data analysis Isotoma notabilis, Isotoma uniens, Micrisotoma achromata, Neelus minimus, and Tullbergia yosii were the only collembolan species that occurred in a majority of all the plots sampled. Therefore, gut content data from only these species was analyzed using statistical analysis. PROC GLM in SAS, was used to assess both inter- and intraspecific differences in the percentages of individuals with guts empty, comprised of at least 50% organic matter, mineral material, hyaline fungal hyphae, and darkly pigmented fungal hyphae, among the three conifer species (SAS Institute Inc. 1989-1996). The occurrence of fungal spores and animal parts was minimal and therefore not included in the statistical analysis. 4.3 RESULTS A variety of materials was observed in the overall survey of the gut content of all collembolan species (Table 4.1). The most commonly observed materials included unidentified organic matter, mineral material, and darkly pigmented and hyaline fungal hyphae. Fungal spores and animal material were also observed, though less frequently and generally in lesser amounts. Recognizable plant parts (i.e. xylem), as well as pollen, diatoms, and other algae were never observed. A notable result of the gut content analysis was the prevalence of animals without visible gut content. Overall, 34% and 51% of all animals collected in November and May respectively, lacked visible gut content. A considerable variation in the percentage of 72 Table 4.1. Percentage of individuals of all species that contain various gut components. Gut component November May Empty gut 34 51 Organic matter (at least 50%) 63 65 Mineral matter (at least 50%) 30 11 No trace of fungal hyphae 32 30 Fungal hyphae (at least 50%) 14 25 Fungal spores (any amount) 10 17 Animal material (any amount) 6 3 individuals within a species without gut content was observed, ranging from 5-100% of individuals. Molting, a process possibly connected to the occurrence of empty guts, can sometimes be detected in mounted animals by the presence of a distinct double cuticle (Figure 4.1). A substantially higher number of individuals sampled in May (5%) was noticeably molting compared with November (0.8%). All animals with a double cuticle had empty guts. The most prevalent constituent of the guts was unidentifiable organic matter. It made up at least half of the gut content of 65% of individuals in May and of 63% of individuals collected in November. A substantial number of individuals in both May (26%) and November (23%) had gut contents comprised entirely of organic matter. Though fungi are considered an important component the collembolan diet, close to one third of individuals with visible gut content, collected at either sampling time, had no trace of fungi in their guts. Furthermore, an additional third of all individuals with visible gut content had no more than a few hyphae found within the contents of the gut. In May, only 25% of 73 Figure 4.1. Double cuticle of Folsomia ozeana. Magnified 150x. 74 individuals had gut contents comprised of at least 50% fungal hyphae. This is nearly double the percentage found in the November sampling (14%). Even fewer individuals had been feeding exclusively on fungal hyphae (4% in May and 2% in November). The amount of hyaline hyphae compared to darkly pigmented hyphae found in the gut contents was somewhat different between sampling times. Generally, in the November sampling, in individuals whose guts were at least half filled with hyphae, virtually equal amounts of the two types were found. However, in the May sampling, the proportion of dark hyphae to hyaline hyphae increased to 3: 1. Fungal spores were a relatively insignificant component of the collembolan diet. Though the odd individual was found with guts containing nothing but spores (9 specimens in total), most often only one to a few spores were observed in a gut. Spores were found in 17% of the individuals collected in May and in 10% of individuals collected in November. The majority of Collembola whose guts contained nothing but spores were, in all likelihood, feeding on mushrooms. The spores within these guts, characteristic of basidiospores, were consistently round, hyaline, and between 3-8 microns in diameter (Figure 4.2). Spores seemingly picked up incidentally, occurring as only one or a few in a gut, varied widely in shape and size and included the following (S.M. Berch, pers. comm.): 1) slightly pigmented, spiny, amerospores 2) spinose, hyaline spores with hyphal attachment 3) subhyaline, smooth-walled didymoconidia 4) dematiaceous, ameroconidia 5) sub-hyaline, thick-walled, angular basidiospores 6) spiny, globose, truffle spores. 75 Figure 4.2. Fungal spores from gut of Isotoma (Desoria) uniens. Magnified 150x. 76 A mere 3% of Collembola (16 individuals) collected in May and 6% (30 individuals) of those collected in November were observed as having ingested animal remains, including setae, scales, and exuvia of Collembola. What may have been cysts of amoebae were observed on two occasions. In May, animal remains most often occurred in the gut in only small amounts, and only 29% ofthe time made up at least half of the gut content. Animal parts were ingested in greater amounts in November and 41% of the time made up at least half of the gut content. 4.3.1 Feeding attributes of individual species I. notabilis, I. uniens, M. achromata, N. minimus, and T. yosii each consumed a similar variety of materials, consisting mainly of organic matter, mineral material and darkly pigmented and hyaline fungal hyphae. The relative percentages of individuals with empty guts and with guts comprised of at least 50% organic matter, mineral material, hyaline and darkly pigmented fungal hyphae are illustrated for each species in Figures 4.3, 4.4, 4.5, 4.6 and 4.7. Statistical analysis revealed no significant difference in the gut contents among these five collembolan species (Table 4.2). The sample size obviously varied among the collembolan species depending on their relative abundance under each conifer species. In November, a total of 69 specimens of I. notabilis, 32 specimens of/, uniens, 66 specimens ofM achromata, 46 of N. minimus, and 45 of T. yosii were included in the analysis. In May, a total 143 specimens of /. notabilis, 69 specimens of I. uniens, 34 specimens ofM achromata, 38 of N. minimus, and 27 of T. yosii were included in the analysis. 0 S ox ox ox x° ox \ ° 0S-O O O O O O o O O O O O ON 00 i> NO CO CN i—i > o T3 PH O PM co > O co t/2 cm O o uo co cd u CD 3 —< ts CO Q 1 B .2 CD Cl to O CD ° £ 00 £ CO ^ O I s 5, CD CD CD a, CO 2^ cd S O CO S || CO •2 oo O NO i f O ON co ON 73 ^ I £ I I P. > CD g 00 7J cd jo CD C CD CD c3 tS <^  a 'I .2 7D C3 « > co <u i-0X1 78 CD 03 CO •cd JS & bJO yph e hy ipty rkh alin uerj rgani p ccj a rgani S a 0 • • H • 0 o o o 00 o NO 0S-o O CN > O PH > O PH co CZ3 O co t/1 CM O \ ° o x O to 3 c<J <D CD — ' O -a. ^ CO <D 00 .a a .a b § * « 1  co rP i-T > GO 2 00 co O •3 Q W) 1  £,PH 8 8 ^ FL S P 1 £ CO 3 $ ca ."K "5 ™ » II CO •2 00 O VO i | O <J\ co II •I I •s iz 00 ca co 3 % CD C cu cd <^ a 1 . 2 > cu u 3 OJD 79 > o P H > O Z OH oo > o o o o x O o o oo r~ 0 s O 6 S -O in O •St" o cn o x O CN X ? O o x O CO p O > o o x O m -•-> CO cd 00 5 s co cd co ° E oo r* HS II '% * 4 u CO I PH 5 CO || C3 5 2 »s; cj C3 ca S h o e O H CO cd C O II -S co o . .co S O C O N CJ O N c<-i cd ° £ CO ^ ^ S • r m . £ O N X ) O N P T< M-H CO CO g 00 § o ^ t-l p CO -PH PH M £ 1 3 c^d To i n TT CU I-0 OX •a CO t3 a 80 o o O ON O 00 © r--o NO O o o CO o CN > O T3 PH > O Z PH GO 00 O co 00 CM O O IT) -M CO 03 CD os y I 3 •3 ^ 03 oj 8 6 ° CD j9 to M CD 5o £ - cd CO I •H co OH 5 a Q CD || CD O § •1—4 s S 53 CO 03 ' 00 co OO NO cu -a <U M CM I/-N 0 O N co O N 13 ~ I 5 • M S T3 CD « £ • M O CM V<-U C °0 "™ 03 J2 "S ^ S -a O CD ^ a S> co • M 3 1 -C 13 c3 Pi > NO CU s. 3 OX) 81 O O O ON O 00 O 0s-O VO "XT 0s-O i n o 0s-O 0s-O > o T 3 PH > O z PH GO 00 !> o GO OO C+H o 0s-O m -*-» GO a a <u o 00 , « S 6 O GO o p GO ^ g GO ra ccS £ "ob 3 g 00 o OH II <U PH •B '(& •r1 O * g :s O . Co GO § CO ^ £ •S 00 £? II ca GO g 00 !S VO E-H ON — ON C3 0s-o ca 3: £ U <+H m O ON GO ON 3 *H l is S •a o •s * cu g 5 -c o <u ^ a > i> CU OX) 82 Table 4.2. Analysis of gut content data to test for differences among five collembolan species. Source NumDF Den DF P value Treatment * Time 4 7 0.4301 Treatment 4 8 0.0946 Time 1 2 0.2271 Despite this, the appearance ofthe gut contents of two species were distinctly different from each other and the remaining three species. The gut contents of N. minimus, always divided into separate pellets, were virtually always comprised of a mixture of mineral matter and organic material. Fungal hyphae were observed in substantial amounts in only one specimen collected from a spruce plot in November. The organic material in the guts of this species was either without form or in minute pieces. The gut contents ofM achromata were noticeably different from those of all other species. Though a few specimens had consumed fairly large quantities of fungal hyphae, in 68% of individuals the entire gut content was comprised of organic matter. Virtually all (97%) individuals had at least half of the gut content comprised of organic matter. The appearance of the organic matter in the guts of these animals was markedly consistent. The guts were very densely packed with large pieces (up to 20 microns) of usually yellowish material with large pieces of darkly pigmented material sometimes mixed in. In both species, the appearance ofthe gut contents was markedly consistent among individuals and thus suggestive of a more selective feeding habit. A rather striking example of selective feeding was observed in one collembolan species, W. denisi. All but one of the 17 specimens observed had gut contents entirely comprised of fungal hyphae 83 (Figure 4.8). Nearly half of the individuals had only darkly pigmented hyphae in their guts. Only one individual had exclusively hyaline hyphae in its gut and, of those containing a mixture of the two types of fungi, only 29% contained more hyaline than darkly pigmented hyphae. A fascinating result of the gut content analysis of /. uniens was the observation of small pieces of bright purple colored material in 28% of the individuals collected in May. The origin of this material remains unknown. 4.3.2 Tree species effect For each collembolan species, no significant differences were found in the gut contents among the three conifer species (Table 4.3). Though not significant, the amount of fungal hyphae consumed appeared to increase in Collembola living under spruce (Table 4.4 and 4.5). This was particularly the case in November where, in all collembolan species except M. achromata, the consumption of fungal hyphae was highest in animals living under spruce compared to the other two conifer species. In November, individuals ofM. achromata living under western redcedar consumed the highest amount of fungal hyphae. In May, no individuals of N. minimus had consumed fungal hyphae. In the remaining three collembolan species, consumption of fungal hyphae was again highest in animals living under spruce. It should be noted that T. yosii was not collected from western redcedar in May, thus preventing a comparison of gut contents among all three conifer species at this sampling time. 85 4.3.3 Time of sampling No significant differences were found between the two sampling times in the relative percentages of individuals with guts empty, comprised of at least 50% organic matter, mineral material, hyaline fungal hyphae, and darkly pigmented fungal hyphae (Tables 4.2 and 4.3). 4.3.4 Parasites Several cases of what were speculated to be internal or external parasites were noted among a number of different species of Collembola. The most striking example involved hyaline, "bucky-ball" type structures, closely resembling an oomycete, which filled the body cavity of two specimens of Tullbergia yosii (Figure 4.9). These structures appeared as a large circular-shaped form measuring approximately 28 pm in diameter in which were contained a number of smaller circular-shaped forms measuring approximately 2.5 pm in width. Several examples of hyaline cells, differing in shape and size, were observed within the body, but conspicuously outside of the gut, in different collembolan species. The location of these cells led to the conjecture of possible parasitic infections. Elongated cells, approximately 10.5 pm in length and 2.5 pm in width were observed on Tullbergia ruseki. These cells, numbering 3 to 10 per collembolan, appeared to be attached externally. 86 Table 4.3. Analysis of collembolan gut content data to test for differences among conifer species. I. notabilis, I. uniens, M. achromata, N. minimus, and T. yosii analysed separately. Source Num DF Den DF P value Isotoma notabilis Treatment * Time 2 4 0.6149 Treatment1 2 4 0.7603 Time2 1 2 0.4224 Isotoma uniens Treatment * Time 2 2 0.9894 Treatment 2 4 0.1604 Time 1 2 0.5154 Micrisotoma achromata Treatment * Time 2 3 0.1095 Treatment 2 4 0.0523 Time 1 2 0.0995 Neelus minimus Treatment * Time 2 2 0.7960 Treatment 2 4 0.0832 Time 1 2 0.7050 Tullbergia yosii Treatment * Time - . -Treatment . 2 3 0.6464 Time 1 2 0.9737 1 Treatment =conifer species (western redcedar, Douglas-fir, and Sitka spruce). 2 Time= November 1995 and May 1996. 87 Table 4.4. Comparison of the quantity of fungal hyphae in gut contents of five species of collembolans collected in November 1995 from three different conifer species. Percentage of individuals with gut contents at least 50% hyaline or darkly pigmented hyphae. Hyaline fungal hyphae Darkly pigmented hyphae Isotoma notabilis Sitka spruce 5 23 Douglas-fir 0 0 Western redcedar 0 6 Isotoma uniens Sitka spruce 0 7 Douglas-fir 0 0 Western redcedar 0 0 Micrisotoma achromata Sitka spruce 0 0 Douglas-fir 0 0 Western redcedar 13 38 Neelus minimus Sitka spruce 4 0 Douglas-fir 0 0 Western redcedar 0 0 Tullbergia yosii Sitka spruce 46 0 Douglas-fir 30 0 Western redcedar 9 9 88 Table 4.5. Comparison of the quantity of fungal hyphae in gut contents of five species of collembolans collected in May 1996 from three different conifer species. Percentage of individuals with gut contents at least 50% hyaline or darkly pigmented hyphae. Hyaline fungal hyphae Darkly pigmented hyphae Isotoma notabilis Sitka spruce 1 22 Douglas-fir 2 14 Western redcedar 0 15 Isotoma uniens Sitka spruce 0 21 Douglas-fir 0 0 Western redcedar 0 0 Micrisotoma achromata Sitka spruce 8 0 Douglas-fir 0 0 Western redcedar 0 0 Neelus minimus Sitka spruce 0 0 Douglas-fir 0 0 Western redcedar 0 0 Tullbergia yosii Sitka spruce 5 15 Douglas-fir 0 17 Western redcedar no data no data 89 Figure 4.9. "Bucky-ball" structures from Tullbergia (Tullbergia) yosii. Magnified 200x. 90 4.4 DISCUSSION The gut contents of the Collembola of this study, including organic material, mineral material, darkly pigmented and hyaline fungal hyphae and, to a lesser degree, fungal spores and animal material, are typical food items of soil Collembola (Bodvarsson 1970; Takeda and Ichimura 1983; Hasegawa and Takeda 1995). The natural feeding habits and food preferences of most species of Collembola have not been well established. Several studies of the gut contents of field-collected Collembola suggest that many species will consume whatever food materials are readily available (Poole 1959; Bodvarsson 1970; Gilmore and Raffensperger 1970; McMillan and Healey 1971; Anderson and Healey 1972; Takeda and Ichimura 1983). Hasegawa and Takeda (1995) revealed both generalist and specialized feeding strategies in a study of the feeding attributes of four collembolan species, as they related to decomposition processes of pine needle litter. The authors found that generalist feeders, able to switch their feeding habits according to the availability of different food items (fungi and detritus), were also the numerically dominant species. The ability to switch feeding habits likely allows a collembolan species to exploit a wider food niche (Hasegawa and Takeda 1995). The present study, which failed to detect significant differences among the gut contents of I. notabilis, I. uniens, M. achromata, N. minimus, and T. yosii, suggests that these species are generalist feeders and non-selective in their feeding habits. This may help to explain the widespread distribution of these species. Definite preferences for certain foods in some collembolan species have been demonstrated in laboratory experiments (Mills and Sinha 1971; Visser and Whittaker 1977; Aitchison 1983; Moore et al. 1987; Klironomos et al. 1992). As well, field studies have shown some partitioning of food resources, or food specialization, among co-existing species, though 91 often with a tremendous amount of overlap between respective diets (Vegter 1983; Verhoef et al. 1988; Ponge 1991; Hasegawa and Takeda 1995; Chen et al. 1996). The species M. achromata may be a somewhat specialized feeder, based on the consistent appearance of the gut contents of most specimens in this study. Unfortunately, the broad definition of "organic matter" used in this study could not account for the unique appearance of its gut content, compared with the other collembolan species studied. For each collembolan species, no significant differences were found in the gut contents among the three conifer species. The consumption of fungal hyphae did however appear to increase in Collembola living under spruce. Chen and Wise (1997) found that enhancement of a fungal food resource resulted in higher densities of all collembolan families and suggested that collembolan populations are limited by the availability of this important food source. Knight and Angel (1967) found a high proportion of plant material in the guts of field collected tomocerids, though a preference for fungal material was shown in laboratory feeding experiments. This too, would suggest a scarcity of the preferred fungal food in the field. The increased consumption of fungal hyphae found in the present study, in animals living under spruce, correlated well with an increased abundance of fungal fruiting bodies observed under this conifer, relative to Douglas-fir and western redcedar (Strub 1996). The lower consumption of fungal hyphae in Collembola living under Douglas-fir and western redcedar may be the reflection of its limited availability under these conifer species, which would, in turn, provide an explanation for the relatively higher densities of Collembola observed under spruce (see previous chapter). The ability of certain collembolan species to change feeding habits according to food availability may provide an explanation for the seasonal variation in fungal feeding observed in several studies (Poole 1959; Takeda and Ichimura 1983; Newell 1984; Chen et al. 1996). 92 While no significant difference in gut contents was found between sampling times in this study, the consumption of fungal hyphae by I. notabilis was over twice as great in May compared to November and the over-all consumption of fungal hyphae (all species combined) was nearly twice as great in May. This contradicts the findings of Takeda and Ichimura (1983). These authors, in a study of four collembolan species collected on four sampling dates from a pine forest soil in Japan, reported a tendency by these species to consume greater amounts of hyphae in the autumn and winter compared to spring and summer. The apparent inconsistency between the two studies might be due to the limited nature ofthe sampling in the present study. Rather than indicating a seasonal trend, increased fungal consumption in May might be the result of an unusual feeding condition created by a rainfall following a dry period (Joosse and Testerink 1977). The proportion of empty guts in /. uniens was considerably higher relative to the other four species. This likely corresponds to the size and motility of this species. The morphological adaptations associated with the depth at which a collembolan species lives have long been recognized (Poole 1961) and it is likely that /. uniens, a large, pigmented species with well-developed eyes, lives in the litter and higher vegetation. In constrast, M. achromata, N. minimus, and T. yosii, are small, blind and white, which is typical of species that occupy the spaces between soil particles. Bodvarsson (1970) found that species living deeper in the soil, compared with species living in litter or vegetation, more frequently had filled guts containing relatively less fungal material. The author speculated that the food of species living deeper in the soil is qualitatively inferior to the food of those species living more superficially and therefore, must pass through the gut in greater quantities. This provides a possible explanation for the high incidence of empty guts observed in /. uniens in the present study. 93 Interestingly, the ingestion of hyaline fungi was conspicuously high in T. yosii, particularly in the November sampling. This presumably selective feeding on hyaline fungal hyphae was unusual. It was not apparent in the other common species of this study, nor was it apparent in the over-all results. Collembolan species used in laboratory feeding experiments have most often demonstrated a preference for darkly pigmented hyphae compared to hyaline fungal hyphae, though the reasons for this apparent feeding preference remain unclear (Mills and Sinha 1971; Visser and Whittaker 1977; Moore et al. 1987; Klironomos et al. 1992). Klironomos et al. (1992) observed significantly higher growth rates and fecundity in some collembolan species feeding on darkly pigmented hyphae compared with those feeding on hyaline fungal hyphae. The authors surmised that selective feeding on darkly pigmented hyphae may therefore be of some adaptive significance. In the case of T. yosii, perhaps its increased feeding on hyaline fungal hyphae represents a mechanism to reduce competition for a limited fungal resource, though this was not clearly demonstrated in this study. This study, similar to most other studies, found a high percentage of individuals lacking visible gut content (Poole 1959; Knight and Angel 1967; Bodvarsson 1970; Anderson and Healey 1972; Marshall 1978; Chen et al. 1996). This may be due to several factors related both to the diet of certain species and the influence of climate on food supply and subsequent feeding patterns. Certain species of Collembola, placed within the Subfamily Neanurinae, are distinguished from all other Collembola by the absence of a molar plate (Christiansen and Bellinger 1980-1981). It has been hypothesized that these species are fluid or suspension feeders (Macnamara 1924; Poole 1959; Christiansen 1964; Adams and Salmon 1972). This may account for the high percentage (93%) of Neanurinae species found in this study that lacked visible gut contents, compared to 31% (November) and 48% (May) of all other species. 94 For species possessing a molar plate, the gut may appear empty yet contain material (i.e. bacteria) not visible in mounted specimens (Poole 1959; Christiansen 1964). It is also possible that the gut is empty due to a cessation in feeding associated with molting. Collembola molt throughout their lives, the process continuing long after sexual maturity has been reached. A few days prior to ecdysis, an individual will stop moving and feeding, during which time the new cuticle is secreted (de With and Joosse 1971). This would explain why 100% of the animals in the present study which possessed a distinct double cuticle also had no visible gut content. Since the time spent feeding is limited by the molting process, Joosse and Testerink (1977) estimated that, with an adequate food supply and temperatures above 10°C, from only 50 to 60% of animals in a field population are in a fed state at any one time. The occurrence of empty guts appears to be related to soil temperature as well as food availability. At lower temperatures, the percentage of animals without gut contents becomes higher (de With and Joosse 1971; Joosse and Testerink 1977; Takeda and Ichimura 1983). This was not apparent in the present study as the percentage of individuals without visible gut content was higher in May 1996 compared with November 1995. However, since sampling occurred on only two occasions, this is not necessarily indicative of a seasonal trend. Perhaps the higher number of individuals lacking gut content in May is a reflection of an exceptional feeding condition as revealed by Joosse and Testerink (1977). These authors found unexpectedly low percentages of Orchesella cincta (Linne) with food in their guts corresponding to dry periods, which presumably created a food shortage. However, sampling approximately one week after a rainfall event, which had been preceded by a prolonged dry period, also resulted in an unexpectedly low percentage of animals with gut content. The authors surmised that rainfall following a dry period brought about a sudden change in food 95 supply which subsequently induced a high level of feeding activity. This renewed feeding activity led to a temporary synchronization in the molting rhythm, the molting interval being about 4-5 days. The synchronization in molting disappeared after approximately 2 weeks as a result of individual variations in the duration of instars. Unfortunately, the weather patterns immediately preceding the May sampling ofthe present study are unknown. However, a substantially higher percentage of individuals with distinct double cuticles were found in these samples, which supports the possibility that the low percentage of individuals with gut content is the result of a recent synchronization of feeding and molting. Joosse and Testerink (1977) found an extremely high percentage (91%) of individuals extracted using the Tullgren method had no visible gut content when compared to those collected by hand sorting and thus suggested that funnel extraction was inappropriate for gut content analyses. On the other hand, Marshall (1978) reported a large percentage of individuals with no visible gut content in specimens collected from the field by hand and placed directly into alcohol. The present study, despite having used a funnel extraction method, yielded a reasonable number of Collembola with visible gut content (69% in November and 52% in May), which is similar to the findings of Chen et al. (1996). The extraction of all samples was carried out under the same conditions, so the relative amount of loss should be constant between the different sampling dates. The results of this study found no significant differences among the diets of five common collembolan species. Also, no significant differences were found in the feeding habits of any Collembola among three different conifer species. The diet categories used in this study were very broad and at the scale used, these five collembolan species appear to be non-selective feeders with similar diets. However, the consumption of fungal hyphae was greater in some collembolan species living under Sitka spruce compared to Douglas-fir and western redcedar 96 plantations. This may be the reflection of an increased availability of this food source Sitka spruce, which may, in turn, provide an explanation for the increased density of Collembola found under this tree species. GENERAL CONCLUSIONS A total of 62 collembolan species was identified in this study, including four new records for Canada and three new records for British Columbia. Collembolan densities as well as collembolan richness were significantly different among single species stands of Sitka spruce, Douglas-fir, western hemlock, and western redcedar. Several studies have found collembolan populations to be correlated with soil characteristics, including those measured here (Hagvar and Abrahamsen 1984; Blair et al. 1994; Geissen et al. 1997). However, as no significant differences were detected in the soil characteristics measured, which included moisture content, pH, and bulk density, these parameters did not correlate with differences in collembolan populations. Results from the gut content analyses of I. notabilis, I. aniens, M. achromata, N. minimus, and T. yosii revealed no significant differences in the diets among these five more 1 common collembolan species. In addition, no significant differences were found in the diets of any one of these species living under different conifer species. This suggests that these Collembola were non-selective in their feeding. However, there was an increase, though not significant, in the amount of fungal hyphae consumed by certain species living under spruce compared with those found under Douglas-fir or western redcedar. Hasegawa and Takeda (1995) found that certain collembolan species were able to switch food, depending on its availability. Preliminary findings (Strub 1996) suggested there was a greater abundance of fungal fruiting bodies under spruce compared to the other two conifer species. This may account for the increased consumption of fungal hyphae observed under this conifer species. Though the majority of Collembola consume a wide variety of materials, fiingi may be the most important source of nutrition and energy (Moore et al. 1988). Collembolan densities increased in the field with an enhanced fungal food source, suggesting that this food source is 98 an important limiting factor in collembolan population dynamics (Chen and Wise 1997). An increase in the consumption of fungal hyphae under spruce compared to Douglas-fir and western redcedar suggested its increased availability under this conifer species. This, in turn, might provide an explanation for the increased densities of collembola found under spruce, compared to three other conifer species. ) The issue of biodiversity and its importance to ecosystem health has received increased attention in recent years (Probst and Crow 1991; Naeem et al. 1995; Johnson et al. 1996; Wardle et al. 1997). Several hypotheses regarding the nature of the relationship between biological diversity and productivity and stability of ecosystems have been put forth. These include the widely accepted diversity-stability hypothesis, in addition to the more recently proposed rivet, redundancy, and idiosyncratic hypotheses (Johnson et al. 1996). Regardless of the outcome of this controversy, knowledge ofthe biological diversity of an ecosystem, in addition to the interaction of the organisms within it, is fundamental to an understanding of the effect of human-mediated disturbances to a system. The forest industry plays a vital role in the economy of British Columbia and thus, the factors affecting the long-term productivity of its forests deserve investigation. This includes research into the impacts of forest management practices on below-ground diversity. Unfortunately, studies of soil arthropod species diversity and ecology in B. C.'s forests are extremely limited (Craig 1995; Rusek and Marshall 1995) and, as Marshall (1993) points out, basic taxonomic and ecological research on British Columbia's soil fauna are desperately needed. 99 LITERATURE CITED Adams, E.C.G. and Salmon, J.T. 1972. The mouth-parts and feeding methods of Brachystomella parvula (Schaeffer) (Collembola: Brachystomellidae). Transactions of the Royal Entomological Society of London. 124: 269-286. Addison, J.A. 1980. Influence of individual plant species on the distribution of Arctic Collembola. In: Dindal, D.L. (editor). Soil biology as related to land use practices. Proceedings, 7th International Soil Zoology Colloquium ofthe International Society of Soil Science. Office of Pesticide and Toxic Substances, Environmental Protection Agency, Washington. D.C. pp. 704-715. Aitchison, CW. 1983. Low temperature and preferred feeding by winter-active Collembola (Insecta, Apterygota). Pedobiologia 25: 27-36. Al-Assiuty, A.I., Bayoumi, B.M, Khalil, M.A., and van Straalan, N.M. 1993. The influence of vegetational type on seasonal abundance and species composition of soil fauna at different localities in Egypt. Pedobiologia 37: 210-222. Anderson, J.M. and Healey, I.N. 1972. Seasonal and inter-specific variation on major components of the gut contents of some woodland collembola. Journal of Animal Ecology 41: 359-368. Barra, J.A. and Christiansen, K. 1975. Experimental study of aggregation during the development of Pseudosinella impediens (Collembola, Entomobryidae). Pedobiologia 15: 343-347. Battigelli, J.P. Berch, S.M., and Marshall, V.G. 1994. Soil fauna communities in two distinct but adjacent forest types on northern Vancouver Island, British Columbia. Canadian Journal of Forest Research 24: 1557-1566. Battigelli, J.P. and Marshall, V.G. 1993. Relationships between soil fauna and soil pollutants. In: Marshall, V.G. (compiler). Proceedings of the Forest Ecosystem Dynamics Workshop. February 10-11,1993. FRDA Report 210, Forestry Canada and British Columbia Ministry of Forests, Victoria, BC. pp. 31-34. Bauer, T. 1982. Predation by a carabid beetle specialised for catching collembola. Pedobiologia 24: 169-79. Bauer, T. 1985. Beetles which use a setal trap to hunt springtails: the hunting strategy and apparatus ofLeistus (Coleoptera, Carabidae). Pedobiologia 28: 275-287. Bengtsson, G. and Rundgren, S. 1983. Respiration and growth of a fungus, Mortierella isabellina, in response to grazing by Onychiurus armatus (Collembola). Soil Biology and Biochemistry 15: 469-473. 100 Berch, S.M., Deom, E., and Willingdon, T. 1991. Western red cedar growth and vesicular-arbuscular mycorrhizal colonization in fumigated and nonfumigated nursery beds. Tree Planters' Notes 42: 14-16. Berch, S.M., Deom, E., Roth, A. and Beese, WJ. 1992. Vesicular-arbuscular mycorrhizae of western redcedar in container nurseries and on field sites after slash burning. Tree Planters' Notes 44: 33-37. Berg, N.W. and Pawluk, S. 1984. Soil mesofaunal studies under different vegetative regimes in north central Alberta. Canadian Journal of Soil Science 64: 209-223. Binkley, D. 1995. Species influence on forest soils-processes and patterns. In: Mead, D.J. and Cornforth, I.S. (editors). Proceedings of the Trees and Soils Workshop. Lincoln University, Canterbury, New Zealand, February 28-March 2, 1994. Agronomy Society of New Zealand. Special Publication No. 10. pp.1-33. Blair, J.M., Parmelee, R.W., Wyman, R.L. 1994. A comparison of the forest floor invertebrate communities of four forest types in the northeastern U.S. Pedobiologia 38: 146-160. Blancquaert, J.-P., Coessens, R. and Mertens, J. 1981. Life history of some Symphypleona (Collembola) under experimental conditions. II. - Postembryonal development and reproduction. Revue d'Ecologie et de Biologie du Sol 18: 373-390. Bodvarsson, H. 1970. Alimentary studies of seven common soil-inhabiting collembola of southern Sweden. Entomologica Scandinavica. 1: 74-80. Booth, R.G. and Anderson, J.M. 1979. The influence of fungal food quality on the growth and fecundity of Folsomia Candida (Collembola: Isotomidae). Oecologia 38: 317-323. Chen, B., Snider, R.J., and Snider, R.M. 1995. Food preferences and effects of food type on the life history of some soil Collembola. Pedobiologia 39: 496-505. Chen, B. Snider, R.J., and Snider, R.M. 1996. Food consumption by Collembola from northern Michigan deciduous forest. Pedobiologia 40: 149-161. Chen, B. and Wise, D.H. 1997. Responses of forest-floor fungivores to experimental food enhancement. Pedobiologia 41:316-326. Choudhuri, D.K., Bhattacharyya, B., Mukherjee, P., and Pande, T. 1979. The effects of various physical factors on the biology of Lobelia (Lobelia) maxillaris Yosii, 1966 (Collembola: Insecta). Revue d'Ecologie et de Biologie du Sol 16: 241-257. Christiansen, K. 1964. Bionomics of Collembola. Annual Review of Entomology 9: 147-178. Christiansen, K. 1970. Experimental studies on the aggregation and dispersion of Collembola. Pedobiologia 10: 180-198. 101 Christiansen, K. 1990. Insecta: Collembola. In: Dindal, D.L. (editor). Soil biology guide. John Wiley and Sons, New York, NY. pp. 965-995. Christiansen, K. and Bellinger, P. 1980-81. The Collembola of North America. North of the Rio Grande. Grinnell College, Iowa. Craig, K. 1995. Variation in Carabid community structure associated with Coastal Douglas-fir forest sucessional stages. Master's Thesis. Department of Zoology. University of British Columbia. Curry, J.P. 1978. Relationships between microarthropod commumties and soil and vegetational types. Scientific Proceedings ofthe Royal Dublin Society. Series A 6: 131-141. Davidson, M.M. and Broady, P. A. 1996. Analysis of gut contents of Gomphiocephalus hodgsoni Carpenter (Collembola: Hypogastruridae) at Cape Geology, Antarctica. Polar Biology 16: 463-467. de With, N.D. and Joosse, E.N.G. 1971. The ecological effects of moulting in collembola. Revue d'Ecologie et de Biologie du Sol 8:111-117. Faber, J.H. and Verhoeff, H.A. 1991. Functional differences between closely-related soil arthropods with respect to decomposition processes in the presence or absence of pine tree roots. Soil Biology and Biochemistry 23: 15-23. Finlay, R.D. 1985. Interactions between soil micro-arthropods and endomycorrhizal associations of higher plants. In: Fitter, A.H., Atkinson, D., Read, K.J., and M.B. Usher (editors). Ecological interactions in soil: plants, microbes, and animals. Blackwell, Oxford, pp. 319-331. Fjellberg, A. 1985. Arctic Collembola I - Alaskan Collembola of the families Poduridae, Hypogastruridae, Odontellidae, Brachystomellidae and Neanuridae. Entomologica Scandinavica Supplement No. 21. Geissen, V., Illmarm, J., Flohr, A., Kahrer, R. and Brummer, G. 1997. Effects of liming and fertilization on Collembola in forest soils in relation to soil chemical parameters. Pedobiologia 41:194-201. Gilmore, S.K. and Potter, D.A. 1993. Potential role of Collembola as biotic mortality agents for entomopathogenic Nematodes. Pedobiologia 37:30-38. Gilmore, S.K. and Raffensperger, E.M. 1970. Foods ingested by Tomocerus spp. (Collembola, Entomobryidae), in relation to habitat. Pedobiologia 10: 135-140. Hagvar, S. 1982. Collembola in Norwegian coniferous forest soils. I. Relations to plant communities and soil fertility. Pedobiologia 24: 255-296. 102 Hagvar, S. and Abrahamsen, G. 1984. Collembola in Norwegian coniferous forest soils III. Relations to soil chemistry. Pedobiologia 27: 331-339. Hale, W. G. 1966. A population study of moorland Collembola. Pedobiologia 6: 65-99. Hanlon, R.D.G. 1981. Influence of grazing by Collembola on the activity of senescent fungal colonies grown on media of different nutrient concentration. Oikos 36: 362-367. Hanlon, R.D.G. and Anderson, J.M. 1979. The effects of Collembola grazing on microbial activity in decomposing leaf litter. Oecologia 38: 93-99. Hasegawa, M. and Takeda, H. 1995. Changes in feeding attributes of four collembolan populations during the decomposition process of pine needles. Pedobiologia 39: 155-169. Hodkinson, I.D., Coulson, S., Webb, N.R., Block, W., Strathdee, A.T., and Bale, J.S. 1994. Feeding studies on Onychiurus arcticus (Tullberg) (Collembola: Onychiuridae) on West Spitsbergen. Polar Biology 14: 17-19. Hopkin, S. 1997. Biology of Springtails (Insecta: Collembola). Oxford University Press, Inc., New York. Ineson, P., Leonard, M.A. and Anderson, J.M. 1982. Effect of Collembolan grazing upon nitrogen and cation leaching from decomposing leaf litter. Soil Biology and Biochemistry 14: 601-605. Johnson, K.H., Vogt, K.A., Clark, H.J., Schmitz, O.J., and Vogt, D.J. 1996. Biodiversity and the productivity and stability of ecosystems. Tree. 11: 372-377. Joosse, E.N.G. 1970. The formation and biological significance of aggregations in the distribution of collembola. Netherlands Journal of Zoology 20: 299-314. Joosse, N.G. and Testerink, G.J. 1977. The role of food in the population dynamics of Orchesella cincta (Linne) (Collembola). Oecologia 29: 189-204. Kalra, Y.P. and Maynard, D.G. 1991. Methods Manual for Forest Soil and Plant Analysis. Forestry Canada. Information Report. NOR-X-319. Kevan, P.G. and Kevan, D.K. 1970. Collembola as pollen feeders and flower visitors with observations from the high arctic. Quaestiones entomologicae 6: 311-326. Klinka, K , Wang, Q., Sivak, B., Wang, G.G. and. Elstone, D. 1994. Ecological and supplementary height growth analyses for experimental projects 368 and 571. Contract Report to B.C. Ministry of Forests Research Branch. Victoria, B.C. Klironomos, J.N. and Kendrick, B. 1995. Relationships among microarthropods, fungi, and their environment. Plant and Soil 170: 183-197. 103 Klironomos, J.N. and Kendrick, W.B. 1996. Palatability of microfungi to soil arthropods in relation to the functioning of arbuscular mycorrhizae. Biology and Fertility of Soils 21: 43-52. Klironomos, J.N., Widden, P. and Deslandes, I. 1992. Feeding preferences of the collembolan Folsomia Candida in relation to microfungal successions on decaying litter. Soil Biology and Biochemistry 24: 685-692. Knight, C.B. and Angel, R.A. 1967. A preliminary study of the dietary requirements of Tomocerus (Collembola). The American Midland Naturalist 77: 510-517. Krebs, C.J. 1989. Similarity Coefficients and Cluster Analysis. In: Pisano, S. and Wilson, CM. (editors). Ecological Methodology. Harper Collins, NY. pp. 293-327. Lavy, D. and Verhoef, H.A. 1996. Effects of food quality on growth and body composition of the collembolan Orchesella cincta. Physiological Entomology 21: 64-70. Lussenhop, J. 1971. A simplified canister-type soil arthropod extractor. Pedobiologia 11: 40-45. Lussenhop, J. 1992. Mechanisms of microarthropod-microbial interactions in soil. In: Begon, M. and Fitter, A.H. (editors). Advances in Ecological Research. Volume 23. pp. 1-33. Macnamara, C. 1924. The food of collembola. The Canadian Entomologist 56: 99-105. Marshall, V.G. 1974. Seasonal and vertical distribution of soil fauna in a thinned and urea fertilized Douglas-fir forest. Canadian Journal of Soil Science 54: 491-500. Marshall, V.G. 1978. Gut content analysis ofthe Collembolan Bourletiella hortensis (Fitch) from a forest nursery. Revue d'Ecologie et de Biologie du Sol 15: 243-250. Marshall, V.G. 1993. Sustainable forestry and soil fauna diversity. In: Fenger, M.A., Miller, E.H., Johnson, J.A. and E.J.R. Williams (editors). Our Living Legacy : Proceedings of a Symposium on Biological Diversity Royal British Columbia Museum. Victoria, B.C. pp. 239-248. McMillan, J.H. 1975. Interspecific and seasonal analyses of the gut contents of three Collembola (Family Onychiuridae). Revue d'Ecologie et de Biologie du Sol 12: 449-457. McMillan, J.H. and Healey, I.N. 1971. A quantitative technique for the analysis ofthe gut contents of Collembola. Revue d'Ecologie et de Biologie du Sol 8: 295-300. Meidinger, D. and Pojar, J. 1991. Ecosystems of British Columbia. Special Report Series No. 6. British Columbia Ministry of Forests. 104 Mills, J.T. and Sinha, R.N. 1971. Interactions between a springtail Hypogastrura tullbergia and soil borne fungi. Journal of Economic Entomology 64: 398-401. Milne, S. 1962. Phenology of a natural population of soil Collembola. Pedobiologia 2: 41-52. Moore, J.C, Ingham, E.R., and Coleman, D.C. 1987. Inter- and intraspecific feeding selectivity of Folsomia Candida (Willem) (Collembola, Isotomidae) on fungi. Biology and Fertility of Soils 5: 6-12. Moore, J.C, Walter, D.E., and Hunt, H.W. 1988. Arthropod regulation of micro- and mesobiota in below-ground detrital food webs. Annual Review of Entomology 33: 419-39. Morisita, M. 1959. Measuring interspecific association and similarity between communities. Memoirs of the Faculty of Science, Kyushu University. Series. E (Biology) 3: 65-80. Morton, J. 1990. In: International culture collection of vesicular-arbuscular mycorrhizal fungi (INVAM) Newsletter 1: 6. Naeem, S. Thompson, L.J., Lawler, S.P., Lawton, J.H., and Woodfin, R.M. 1995. Empirical evidence that declining species diversity may alter the performance of terrestrial ecosystems. Philosophical Transactions of the Royal Society of London. 347: 249-262. Newell, K. 1984. Interaction between two decomposer basidiomycetes and a collembolan under Sitka spruce: distribution, abundance, and selective grazing. Soil Biology and Biochemistry 16: 227-233. Niijima, K. 1971. Seasonal changes in collembolan populations in a warm temperate forest of Japan. Pedobiologia 11: 11-26. Omule, S.A.Y. and Krumlik, G.J. 1987. Junvenile height growth of four species on four sites in the CWHbl Variant. FRDA Report 007. Canadian Forestry Service and B.C. Ministry of Forests and Lands, Victoria, B.C. Parkinson, D., Visser, S. and Whittaker, J.B. 1979. Effects of collembolan grazing on fungal colonization of leaf litter. Soil Biology and Biochemistry 11: 529-535. Payne, J.A., King, E.W. and Beinhart, G. 1968. Arthropod succession and decomposition of buried pigs. Nature 219: 1180-1181. Petersen, H. 1994. A review of collembolan ecology in ecosystem context. Acta Zoologica Fennica 195: 111-118. Petersen, H. and Luxton, M. 1982. A comparative analysis of soil fauna populations and their role in decomposition processes. Oikos 39: 291-388. 105 Pinto, C , Sousa, J.P., Graca, M.A.S. and da Gama, M.M. 1997. Forest soil Collembola. Do tree introductions make a difference? Pedobiologia 41: 131-138. Ponge, J.F. 1991. Food resources and diets of small animals in a small area of Scots pine litter. Geoderma 49: 33-62. Poole, T.B. 1959. Studies on the food of Collembola in a Douglas fir plantation. Proceedings of the Zoological Society of London 132: 71-82. Poole, T.B. 1961. An ecological study of the Collembola in a coniferous forest soil. Pedobiologia 1: 113-137. Poole, T.B. 1962. The effect of some environmental factors on the pattern of distribution of soil Collembola in a coniferous woodland. Pedobiologia 2: 169-182. Pozo, J. 1986. Ecological factors affecting Collembola populations. Ordination of communities. Revue d'Ecologie et de Biologie du Sol 23: 299-311. Prescott, C.E., Pratt, J. and Staley, CL. 1995. The influence of tree species on rates of nitrogen mineralization and litter decomposition in coastal British Columbia. Contract Report to B.C. Ministry of Forests, Research Branch. Prescott, C.E. and Preston, CM. 1994. Nitrogen mineralization and decomposition in forest floors in adjacent plantations of western red cedar, western hemlock, and Douglas-fir. Canadian Journal of Forest Research 24: 2424-2431. Probst, J.R. and Crow, T.R. 1991. Integrating biological diversity and resource management. Journal of Forestry. 89: 12-17. Routledge, R.D. 1979. Diversity indices: which ones are admissible? Journal of Theoretical Biology 76: 503-515. Rusek, J. 1974. Die Preparation von Kleininsekten. Mickrokosmos 63: 10-12. Rusek, J. and Marshall, V.G.. 1995. Long-term changes of collembolan communities in forest soils. Bulletin Entomologique de Pologne 64: 159-171. SAS Institute Inc. 1989-1996. Windows NT Version 4.0.13891. Cary, NC. Seastedt, T.R. 1984. The role of microarthropods in decomposition and mineralization processes. Annual Review of Entomology 29: 25-46. Setala, H. 1995. Growth of birch and pine seedlings in relation to grazing by soil fauna on ectomycorrhizal fungi. Ecology. 76: 1844-1851. 106 Setala, H., Marshall, V.G. and Trofymow, J. A. 1995. Influence of micro- and macro-habitat factors on collembolan communities in Douglas-fir stumps during forest succession. Applied Soil Ecology. 2: 227-242. Setala, H. and Marshall, V.G. 1994. Stumps as a habitat for Collembola during succession from clear-cuts to old-growth Douglas-fir forests. Pedobiologia 38: 307-326. Shaw, P. J.A. 1985. Grazing preferences oi Onychiurus armatus (Insecta: Collembola) for mycorrhizal and saprophytic fungi of pine plantations. In: Fitter, A.H., Atkinson, D., Read, K.J., and M.B. Usher (editors). Ecological interactions in the soil: plants, microbes, and animals. Blackwell, Oxford, pp. 333-337. Shaw, P. J. A. 1988. A consistent hierarchy in the fungal feeding preferences ofthe Collembola Onychiurus armatus. Pedobiologia 31: 179-187. Skidmore, R.E. 1995. Checklist of Collembola (Insecta: Apterygota) of Canada and Alaska. Proceedings ofthe Entomological Society of Ontario 126: 45-76. Strub, N. 1996. Auswirkungen forstlicher Monokulturen auf Pilzflora und Bodenbeschaffenheit. Thesis. University of Hamburg. Takeda, H. and Ichimura, T. 1983. Feeding attributes of four species of Collembola in a pine forest soil. Pedobiologia 25: 373-381. Teuben, A. and Smidt, G.R.B. 1992. Soil arthropod numbers and biomass in two pine forests on different soils, related to functional groups. Pedobiologia 36: 79-89. Teuben, A. and Verhoef, H.A. 1992. Direct contribution by soil arthropods to nutrient availability through body and faecal nutrient content. Biology and Fertility of Soils 14: 71-75. Thimm, T. and Larink, O. 1995. Grazing preferences of some collembola for endomycorrhizal fungi. Biology and Fertility of Soils 19: 266-268. Tomlin, A.D. and Miller, J. J. 1987. Composition of soil fauna in forested and grassy plots at Delhi, Ontario. Canadian Journal of Zoology 65: 3048-3055. Usher, M.B. 1970. Seasonal and vertical distribution of a population of soil arthropods: Collembola. Pedobiologia 10: 224-236. Usher, M.B. and Hider, M. 1975. Studies on populations of Folsomia Candida (Insecta: Collembola): causes of aggregations. Pedobiologia 15: 276-283. van Straalen, N.M. 1997. Community structure of soil arthropods as a bioindicator of soil health. In: Pankhurst, C.E., Doube, B.M. and Gupta, V.V.S.R. (editors). Biological indicators of soil health. CAB International, pp. 235-264. 107 van Straalen, N.M. and Joosse, E.N.G. 1985. Temperature responses of egg production and egg development in two species of Collembola. Pedobiologia 28: 265-273. Vegter, J.J. 1983. Food and habitat specialization in coexisting springtails (Collembola, Entomobryidae). Pedobiologia 25: 253-262. Verhoef, H. A., Nagelkerke, C.J., and Joosse, E.N.G. 1977. Aggregation pheromones in Collembola (Apterygota); a biotic cause of aggregation. Revue d'Ecologie et de Biologie du Sol 14: 21-25. Verhoef, H.A., Prast, J.E., and Verwiej, R.A. 1988. Relative importance of fungi and algae in the diet and nitrogen nutrition of Orchesella cincta (L.) and Tomocerus minor (Lubbock) (Collembola). Functional Biology 2: 195-201. Visser, S. 1985. Role of the soil invertebrates in determining the composition of soil microbial communities. In: Fitter, A.H., Atkinson, D., Read, D.J. and M.B. Usher (editors). Ecological interactions in the soil: plants, microbes and animals. Special Publication No. 4 of the British Ecological Society. Blackwell Scientific Publications, Oxford. Visser, S. and Whittaker, J.B. 1977. Feeding preferences for certain litter fungi by Onychiurus subtenuis (Collembola). Oikos 29: 320-325. Vlug, H. and Borden, J.H. 1973. Soil Acari and Collembola populations affected by logging and slash burning in a coastal British Columbia coniferous forest. Environmental Entomology 2: 1016-1023. Walsh, M.I. and Bolger, T. 1990. Effects of diet on the growth and reproduction of some Collembola in laboratory cultures. Pedobiologia 34: 161-171. Walter, D.E. 1987. Trophic behaviour of "mycophagous" microarthropods. Ecology 68: 226-229. Walter, D.E., Kethley, J., and Moore, J.C. 1987. A heptane flotation method for recovering microarthropods from semiarid soils, with comparison to the Merchant-Crossley high-gradient extraction method and estimates of microarthropod biomass. Pedobiologia. 30: 221-232. Wardle, D.A., Bonner, K.I. and Nicholson, K.S. 1997. Biodiversity and plant litter: experimental evidence which does not support the view that enhanced species richness improves ecosystem function. Oikos. 79: 247-258. Wolda, H. 1981. Similarity indices, sample size and diversity. Oecologia 50: 296-302. Wood, T.G. 1966. The fauna of grassland soils with special reference to Acari and Collembola. New Zealand Ecological Society Proceedings 13: 79-85. 108 Wood, T.G. 1967. Acari and Collembola of moorland soils from Yorkshire, England. III. The micro-arthropod communities. Oikos 18: 277-292. Zettel, J. 1982. Dietary influence on the biology and metabolism of Isotoma viridis (Collembola). Journal of Insect Physiology 28: 453-458. 109 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0088992/manifest

Comment

Related Items