UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Variation in carabid community structure associated with coastal Douglas-fir forest successional stages Craig, Katherine G. 1995

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

Item Metadata

Download

Media
831-ubc_1995-0332.pdf [ 3.02MB ]
Metadata
JSON: 831-1.0086840.json
JSON-LD: 831-1.0086840-ld.json
RDF/XML (Pretty): 831-1.0086840-rdf.xml
RDF/JSON: 831-1.0086840-rdf.json
Turtle: 831-1.0086840-turtle.txt
N-Triples: 831-1.0086840-rdf-ntriples.txt
Original Record: 831-1.0086840-source.json
Full Text
831-1.0086840-fulltext.txt
Citation
831-1.0086840.ris

Full Text

V A R I A T I O N IN C A R A B ID C O M M U N I T Y STRUCTURE ASSOCIATED WITH C O A S T A L DOUGLAS-FLR FOREST SUCCESSIONAL STAGES K A T H E R I N E G. CRAIG B.Sc , The University of British Columbia, 1987 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF M A S T E R OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES (DEPARTMENT OF ZOOLOGY) We accept this thesis as conforming "to, the required standard , THE UNIVERSITY OF BRITISH C O L U M B I A July 1995 © Katherine G. Craig, 1995 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 Jg^PcP/i The University of British Columbia Vancouver, Canada DE-6 (2/88) 11 A B S T R A C T Carabid beetles were collected by means of pitfall trapping in four forest successional stages: 1) Regeneration (3-8 years); 2) Immature (25-45 years); 3) Mature (65-85 years); and 4) Old-Growth (>200 years). The study was conducted at two locations, Victoria Watershed South, and Koksilah, in Coastal Douglas-fir forests on Vancouver Island. A total of 28 species was collected during the year of collecting. Intraspecific comparisons were made and six distributional patterns were identified. These are: 1) Regeneration specialists; 2) Generalists; 3) Forest species; 4) Recovering species; 5) Old-Growth specialists; 6) Unexpected pattern. A corrected species richness measure was calculated and showed the regeneration sites to have the greatest species richness. There was no replicated, significant difference among the other three stages. Other diversity measures are discussed, as are implications for the forest industry. T A B L E OF CONTENTS Pag Abstract i i Table of Contents i i i List of Tables iv List of Figures ; v Acknowledgments vi i i Introduction 1 Materials and Methods 6 Results 15 Discussion 32 Conclusions 48 Literature Cited 49 Appendix 56 iv LIST OF TABLES Table 1. Pitfall trap collection dates. Data collected after May 1993 has not been analyzed. Table 2. Total species collected, arranged according to distributional pattern. Table 3a. Number of species, corrected number of species E(S)3oo, with SD, Simpson index (1-D), Brillouin's index (H),and the Shannon-Wiener index (H1) for Victoria Watershed at all successional stages. Table 3b. Number of species, corrected number of species E(S)3oo, with SD, Simpson index (1-D), Brillouin's index (H),and the Shannon-Wiener index (H1) for Koksilah at all successional stages. Table 4. Number of species, Simpson index (1-D), Brillouin's index (H),and the Shannon-Wiener index (H1) comparing Victoria Watershed and Koksilah. Table 5. Horn's Index of Similarity Coefficients for Victoria Watershed and Koksilah at all successional stages. V LIST OF F I G U R E S Figure 1. Locations of the two forest chronosequences (referred to as Victoria Watershed South and Koksilah) where beetle collection for this study took place. (Reprinted with permission from Pollard and Trofymow 1993.) Figure 2. Example of the plot layout and subplot assignments provided by the Canadian Forest Service. Circled areas-numbered 1-8 were used for carabid beetle collection. (Reprinted with permission from Blackwell and Trofymow 1993.) Figure 3a. Mean number of individuals (with standard error) of Pterostichus amethystinus Mann, trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 3b. Mean number of individuals (with standard error) of Pterostichus amethystinus Mann, trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 4a. Mean number of individuals (with standard error) of Pterostichus lama (Men.) trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 4b. Mean number of individuals (with standard error) of Pterostichus lama (Men.) trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. vi Figure 5a. Mean number of individuals (with standard error) of Harpalus somnulentus Dej. trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 5b. Mean number of individuals (with standard error) of Harpalus cautus Dej. trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 6a. Mean number of individuals (with standard error) of Zacotus matthewsii LeC. trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 6b. Mean number of individuals (with standard error) of Zacotus matthewsii LeC. trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 7. Mean number of individuals (with standard error) of Carabus taedatus Fab. trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 8a. Mean number of individuals (with standard error) of Pterostichus herculaneus Mann, trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 8b. Mean number of individuals (with standard error) of Pterostichus herculaneus Mann, trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 9a. Mean number of individuals (with standard error) of Scaphinotus angusticollis (F.v Wald.) trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 9b. Mean number of individuals (with standard error) of Scaphinotus angusticollis (F.v Wald.) trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 10a. Mean number of individuals (with standard error) of Pterostichus algidus LeC. trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 10b. Mean number of individuals (with standard error) of Pterostichus algidus LeC. trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H , and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Figure 11. Dendrogram resulting from a single linkage cluster analysis of Horn's similarity coefficients. Figure 12. Fluctuation in numbers of P. herculaneus in the four successional stages at the Victoria Watershed. ACKNOWLEDGMENTS I would like to thank my committee members: 1) Dr. Geoffrey Scudder, my supervisor, for his support and guidance in all my endeavors, and for being one of the all-time great teachers. 2) Dr. Valin Marshall, for making this project possible. In addition, his assistance both in the field and with this manuscript are greatly appreciated. 3) Dr. Judy Myers, for her friendship and support over the years. I would also like to thank: Dr. George Ball, for his species identifications and pinning advice. Dr. Doug Currie, for teaching me the basics of beetle identification and Doug-speak. Dr. Rick Taylor and Shannon Bennett for their statistical help. Launi Lucas for her computer artistry. Bob Rowswell and Marilyn Clayton, at the Canadian Forest Service, for their field help. Stephen Connor who helped with every aspect of this work (and never complained). This work was supported by grants from: 1) The Natural Sciences and Engineering Research Council (to G.G.E. Scudder). 2) The Forestry Practices component of Forestry Canada's Green Plan. 1 I N T R O D U C T I O N Non-native forestry in British Columbia got its start over 200 years ago in Nootka Sound on the west coast of Vancouver Island when English and Spanish explorers replaced their ships' rotting masts with Douglas-fir. It was another 75 years before the first sawmill was built by the Hudson's Bay Company in 1848. It was a water-powered mill, located near Victoria, on Millstream, at the head of Esquimalt Harbour (Gold 1985). During this time timber licenses and leases were easily obtained, so the lumbermen and companies of the day selected sites with the best and most easily accessible timber. On the southern coast of Vancouver Island, sites in valley bottoms, at river mouths, or those with ocean access were especially sought after (Gould 1975). Since those early days, the forest industry has seen tremendous technological advances that now allow logging in areas previously regarded as physically inaccessible. Now the main impediments to the forest industry's harvesting of an area are lack of legal entitlement, poor timber quality, or economic infeasibility. About 100 years ago, silviculture practices such as clearcutting, slashburning, herbicide spraying, fertilizing, and seedling planting were brought from Europe to Canada. Today, around 90% of the trees harvested in British Columbia's public forests are harvested by clearcutting (Hammond 1993), and, according to Travers (1993), the rate of clearcutting is on the rise. Ministry of Environment figures to 1989 show that well over half of the timber ever cut in British Columbia had been cut in the previous 20 years (Travers 1993). There have always been naturally-occurring losses of habitat such as those from lightning fires, wind, disease, or insect outbreaks, and so it is argued by some that clearcutting is really no different (Pojar 1993; Hammond 1993). However, the scale of natural deforestation events tends to be smaller, and generally some trees, both as individuals and groups, survive. Further, the burned, diseased, or blown-over trees are not removed from the site, but left to decay where they fall, thereby returning nutrients (contained in the 2 boles of the trees) to the soil, and acting as physical barriers to erosion. Finally, natural disturbances are random both spatially and temporally, as compared to clearcut logging which exerts its effects over spatially broader areas and sustained periods of time, and results in the complete stripping of vegetation from affected areas, leaving them depleted of nutrients and vulnerable to erosion. (Hammond 1993). Although the concept of tree planting, or reforestation, was also introduced to Canada about 100 years ago, the effort to reforest has been ineffective in some areas and non-existent in others, leading to a backlog of Not Satisfactorily Restocked (NSR) lands. The exact extent of NSR lands is difficult to determine from the literature, but conservatively amounts to about 4% of British Columbia forests (some 874,000 hectares) according to 1988 figures released by the Ministry of Forests (in Thompson et al. 1992). In areas that are reforested, the seedlings typically come from nursery monocultures, or near-monocultures. This loss of genetic diversity renders the plantations potentially more susceptible to disease, and pest outbreak, and less able to adapt to environmental changes (Smith 1990). Perhaps more problematic for local biodiversity is the fact that the planted trees are of a single species, often not the species that dominated the pre-harvest area (Fritz 1989). Communities adapted to the pre-harvest forest may find it difficult, or impossible, to survive in the new monoculture plantations. In recent years the forest industry has come under increasingly sharp attacks for these utilization and management strategies. During the 1960's and 1970's there was an acceleration in development of ecological theories and the collection of data in specializations such as island biogeography, and population and community ecology. Around the same time the environmental movement was beginning to get the public's attention. Scientific evidence was accumulating to show that tropical deforestation and habitat destruction were leading to species' extinctions (Soule 1986a). The diversity of living things in tropical rain forests is recognized to be far greater than that in temperate forests (Wilson 3 1988; Soule 1986b). According to Wilson (1988) over 50% of the world's species are found in tropical rain forests, which only account for 7% of the Earth's terrestrial area. Not surprisingly then, deforestation in these areas has led, and continues to lead, to hundreds of extinctions each year. Species in temperate zones have also been driven to extinction. In many cases this occurred years ago as a result of both habitat destruction and hunting (Wilcove et al. 1986). Recently, the plight of the northern spotted owl, Strix occidentalis (Xantus de Vesey), found in western North American forests, has focused attention on the fact that i f present old-growth harvesting rates continue, the habitat of this species wil l be gone within the next few decades (Bart and Forsman 1992). Unfortunately, for every species about which we know something regarding how they are affected by current forestry practices, there are many times more species for which we have no information. This is particularly true of insects. While many people would probably be just as happy to have a world with fewer insects, they are nevertheless a vital part of most ecosystems (Samways 1993). The class Insecta is thought to contain significantly more species than any other class of animals (Wilson 1988; Samways 1993), and while it is estimated that less than 10% of insect species have been described, it is known that insects are essential for many processes on Earth. Though most people would recognize the role insects play in pollination, it is probably less widely appreciated that phytophagous insects can be involved in modifying plant communities, and helping with decomposition and soil fertility, while predacious insects keep other species in check. Lastly, all insects contribute to the food chain (Samways 1993). In other words, the abundance and distribution of terrestrial plants and animals depends to a large extent on insects (Miller 1993). The problem is that we have no idea which insects are vital parts of a given ecosystem, and which, i f any, are expendable, and yet we continue to destroy habitat, and the species therein, at an ever-increasing rate. 4 Forestry practices must be based on a thorough knowledge of the structure and dynamics of forest ecosystems. In order to accomplish this, fundamental research, beginning with identifying the species found in old-growth forests is needed. In Canada, Marshall et al. (1982) estimate that less than 50% of soil arthropod species have been described. Taxonomic information is only the first step. Further information about life histories, habitat ranges, and environmental interactions is needed before a complete understanding of the impact of clearcutting on forest biodiversity can begin to develop. This study was intended to assess how current forestry practices affect one small subset of insects, the ground-beetles (Coleoptera, Carabidae). Specifically, this study looks at how ground-beetle assemblages vary with forest successional stage. Carabids were selected for this work for five reasons. First, from a practical stand point, they are easy to catch in relatively large numbers with the use of simple, inexpensive pitfall traps (Winchester and Scudder 1993; Marshall et al. 1994). Second, much of the taxonomy of this family of beetles has been worked out for North American species. There are good keys available for identification to the species level, and expert systematists are available to help with more difficult identifications. Third, carabids are found world-wide, in relatively large numbers, suggesting their importance in ecosystem functioning (Thiele 1977). Fourth, they are ground-dwelling (at least most North American species). When a comparison is to be made between treed and non-treed sites, it helps to use a group whose members' life-cycles are not partially, or completely, arboreal. Finally, there is considerable interest and research underway, examining the use of carabids in monitoring the health of various habitats (for examples see Rushton et al. 1994; Haila et al. 1994; Van Essen 1994). According to Goodall (1970), ecological work falls into two categories: experimental, and observational. This study has both components. First, there is the experimental part having as its hypothesis that carabid beetle communities (defined herein as an assemblage of carabid species in a prescribed area) differ among four successional stages of coastal Douglas-fir forest. This will be looked at on a species level. As well, overall beetle 5 diversity wil l be compared among the four successional stages. Literature from various sources suggests that carabid beetle diversity will be highest in the recently disturbed regeneration stages, probably as a result of an increase in plant diversity, and then wil l drop off once the canopy has been re-established (Southwood 1978; Halme and Niemela 1993; Hailaetal. 1994). The second, observational part is an a posteriori look at the species distribution patterns discovered during this study. This will be coupled with known information about the habitat preferences and behaviour of the trapped beetle species. 6 M A T E R I A L S A N D M E T H O D S Study Areas Two locations on southeast Vancouver Island were selected as study sites. The first is located at the south end of Shawnigan lake and is referred to as the Victoria Watershed South (here after known as Victoria Watershed). The second location is southwest of Duncan and is known as Koksilah (Fig. 1). These locations were among several selected by the Canadian Forest Service in which a variety of studies were to simultaneously take place (for exact site locations and descriptions see appendix). Criteria for selection by the Canadian Forest Service included finding locations 5 km x 5 km (or smaller) within which were four ages of stands (referred to as a chronosequence). These stands were to have a similar slope, aspect, and elevation (within 200 m, and a mid slope below 600 m) (Pollard and Trofymow 1993). The chronosequence was defined, with 1990 as the reference year, as having the following successional stages: Regeneration (3-8 years), Immature (25-45 years), Mature (65-85 years), and Old-Growth (>200 years). "Years" refers to post-harvest and burning time, or, in the case of some mature stands, post-wildfire or landslide (Pollard and Trofymow 1993). The old-growth sites were treated as experimental controls. Both Victoria Watershed and Koksilah are in the Dry Coastal Western Hemlock biogeoclimatic zone (Meidinger and Pojar 1991; Pollard and Trofymow 1993). The dominant vegetation of this particular subzone is composed of Douglas-fir (Pseudotsuga menziesii (Mirb.)Franco), salal (Gaultheria shallon Pursh), step moss (Hylocomium splendens (Hedw.) B.S.G.), and Western hemlock (Tsuga heterophylla (Raf.) Sarg.) (Pojar et al. 1991). In each successional stage, a 60 m x 60 m plot was marked off by the Canadian Forest Service. Within this plot eight 10 m x 10 m subplots were allocated for this study (Fig. 2), (Blackwell and Trofymow 1993). 7 B I O G E O C L I M A T I C ZONES Coastal Western Hemlock * The western part of the Coastal Douglas-fir zone has recently been transferred to the Dry Coastal Western Hemlock zone (see Meidinger and Pojar 1991) Figure 1. Locations of the two forest chronosequences (referred to as Victoria Watershed South and Koksilah) where beetle collection for this study took place. (Reprinted with permission from Pollard and Trofymow 1993.) 8 Plot markers [3 5 c m x 5 c m x 1.5 m c e d a r s t a k e s with c o l o u r e d f l a g g i n g - d e f i n e s s u b p l o t centres a n d triangle c o r n e r s - 1 9 9 1 / 9 2 L 2 . 5 c m x 2 . 5 c m a n g l e a l u m i n u m - d e f i n e s c o m e r s of 6 0 x 6 0 m plot - 1 9 9 1 < 1 2 m m x 12 m m a n g l e a l u m i n u m - d e f i n e s m i d p o i n t s of 6 0 x 6 0 m p l o t - 1 9 9 1 • 2 . 5 c m x 2 . 5 c m x 2 m c e d a r s t a k e s p a i n t e d y e l l o w a n d white - 1 9 9 2 * 2 . 5 c m d i a m e t e r x 2 . 0 m b l u e - t i p p e d P V C p i p e a n d b r u s h e d trail - d e f i n e s inner 4 0 x 4 0 m plot - 1 9 9 2 Subplot assignments A-C F o r e s t floor, s o i l , w o o d y litter s a m p l i n g , d e c a y a n d m i c r o b i a l activity s t u d i e s (Trofymow, P r e s t o n ) A m p h i b i a c o v e r object ( D a v i s ) ft U n d i s t u r b e d c o r e area S u b p l o t for s o i l z o o l o g y , m i c r o f l o r a ( M a r s h a l l , Trofymow, P a n e s a r , G o o d m a n , C r a i g ) Q F i n e litterfall traps (to be p l a c e d - P o l l a r d ) '1-8 Figure 2. Example o f the plot layout and subplot assignments provided by the Canadian Forest Service. C i rc led areas numbered 1-8 were used for carabid beetle collection. (Reprinted with permission from Blackwel l and Tro fymow 1993.) 9 Traps and Trap placement In each of the eight 10 m x 10 m subplots three pitfall traps were randomly placed. This was accomplished by hypothetically dividing the subplot into one meter squares numbered one to one hundred and then using a random number table (from Zar, 1984) to select three squares. A pitfall trap was then placed in the centre of each of the selected squares. The traps themselves consist of two 450 ml (9 cm diameter opening and 35 cm in depth) disposable plastic drinking cups nested one inside the other. The two cups differed slightly from one another so that when nested their rims were flush with one another. The inner cup was made by Progress Plastic Disposables Ltd., the outer cup by Portion Packaging. The nested cups were placed into holes in the ground and fixed so that the cup rim was flush or slightly below the ground's surface as recommended by Winchester and Scudder (1993). Where the designated site for the trap was unusable (e.g. owing to an outcropping of rock, a tree, log, etc.), the cup was placed in the ground at the nearest accessible position. The inner cup was filled about 1/3 to 1/2 with a 35% solution of non-toxic propylene-glycol (made for veterinary use). Organisms falling into the cup would be killed (by drowning) and preserved. This prevents the problem of predation, where, for example, one large species may feed on smaller ones and thereby affect the study's outcome. This very non-volatile fluid allowed the cups to remain in the field for long periods of time. There is no evidence the propylene-glycol acted as a bait. Each trap was covered by a 19 cm by 19 cm rough-cut cedar roof to protect it from rain, snow, and debris. The roofs were raised 4 cm off the ground by four wooden "legs". These legs were placed at the corners of the roof and angled at 45° from each side. Organisms walking into the legs would be directed toward the pitfall cup. Traps were placed in the ground at Victoria Watershed on May 20, 1992, and at the Koksilah site on May 26, 1992. Table 1 shows the dates on which the traps' contents were 10 collected. In order to accommodate different species' phenologies, trapping and analysis had to involve a full year. Table 1. Pitfall trap collection dates.* Collection number Victoria Watershed Koksilah 1 June 17, 1992 June 23, 1992 2 July 15, 1992 July 28, 1992 3 August 12, 1992 September 1, 1992 4 September 9, 1992 October 7, 1992 5 October 6, 1992 November 25, 1992 6 November 24, 1992 April 7, 1993 7 April 6, 1993. May 12, 1993 8 May 11, 1993 June 9, 1993 9 June 8, 1993 July 7, 1993 10 July 6, 1993 August 18, 1993 11 August 17, 1993 Data collected after May 1993 have not been analyzed. To collect the contents of a trap, the inner cup (containing the propylene-glycol and trapped organisms) was removed from the ground. The propylene-glycol was strained through a fine-mesh, plastic strainer into a new inner cup which was then placed in the original undisturbed outer cup. The contents of the strainer were transferred to plastic specimen jars and covered with 70% ethyl alcohol as recommended by Martin (1977). These jars were then transferred to the laboratory at the University of British Columbia for sorting and identification. On each collection date, traps were inspected to make sure that the cups remained flush with the ground surrounding them. Additionally, extra propylene-glycol was added to cups showing signs of evaporation. Throughout the year the occasional outer cup would become brittle and cracked and would therefore need replacing. This was done as seldom as possible so as not to disturb the surrounding soil. 11 Over the course of the year, only 27 of the roughly 2000 cups collected were damaged. Damage seemed to be caused by either humans (i.e. other researchers stepping on the traps) or some curious animals. Several traps were ripped out of the ground and the cups showed what appeared to be claw marks. These damaged or missing traps were recorded and replaced with new cups. S o r t i n g a n d Identi f icat ion In the laboratory at the University of British Columbia, larger organisms such as shrews, salamanders, and frogs were removed and their presence recorded. Because of the time involved in sorting and identifying carabids from each individual cup, only about one year's-worth of the traps was ultimately used. From these almost 1300 samples, the ground-beetles were removed. Identification to the family Carabidae was done using the key by Borror et al. (1981). Initially, all carabids were removed from the alcohol and washed using a mild shampoo and water solution. The beetles were then direct-pinned or put on points (depending on their size), positioned and dried following the protocol outlined in Martin (1977). Beetles whose elytra or pronotums appeared greasy were gently wiped with a 10% ammonium hydroxide solution as suggested by Dr. George Ball, University of Alberta, Edmonton (personal communication). Pinned, dried beetles were individually labelled with the location, successional stage, trap number, latitude, longitude, range of dates the trap was in the field, and the collectors name, as recommended by Winchester and Scudder (1993). Pinned carabids were identified to species primarily using Lindroth's keys (1961, 1963, 1966, 1968, 1969). On occasion Hatch's (1953) keys were used as reference. Noonan's (1991) key to Harpalus was used to identify species in this difficult genus. In cases where a species was relatively easy to identify without being pinned, and had been trapped in large numbers, about 100 individuals were pinned; the remainder were identified, labelled and placed in vials of 70% ethyl alcohol. On two occasions a subsets of 12 pinned and tentatively identified beetles were sent to Dr. George Ball at the University of Alberta, in Edmonton, for his expert determination and confirmation of the identified species. Methods of Data Analysis: Missing Data As described earlier, 8 groups of 3 traps were laid within each successional stage at each location. For the purposes of statistical analysis, beetles collected by the three traps were pooled, so that in effect each successional stage now only had data from 8 traps. In the few cases (19 out of 1248) where a trap was damaged, the missing data were dealt with in two ways, depending on what the data were to be used for. For analysis of the intraspecific comparisons, the missing data were estimated as follows: If in one group of three traps one cup was damaged, the data from the other two were averaged and this average multiplied by three. In a couple of instances two of the three cups were missing - in this case the data from the single cup were multiplied by three. In no case were all three of the cups in one group damaged. In order to examine diversity and related concepts, the missing data could not be estimated and therefore the data were handled differently. For example, i f during one collection period two cups (out of 24) were destroyed in the regeneration site, but all 24 cups in each of the other three sites were fine, then with the use of a random number table, two out of the 24 cups from each of the other three sites were selected and those data were eliminated. This same procedure was used to ensure that the two locations (i.e. Koksilah and Victoria Watershed) were sampled evenly. 13 Intraspecific Comparisons For graphical presentation of the intraspecific comparisons, the data from the different collection dates were pooled. This resulted in eight numbers (one for each group of three cups) for each successional stage. The mean and standard error of these eight numbers for each successional stage, at each location, were plotted. A nonparametric analysis of variance test, the Kruskal-Wallis rank test, was used to test intersuccessional stage differences. This test is recommended in instances where the samples do not come from a normal population and/or the variances are heterogeneous. Both of these situations apply to the beetle data. The Kruskal-Wallis test statistic, H , was calculated and the probability, P, was noted. (The graphs and Kruskal-Wallis test were produced using the computer program Systat for Windows, version 5.05.) Species that showed a significant difference among stages were further tested with the nonparametric multiple comparison test to resolve exactly which stages differed significantly. These calculations were done by hand following the method outlined in Zar (1984). Species Diversity and Related Measures A statistical method known as rarefaction was used to estimate the number of species one would expect to find at each successional stage if all the samples were of a standard size (Magurran 1988; Krebs 1989; Ludwig and Reynolds 1988). A sample size of 300 was selected because it is approximately the size of the smallest group of individuals trapped at any one stage (320 individuals were trapped in the Koksilah regeneration site). (The computer program provided in Ludwig and Reynolds (1988) was used to compute these values.) Several of the so-called nonparametric diversity indices were calculated for each successional stage (Southwood 1978). Diversity indices for the two locations were also 14 calculated. These measures take into account both species richness and species evenness, consequently they are often referred to as heterogeneous measures (Magurran 1988; Krebs 1989). The Simpson index, 1-D, the Brillouin index, H , and the Shannon-Wiener index, Ff, were all calculated. (The computer package used to calculate this was Ecological Methods by Krebs 1989.) Similarity Coefficients and Cluster Analysis In order to assess how the different successional stages compare to one another, Horn's Index of Similarity was calculated (using the computer program Ecological Methodology by Krebs, 1989). This index was designed to be used for count data, and it has the advantage of not being biased by sample size (except in cases of very small samples) (Krebs 1989). Prior to this calculation the data were transformed by taking the square-root of (x+0.5), where x is the original count. This transformation is recommended for count data, and wil l effectively diminish the effect of very abundant species, while enhancing the effect of very rare species (Zar 1984). The resulting Horn's Similarity Matrix was then used in a cluster analysis program in order to group similar successional stages. A single linkage cluster analysis program (in Systat for Windows, V.5.05) was used to produce a tree diagram. Other For discussion purposes, the number of individuals of P. herculaneus caught per collection date in all four successional stages was graphed. 15 RESULTS A total of 28 species of Carabidae was identified from among the 6,849 beetles captured at the 8 plots sampled, over the year (Table 2). They represent 11 different tribes, and belong to 12 different genera. Eleven species were very rare (less than 10 individuals per species trapped); six of these species were represented by single individuals. Four species accounted for the majority (5,880) of individuals caught. Intraspecific comparisons Examination of the data suggests six patterns of beetle distribution among the four successional stages. This accounts for 13 of the 28 species caught. Patterns for the remaining species could not be resolved. As might be expected with field data, some trends are more tenuous than others. Generalists. The first, and weakest, pattern is from beetles I describe as generalists. For illustration purposes, I use two different species because in the case of the first species, Pterostichus amethystinus Mannerheim, only 17 beetles were ever caught; and for the second species, Pterostichus lama (Menetries) the pattern is not as clear. P. amethystinus (while in low numbers) does not appear to be affected by the different forest habitats provided by the different successional stages. This is confirmed statistically with the Kruskal-Wallis analysis of variance rank test (Fig. 3a & 3b). In the case of P. lama there are significant differences among stages, however, the two locations vary considerably as to which stages are different. The multiple comparison tests reveal that at the Victoria Watershed site the regeneration stage differed significantly from the immature and mature, but not the old-growth. At the Koksilah location, the immature stage differed significantly from the regeneration and the mature (Fig. 4a & 4b). This, coupled with the fact that these beetles were relatively abundant in all successional stages leads me to describe them as generalists. Table 2. T O T A L SPECIES C O L L E C T E D Species Victoria Watershed South Ret-, lnmi. Mat. O G. Generalists: 1 Pterostichus lama 51 165 264 128 (Menetries) 2 Pterostichus amethystinus 3 2 2 1 Mannerheim Regeneration specialists: 3 Harpalus somnulentus 26 -•' ' 0 0 0 Dejean 4 Harpalus cau tus D ej ean 1 0 0 0 5 Calathus fuscipes (Goeze) 17 0 0 0 6 Metabletus americanus 16 0 0 0 (Dejean) 7 Amara sinuosa (Casey) 3 0 0 0 8 Amara sanjuanensisHatch- 1 0 0 0 Old-growth specialists: 9 Zacotus matthewsii 0 1 12 30 LeConte 10 Carabus taedatus 0 0 0 0 Fabricius Recovering species: 11 Pterostichus herculaneus 103 196 185 305 Mannerheim Forest species: 12 Scaphinotus angusticollis 40 744 390 351 (Fischer von Waldheim) Unexpected pattern: 13 Pterostichus algidus 65 3 528 114 LeConte Unresolved species 14 Omus dejeani Reiche 1 5 1 7 15 Cychrus tuberculatus 3 3 4 5 Harris 16 Scaphinotus angulatus 0 1 2 15 (Harris) 17 Promecognathus crassus 12 4 0 13 LeConte 18 Scaphinotus marginatus 0 0 1 0 (Fischer von Waldheim) 19 Pterostichus crenicollis 0 4 2 0 LeConte 20 Notiophilus sylvaticus 0 0 0 2 Eschscholtz 21 Pterostichus pumilus 0 0 0 0 Casey 22 Pterostichus neobrunneus 0 0 0 0 Lindroth Koksilah Total Reg 1mm Mat. O.G. 39 156 21 45 869 3 1 2 3 17 2 0 0 0 28 17 0 0 0 18 1 0 1 0 19 1 0 0 0 17 3 0 0 0 6 4 0 0 0 5 0 0 9 41 93 21 7 34 280 342 56 56 183 154 1238 1 229 503 349 2607 106 0 172 178 1166 32 11 26 103 186 20 4 35 36 110 5 1 16 1 41 7 0 2 1 39 0 1 22 0 24 0 0 5 0 11 0 0 1 0 3 0 1 1 0 2 0 1 1 0 2 Table 2. T O T A L SPECIES C O L L E C T E D 17 Species Unresolved species: (cont'd.) 23 Pterostichus castaneus (Dejean) 24 Pterostichus melanarius (Illiger) 25 Amara littoralis Mannerheim 26 Trachypachus holmbergi Mannerheim 27 Harpalus cordifer Notman 28 Dicheirus piceus (Menetries) Victoria Watershed South Koksilah Total Reu. I in in. Mat . O . G . Reg. 1mm. Mat . O . G . 0 0 1 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 Total 343 1130 1392 971 320 468 1034 1191 6849 18 H P 0.351 0.950 M A T . EvTM. O.G. REG. Reg. Imm. Mat. O.G. Successional Stage Figure 3a Mean number of individuals (with standard error) of Pterostichus amethystinus Mann, trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. Reg. Imm. Mat. O.G. Successional Stage H P 1.419 0.701 LMM. M A T . O.G. REG. Figure 3b Mean number of individuals (with standard error) of Pterostichus amethystinus Mann, trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 19 H P 18.069 0.0005 REG. O.G. IMM. MAT. Reg. Imm. Mat. O.G. Successional Stage Figure 4a Mean number of individuals (with standard error) of Pterostichus lama (Men.) trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. H P 15.234 0.002 MAT. REG. O.G. IMM. Reg. Imm. Mat. O.G. Successional Stage Figure 4b Mean number of individuals (with standard error) of Pterostichus lama (Men.) trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 20 Regeneration specialists: Several species were found exclusively in the regeneration sites. Unfortunately, the number of individuals within these species was either quite low, or as in the case of the following two species, the number was different at the two locations. Harpalus somnulentus Dejean was found exclusively in the regeneration sites at Victoria Watershed and Koksilah, but in very low numbers at Koksilah. Harpalus cautus Dejean followed the same pattern but was found in low numbers at the Victoria Watershed site. The Kruskal-Wallis test statistic for H. somnulentus at Victoria Watershed shows a very significant probability of P < 0.0005. Similarly, the probability for H. cautus at Koksilah is significant at P = 0.001. While it is obvious that the difference being detected is between the regeneration stage and the other three stages, the Tukey-like multiple comparison test is unable to resolve this (Fig. 5a & 5b). Old-growth specialists: The old-growth specialist, Zacotus matthewsii LeConte, was virtually absent from the regeneration and immature stages. The mature sites had some individuals, but were not significantly different from the regeneration and immature stages. The old-growth forests were statistically different from the regeneration and immature stages, but not from the mature. This pattern was seen at both locations (Fig. 6a & 6b). Another species, Carabus taedatus Fabricius, also appears to be an old-growth specialist. This species has the same distribution pattern as Z. matthewsii, except that it was only trapped at the Koksilah site (Fig. 7). Some 342 individuals were caught over the year-long period at the Koksilah site, and not a single individual was caught at Victoria Watershed. Recovering species: The fourth pattern follows what one might expect to see after a major habitat disturbance. The number of individuals drops off after the disturbance and then gradually the species recovers. Pterostichus herculaneus Mannerheim shows this trend. Again, this is 21 H P 21.173 0.0005 IMM. MAT. O.G. REG. Reg. I m m . Mat. Successional Stage O.G. Figure 5a Mean number of individuals (with standard error) of Harpalus somnulentus Dej. trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as deterrnined by a nonpararmetric multiple comparison test. H P 17.072 0.001 IMM. MAT. O.G. REG. Reg. Imm. Mat. O.G. Successional Stage Figure 5b Mean number of individuals (with standard error) of Harpalus cautus Dej. trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 22 10 Reg. Mat. O.G. Successional Stage Imm. H P 17.075 0.001 REG. IMM. MAT. O.G. Figure 6a Mean number of individuals (with standard error) of Zacotus matthewsii LeC. trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 10 Figure 6b Mean number of individuals (with standard error) of Zacotus matthewsii LeC. trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 23 Figure 7. Mean number of individuals (with standard error) of Carabus taedatus Fab. trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 24 a case where the Kruskal-Wallis test indicates a significant difference among stages, yet the Tukey-like test could not resolve the difference at the Koksilah site (Fig. 8a & 8b). Forest species: Scaphinotus angusticollis (Fischer von Waldheim) best exemplifies the fifth pattern. This species was the most abundant species trapped (some 2,607 individuals in all). Figures 9a & 9b show that this species was quite uncommon in the regeneration forests. Unexpected pattern: The most unexpected pattern is that of Pterostichus algidus LeConte. This species is found in all but the immature stages. The multiple comparison test found that the immature differed from the mature and old-growth, at both sites, but, it was unable to resolve a difference between the regeneration and immature forests (Fig. 10a & 10b). Unresolved species: Nine of these species were trapped in very low numbers (three individuals or less), obviously making distributional pattern detection impossible (Table 2). Of the remaining six species, i f a pattern was detected at one location it was not replicated at the other, or the numbers of individuals trapped at the second location were so low that they effectively eliminated a replication. Species Diversity and Related Measures The results of the species richness index ( E { S 3 0 o } ) and the standard deviation (SD) produced by rarefaction are given in Tables 3a and 3b. It is clear that the regeneration stage at both locations has the highest corrected species richness. The next most specious environment, determined by rarefaction, depends on the location; at Victoria Watershed it is the old-growth, whereas, at Koksilah, it is the mature. There was agreement among the three heterogeneous diversity measures used (Tables 3a & 3b). The regeneration sites were the most diverse, followed in order by the old-growth, Figure 8a Mean number of individuals (with standard error) of Pterostichus herculaneus Mann, trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 2. 30 d 20 H P 12.292 0.006 1MM. REG. O.G. MAT. Reg. Imm. Mat. O.G. Successional Stage Figure 8b Mean number of individuals (with standard error) of Pterostichus herculaneus Mann, trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 26 120 100 H P 19.915 0.0005 REG. O.G. MAT. IMM. Reg. Imm. Mat. O.G. Successional Stage Figure 9a Mean number of individuals (with standard error) of Scaphinotus angusticollis (F.v Wald.) trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 120 100 H P 21.039 0.0005 REG. IMM. O.G. MAT. Imm. Mat. O.G. Successional Stage Figure 9b Mean number of individuals (with standard error) of Scaphinotus angusticollis (F.v Wald.) trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 27 100 Figure 10a Mean number of individuals (with standard error) of Pterostichus algidus LeC. trapped in each successional stage at the Victoria Watershed site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 100 H P 20.484 0.0005 LMM. R E G . O . G . MAT. Reg. Imm. ' Mat. O.G. Successional Stage Figure 10b Mean number of individuals (with standard error) of Pterostichus algidus LeC. trapped in each successional stage at the Koksilah site. The Kruskal-Wallis test statistic, H, and the statistical probability, P, are the result of intersuccessional stage comparisons. Stages which did not differ from one another are joined with a line, as determined by a nonpararmetric multiple comparison test. 28 Table 3a. Number of species, corrected number of species E(S)3oo, with SD, Simpson index (1-D), Brillouin's index (H),and the Shannon-Wiener index (Ff) for Victoria Watershed at all successional stages. V I C T O R I A W A T E R S H E D Regeneration Immature Mature Old-growth No. Species 15 13 12 11 E(S)300 14.49 7.60 7.37 9.59 SD 0.66 1.42 1.22 0.82 1-D 0.829 0.514 0.724 0.739 H 2.787 1.392 2.011 2.220 H ' 2.896 1.417 2.033 2.254 Table 3b. Number of species, corrected number of species E(S)3oo, with SD, Simpson index (1-D), Brillouin's index (H),and the Shannon-Wiener index (H) for Koksilah at all successional stages. K O K S I L A H Regeneration Immature Mature Old-growth No. Species 18 11 17 11 E(S)300 17.68 9.19 12.92 9.08 SD 0.54 1.07 1.23 0.79 1-D 0.825 0.636 0.701 0.809 H 2.865 1.719 2.290 2.614 H ' 2.999 1.771 2.335 2.644 Table 4. Number of species, Simpson index (1-D), Brillouin's index (H),and the Shannon-Wiener index (H1) comparing Victoria Watershed and Koksilah. Victoria Watershed Koksilah No. Species 23 24 1-D .740 .801 H 2.258 2.765 H ' 2.275 2.789 29 mature, and immature. This pattern held true for both locations. When compared to one another, the Koksilah site was more diverse than the Victoria Watershed site (Table 4). Similarity Coefficients and Cluster Analysis The calculated similarity coefficients (Table 5), and resulting dendrogram (Figure 11), show that the immature sites at both locations cluster together (indicating their similarity to one another). The old-growth and mature sites at each location are more similar to one another than, for instance, the two old-growth sites are to one another. After the initial clusters, the three successional stages (i.e. immature, mature, and old-growth) cluster together, with only the regeneration sites recognizably different. Finally, the two regeneration sites cluster together at the same time they cluster with all the other sites. Table 5. Horn's Index of Similarity Coefficients for Victoria Watershed (V.W.) and Koksilah (Kok.) at all successional stages. V . W . V . W . V . W . V . W . Kok. Kok. Kok. Kok. Reg. Imm. Mat. O.G. Reg. Imm. Mat. O.G. V . W . 1.00 .85 .87 .89 .89 .85 .85 .81 Reg. V . W . 1.00 .90 .94 .75 .96 .89 .84 Imm. V . W . 1.00 .96 .81 .88 .92 .87 Mat. V . W . 1.00 .84 .91 .94 .89 O.G. Kok. 1.00 .81 .86 .87 Reg. Kok. 1.00 .90 .88 Imm. Kok. 1.00 .95 Mat. Kok. 1.00 O.G. 30 SIMILARITIES 1.000 V.W. REG. KOK. REG. KOK. O.G. KOK. MAT. V.W. MAT.-V.W. O.G. V.W. IMM. KOK. IMM.-0.800 0.890 0.890 0.950 0.940 0.960 0.940 0.960 V.W. = Victoria Watershed Kok. = Koksilah Reg. = Regeneration Imm. = Immature Mat. = Mature O.G. = Old-Growth Figure 11. Dendrogram resulting from a single linkage cluster analysis of Horn's similarity coefficients. 31 Other Figure 12 shows the number of individuals of P. herculaneus caught per collection date in all four successional stages. Old-Growth Mature : Immature Regeneration Collection number Collection number refers to the following dates: 1 June 17, 1992 2 July 15, 1992 3 August 12, 1992 4 September 9, 1992 5 October 6, 1992 6 November 24, 1992 7 April 6, 1993 Figure 12. Fluctuation in numbers of P. herculaneus in the four successional stages at the Victoria Watershed. 32 DISCUSSION Limitations of Pitfall Trapping Before discussing the specific results of this study, it is necessary to examine the limitations of pitfall trapping. These limitations have a direct bearing on what information can and cannot be extracted from this study, and consequently, have directed the types of tests performed on data collected. Pitfall traps have been extensively used to collect soil arthropods, particularly Coleoptera, Arachnida, and Collembola (Marshall et al. 1994; Southwood 1978). The popularity of these traps surely results from their ease of use, minimal expense, and efficiency. In the early 1960's, work by Mitchell (1963), Greenslade (1964a) and others demonstrated that trap catches do not depend solely on the actual abundance of a species, but also depend on the activity of its members. This makes intuitive sense - the more a beetle runs around the greater the chance it will run into a trap. Further studies showed that a beetle's size affects the rate of capture, with larger beetles being more susceptible to trapping (Greenslade 1964a; Luff 1975). Small beetles, such as some Notiophilus species, are even able to stop at the edge of a trap and thereby avoid being captured (van der Drift 1951). Halsall and Wratten (1988) suggest that this is the mechanism that results in the correlation of large size with increased capture. That is, different capture rates have to do with the ability of different species to recognize the pitfall edge. They also dispel the notion of van der Drifts that diurnal species (as compared to nocturnal species) are less likely to be caught because they can see the traps. Pitfall traps have been shown to have other limitations as well. Luff (1986) looked at two species; Harpalus rufipes (Deg.) and Pterostichus madidus (F.) whose catches were significantly aggregated in pitfall traps. Laboratory studies indicate that a low concentration of the defensive secretion, formic acid, released by H. rufipes, was an attractant to other H. rufipes. In the case of P. madidus, it seems that a sex pheromone may be responsible for 33 species aggregation. Experiments where individuals of either sex were placed in a pitfall trap resulted in an increase in numbers of the opposite sex. Luff (1975) examined features of the traps themselves to see how such qualities as trap size, shape, and material affect beetle capture. His results suggest that shape of the trap (i.e. whether it is rectangular or circular) does not affect capture rates provided the perimeter is the same. As long as the amount of edge is equal between two differently shaped traps, the efficiency of the traps should also be equal. There is a potential problem with rectangular traps in cases where beetle movement is directional rather than random. If beetles are moving primarily in one direction, then these traps can over-, or underestimate the true activity, depending on whether the broad side or the narrow side of the trap is oriented towards the beetles. While there does not seem to be a logical reason for this, Luffs study also shows that there is a species-specific relationship between trap size and capture. In general smaller species are better caught with smaller traps. The material the trap is made of affects both the capture efficacy and the potential for escape. Glass, plastic, and metal respectively, were shown to be the preferred trap materials. Of course Luffs experiment was looking at retention capacity of these materials and so the traps did not contain preservatives. Preservatives such as formalin have been used to reduce potential escapes (Thiele 1977). In the case of formalin there is evidence that the formalin itself may act as an attractant to some species and a repellent to others (Skuhravy 1970; Luff 1968 ). It also seems that some small species of beetles may be able to escape from formalin (Petruska 1969). For traps that are to remain in the field for any length of time, a preservative, such as propylene glycol, is necessary to prevent predation and decomposition of specimens (Martin 1977). 34 The factor influencing carabid beetle capture that is perhaps the most problematic for this study is the possibility that differences in vegetation surrounding the pitfall traps can lead to differences in capture rates. Gfeenslade (1964a) set out pitfall traps in a uniform plot of grass. One set of traps was set so the lip of the trap (in this case a glass jar) was flush with the top of the grass. A second set was flush with the soil surface, and a third set was also flush with the soil but had a 60 cm diameter area of grass surrounding the traps removed. The result was that these different trap set-ups led to significantly different catches of carabid beetles. The trap with the swath of grass removed caught significantly more beetles than the other two types. Further comparisons among the types of beetles caught by the second and third set of traps revealed that there was no significant difference in the efficacy of the traps for catching species associated with plants. There was however, a significant difference in the capture efficacy of larger ground-dwelling species. When she looked at phytophagous species and divided them into larger and smaller species she found that the trap with the grass swath cut around it caught significantly less smaller phytophagous carabids than the traps just set flush with the soil. She suggests that different species of carabids travel in different layers of the vegetation and/or soil. From this she concludes that one should not compare relative carabid abundances in different environments. I contend that carabids in forest successional stages can be compared by pitfall trapping methods. The understory vegetation (or in the case of the clearcut areas, the vegetation) of all the forests is heterogeneous. At both locations, and at each of the four forest stands, there were areas of mosses, salal, barren rock, decaying wood, and exposed soil. In other words, even within one environment, the microhabitat variation was enormous. By placing a large number of traps (24 in each successional stage) randomly within assigned grids a wide variety of microhabitats was sampled. Greenslade's results show that large ground moving species are more susceptible to being caught in traps cleared of grass. In my study three of the four largest and most abundant species (Pterostichus lama, P. herculaneus, and Scaphinotus angusticollis) were 35 found in the lowest numbers in the regeneration sites. Of all the sites sampled the regeneration sites had the most herbs (Ryan and Fraser 1993). On face value then, this seems to lend support to Greenslade's work. However, there are several reasons why I believe that my results are a true reflection of the lower numbers of these beetles in the regeneration sites, as opposed to a problem with pitfall trap efficacy. First, i f S. angusticollis is taken as an example, there were 40 individuals trapped in the regeneration stage over approximately a one year period at the Victoria Watershed site, and one individual trapped in the regeneration stage over the same period at the Koksilah site. The next lowest number of beetles trapped in any of the other 6 plots was 229. The 40 beetles trapped at the Victoria site were almost all trapped by one group of three traps that was located at one corner of the plot. This location happened to be in a depression formed by the intersection of the low points of three knolls. Contrary to what might be expected (with respect to Greenslade's work) the site was very grassy, and as the year went on it was difficult to find the traps. I believe that more S. angusticollis were caught in these traps because it was the wettest area. Being at the bottom of the knolls it was more shaded and also received more run-off than the rest of the site. Even at this site no beetles were caught until September, when the rains began. Second, the regeneration sites also had large areas covered with shrubs (Ryan and Fraser 1993) and other areas lacking vegetation altogether. If S. angusticollis and some of the other species were able to avoid capture in the grassy areas, they surely would have been caught in the clearer areas. Yet, only one S. angusticollis, was caught in all of the 24 traps in the regeneration site at Koksilah over the one year. Third, I did trap several species of large ground-dwelling carabids (and one huge species of Tenebrionidae) in the regeneration sites in numbers equal to, or even greater than, numbers at other sites (e.g. Cychrus tuberculatus Harris and Pterostichus algidus LeConte). 36 Finally, i f the grasses hindered the trapping of the ground-dwelling beetles then one would expect the number of beetles caught to decrease as the grasses sprouted or began growing again after the winter. In fact, i f one looks at the number of individuals caught per collection of a species such as P. herculaneus in all four successional stages, one can see that, while the relative abundances are different among different stages, the pattern of the numbers caught over time is reasonably similar (Fig. 12). Their numbers do not decrease in the summer as the grasses grow. It may be that the time and way in which my traps were set led to a situation not that different from Greenslade's. The traps were dug into the ground in May before the grasses had really begun to grow. The soil removed from the ground to make room for the cups was patted around the edge of the cup to ensure the cup's rim was flush with the surrounding ground. Then on top of this was placed a wooden roof. Since the roof was opaque nothing was really able to grow underneath it. The roofs were only 19 cm x 19 cm, which is a far cry from the 60 cm swath used by Greenslade, but it may have been sufficient. Because carabid activity is affected by environmental factors such as temperature and moisture (Mitchell 1963), some feel that pitfall trapping is even unreliable at estimating relative abundances (Southwood 1978). Fortunately, work by Baars (1979) and den Boer (1971), suggests that pitfall trapping is reliable provided certain conditions are met. For intraspecific comparisons it is recommended that trapping should be continuous for the complete activity period of the beetles, or for one year. Traps should remain in the ground in the same physical arrangement throughout this period. Baars also advises that i f the microhabitat is very heterogeneous it may be better to use numerous traps. My research followed all of these recommendations. To calculate the absolute population sizes requires knowledge (for each species) of what the relationship is between species density and pitfall efficacy (Baars 1979). Quadrat sampling methods (such as litter washing or sieving a measured area) can be used to determine absolute carabid density, but, in order to gather enough data for any meaningful 37 analysis, are prohibitively labour intensive and habitat destructive (Thiele 1977; Spence and Niemela 1994). In a recent study, Spence and Niemela (1994) compared the relative abundance of species collected by litter washing and pitfall trapping. Interestingly enough, the litter washing seemed to miss larger-sized species of carabids which were however picked up by pitfall traps. The exact mechanism responsible for this discrepancy has not yet been determined. It seems that, despite their limitations, pitfall traps remain the best tool we have to carry out large scale carabid studies such as this one. As the absolute population densities remain unknown for the species in this study, only intraspecific comparisons have been made. Intraspecific comparisons Thiele (1977) credits Rober and Schmidt (1949) with being the first to clearly recognize that there are distinct differences between forest and field species of carabids. Since that time, in Europe, many researchers have contributed, so that there is a better understanding of which carabids inhabit which forest plant community. (See Thiele (1977) for a review of the carabids found in Central Europe.) North America seems to lag behind Europe in this type of research, although the amazing work of Lindroth (1961-1969), and to a lesser extent Hatch (1953) and others, has provided a foundation from which this work can proceed. The focus of research has changed, or at least a new dimension has been added, so that, as well as identifying broad species ranges, attempts are being made to recognize patterns among various species identified. For example, Taglianti and De Felici (1994) have grouped carabids found in beech woods in the Central Apennines (Italy) according to their ecological and faunistic preferences. In Canada, Niemela et al. (1992) compared beetles captured in four forest types and one meadow north of Edmonton, Alberta, and from this 38 were able to describe four distributional patterns for 23 of the 54 species they trapped. They identified: 1) Habitat generalists; 2) Forest generalists; 3) Forest specialists; and 4) Meadow species. Following along this vein, I have described six patterns of carabid distributions among the four successional stages. Regeneration specialists: The species that I call regeneration specialists include members of the genera Amara, Harpalus, Metabletus, and Calathus. While much of the basic biology of the west coast species is yet unknown, what is known is useful in understanding the distribution patterns found. According to Lindroth (1968), members of Amara have full, functional wings, and tend to occur in dry, open country with a tall, weedy vegetation type. The primarily phytophagous adults feed on fruits and seeds. Two (A. sinuosa and A. littoralis) of the three species collected have transcontinental distributions, the third (A. sanjuanensis) has a Pacific Northwest distribution. A. littoralis, he describes as favouring weedy environments such as those produced by human activity. Members of Harpalus are like Amara in their habitat preference and their phytophagous diet. H. somnulentus has a fully developed and functional hind wing. Members of the species H. cautus can exist with fully functional, or vestigial wings - a true wing dimorphism (Lindroth 1968; Noonan 1991). H. cautus was trapped in the regeneration sites primarily at Koksilah (with only one individual collected at Victoria Watershed), whereas H. somnulentus, again trapped exclusively in the regeneration sites, had the opposite pattern, and was caught in greater numbers at the Victoria Watershed. Noonan (1991) found from field studies that H. somnulentus tended to be found near moisture. Perhaps the Victoria Watershed regeneration site will prove to be wetter than the Koksilah regeneration site, explaining the distribution pattern of the two species. However, there are numerous other plausible explanations. Much research shows carabid beetles of various species to be sensitive to factors such as altitude, soil quality (e.g. Eyre and Luff 1994), moisture (e.g. Thiele 1977), food supply (e.g. Lenski 1982b; Hengeveld 1980), or temperature and light 39 (e.g. Mitchell 1963). Any of these factors may be responsible for differences of relative abundance of the two species at the two locations. It seems equally plausible that the difference in distribution of H. cautus and H. somnulentus may be nothing more than the normal stochasticity seen in early colonization. For example, work by Simberloff and Wilson (1970), in which they fumigated six small mangrove islands and then monitored their recolonization, shows that after two years the islands vary somewhat (although there is a lot of overlap) in their species makeup. Until specific research is done for these two species, the underlying mechanism behind their differences in abundance at the two sites wil l not be known. The two other regeneration specialists, C. fuscipes and M. americanus, again have characteristics similar to the Amara and Harpalus species. M. americanus has a transcontinental distribution, occuring in dry open areas with low vegetation. Populations of this wing-dimorphic species are usually comprised of mixed morphs (Lindroth 1968). C. fuscipes is an introduced species, found only in the southwest of British Columbia on open ground with weedy vegetation. Lindroth (1966, p. 543) describes it as "clearly synanthropic" in British Columbia. This is the only regeneration specialist to have exclusively vestigial wings. It is clear that the regeneration specialists as a group, generally share characteristics that make them adept at taking advantage of recently logged sites. A l l but one species is winged constantly or has a winged form. This has obvious implications with respect to their ability to disperse and find new suitable habitats. For the wing-dimorphic species it has been shown that large functional-winged (macropterous) individuals predominate in early successional stages and are replaced in time with short-winged (brachypterous) species (den Boer 1968; Neuman 1971; Holliday 1991). Generalists: Both species found to be generalists belong to the genus Pterostichus, and both species are constantly flightless (Lindroth 1966). Unfortunately, for these two species their 40 biology is not adequately known to explain how they are found in all environments. In fact the habitat preferences described by Lindroth (1966) are similar for both species - dark, dense coniferous forests, often under logs. He goes as far as to call P. amethystinus "a true forest species" (p. 465). This contradicts my finding for these species. Perhaps they only appear to be generalists; since they lack functional wings, their only way out of an undesirable habitat would be to walk. Logging areas are, of course, criss-crossed by roads. Studies have shown that some species of forest carabids (e.g. Abax ater), and some spiders, are unlikely to cross roads, even unpaved, rarely used ones (Mader 1984; Mader etal. 1990). Alternatively, specific information regarding their diets could provide the answer. Though details of the diets of these species are not known, most carabids are considered to be polyphagous predators. On a species by species basis, however, it can be shown by gut analysis that many species rely at least partially on plant matter for food (Skuhravy 1959; Hevengeld 1980). It may be that P. lama and P. amethystinus are able to live in a wide variety of habitats by being able to adapt to changing food sources. Old-growth specialists: Similar issues concerning barriers may be at work with the old-growth specialists. Z. matthewsii is typically found on moist ground in dense coniferous forest. It has a rudimentary hind wing, and is unable to fly (Lindroth 1961). There are several possibilities as to why this species is found predominantly in older forests. It could be that it too is reluctant to cross roads to reach other successional stages. Roads have quite different microclimates, the effects of which extend into the edges of the forest (Mader 1984). These beetles may be sensitive to change in factors such as temperature and moisture, and therefore wil l not cross. They may also be more prone to predation when crossing roads (Primack 1993). Lindroth (1961) reports that remnants of Z. matthewsii can be found in owl pellets. It is 41 possible that birds would have an easier time spotting beetles on open ground, than in dense forest. The other possibility is that there are inherent factors in old-growth forests that these species require. Studies at the H J . Andrews Experimental Forest in Oregon have also found this to be an old-growth species (Lattin 1993). Carabus taedatus is potentially identified as an old-growth specialist. My hesitation is due to the fact that the beetle was only trapped at the Koksilah site, and so there is no replication of data. Also, Lindroth (1961) describes this beetle as favouring dry, open, gravely soil, with sparse vegetation, or open coniferous forests. It is surprising that with this habitat description more C. taedatus were not found in the regeneration sites. Interestingly, Halme and Niemela (1993) found two species of Carabus, in Finland, to be restricted to contiguous forests. Again, this was in contrast to Lindroth's habitat descriptions for these species. Recovering species: The next two distributional patterns (i.e.: recovering, and forest species) may be variations in expression of a common, underlying mechanism. Intuitively, when forests are logged and slash burned, one would expect many of the resident species to be reduced in number i f not eliminated completely. Those that do survive the initial trauma may succumb to such factors as a decrease in available food, or interspecific competition from invading species they normally would not have to contend with. As well, being in a post-burn clearcut with few places to hide may make them more susceptible to predation. For forest beetles that remain viable after this dramatic alteration of habitat, one might expect a decrease in fecundity, and/or larval survival. It is not surprising then, to see a species whose numbers are greatly reduced in the clearcuts, but which seem to recover in number with time and canopy cover. Pterostichus herculaneus shows just such a pattern. 42 Forest species: Forest species such as S. angusticollis can be thought of as an extreme example of the recovering species. These beetles are unable, or unwilling, to reside in the clearcuts. This may be a result of dietary constraints. Members of the tribe Cychrini (to which S. angusticollis belongs) have very characteristic elongated heads. This is thought to be an adaptation that allows them to feed on snails (Lindroth 1961; Thiele 1977; Digweed 1993). While originally marine organisms, some gastropods have adapted to terrestrial environments, however, they are typically restricted to moist or humid areas (Russell-Hunter 1979). During summer months, the clearcuts become very hot and dry and are perhaps therefore unfavourable. This would then lead to a reduction in the beetles' food supply. Another consideration is that these beetles may be able to move the distance necessary to get them out of an undesirable habitat. Neumann (1971) found that the forest species Carabusproblematicus was able to travel greater than 70 metres over one night after being placed on open ground with forest in sight. Of all the beetles trapped, S. angusticollis was caught in greatest number (i.e. 2607 individuals) indicating its general abundance and activity. Maybe this active beetle is capable of walking the distance to get out of the regeneration sites, or to disperse from favourable refuges into sites which are once again favourable. Unexpected pattern: The single species P. algidus is described by the final distribution pattern identified, the unexpected pattern. This species appears to survive clearcutting with a slight reduction in number, but then by 25-45 years post cut (i.e. in the immature stage) virtually no beetles are detected. Sixty-five to 85 years after harvest, this species has recovered. Since the time span between the regeneration and immature stages is quite large it is hard to get a sense of what may be happening. Does the species abundance continually decline with time post harvest, until no beetles are left? Would sampling 15 years after cutting also reveal no beetles? Or is there something specific about the immature sites that is unfavourable for this 43 species? Niemela et al. (1993) found that some species of carabids survived the logging process, but their populations dwindled with time - taking up to nine years to disappear completely. Lindroth (1966) describes P. algidus as being less of a forest species (when compared to other Pterostichus species) - often found in open habitats with rich vegetation. This suggests that P. algidus should do fine in the regeneration sites, and leads one to believe that the problem lies in the immature stand. Unresolved species: Two of the unresolved species (Cychrus tuberculatus Harris and Omus dejeani Reiche) have been identified as old-growth arthropods in the H.J. Andrews Experimental Forest (Lattin 1993). In this present study, O. dejeani is found in greatest number in the old-growth sites at both locations, but overall the numbers are low at the Victoria Watershed site, so that there is no statistical difference among sites. Another replicate may have placed this species as an old-growth specialist, but with the data collected, this is only conjecture. C. tuberculatus does not appear to be restricted to the old-growth forests. However, its exact distributional pattern is not obvious. This species, along with two of the other unresolved species (Scaphinotus angulatus (Harris) and Scaphinotus marginatus (Fischer von Waldheim), belong to the tribe Cychrini, which, as mentioned above, feed on gastropods. Given this, one might expect them to be restricted to the forests as S. angusticollis appears to be. The limited data collected on C. tuberculatus, do not support this, as this species was found in the regeneration sites in reasonable numbers. More information on the dietary requirements of this species would be helpful. Forest Successional Stage Comparisons Species Diversity and Related Measures: In this study heterogeneous species diversity indices have been calculated for all sites using the relative abundance data from the pitfall traps. These indices must, however, be viewed with caution. They are based on two things: 1) the number of species (richness) and, 44 2) the number of individuals per species collected (evenness). Since the number of individuals trapped varies according to their activity, and not true abundance, the second parameter of these indices is affected. The species richness index calculated is technically a more appropriate measure for this type of data. The results of this index agree with the heterogeneous diversity measures for some, but not all, sites, therefore, again, care must be taken when interpreting the heterogeneous diversity measures. Clearly, the regeneration sites were the most diverse. This finding agrees with other carabid research from a variety of forest types (Lenski 1982a; Halme and Niemela 1993), although some have noted a concomitant reduction of diversity at the generic level (Lenski 1982a; Boyle 1991). The increase in diversity is postulated to be a result of the increase in habitat heterogeneity (Southwood et al. 1979; Haila et al. 1994; Halme and Niemela 1993), and the disruption of competitive exclusion (Lenski 1982 a & b). Of the remaining three forest stages, no trend is obvious. The species richness measures at Victoria Watershed are not significantly different among the three sites, and the apparent order differs from that of the Koksilah site. The heterogeneous diversity measures show that the old-growth sites are more diverse than the mature which in turn are more diverse than the immature sites. Similarity Coefficients and Cluster Analysis: The similarity coefficients, cluster analysis, and resulting dendrogram were done in an attempt to see if there was a pattern discernible among the eight sites. Unfortunately, like the diversity measures, only limited information can be gained. Carabid community structure does vary with successional stage, although the difference is most pronounced between forested and regeneration sites. The fact that the old-growth and mature stages at one location are more similar to one another than the two old-growths, or two matures are to each other is not surprising. As 45 mentioned above carabids are very sensitive to a number of environmental factors including elevation and moisture levels. The two locations do differ in elevation (Appendix 1) and may differ in other ways as well. The dendrogram is otherwise unremarkable, with the similar aged stands clustering together. Only the regeneration sites are clearly distinct. These sites are as different (or similar) to one another, as they are to the remaining six sites. Perhaps, this is further evidence of the stochastic nature of early colonization discussed above. Forest Management and Carabid Beetle Diversity It has long been a generally accepted tenet that diversity is good. The seemingly logical next step is that i f diversity is good, increasing diversity is better. Historically, Leopold (1933) recognized that humans could manage woodlots, fields, and ponds in such a manner that the complexity and species richness could be increased. This same principle was used by wildlife managers who were actually primarily concerned with game species (Harris 1984). They observed that where two different habitats meld there is an increase in the number of plant and animal species (Odum 1971; Wiens 1976; Wing 1951). This increase in species was termed an edge effect, and was originally thought of as a positive effect. Over the past thirty years or so, there has been mounting evidence to suggest that edge effects are not necessarily good (Angelstam 1992). The term is now used to apply to changes (such as in temperature, light, and wind) that occur in one habitat (e.g. a forest), when an adjoining area is abruptly altered (e.g. deforested), rather than just an increase in species (Lovejoy et al. 1986). In temperate forests, edge effects go hand-in-hand with habitat fragmentation, and both of these have an impact on diversity. Forest fragmentation is the process whereby large tracts of forested land are converted into a patchwork of smaller habitats unlike the original. The resulting patches of habitat are subject to a much greater area of edge (Wilcove et al. 1986; Primack 1993). Additionally, the diversity of species found in the remaining patches 46 or fragments of old-growth can also be affected. This has very much to do with, among other things, the particular organisms, their ability to disperse, their population size, the size of the remaining fragment, and the proximity to other fragments (Wilcove et al. 1986; Harris 1984). For example, many forest species require large areas of undisturbed habitat, and may fair poorly in remnant fragments of their original range (Harris 1984). It is now recognized that while on a global scale it is optimal to maintain maximal diversity, on a local scale this is not always a desirable goal. McNeely (1994) points to Hawaii as an example of this phenomenon. He suggests that Hawaii probably has more species now than it ever had, yet few would claim Hawaii as a triumph of wise diversity management. Most of the species present have been introduced through humanity's activities, to the detriment of the indigenous species. Counting numbers of species in this case ignores the large number of species found nowhere else that have been lost. In my study, while the regeneration sites were most diverse, and therefore represent an increase in local biodiversity, the species that make up this diversity are common species (many with transcontinental distributions), and are gained at the cost of many of the original forest species, leading to a net loss in global biodiversity. It has also been shown that, contrary to common wisdom, simple communities are often more stable than diverse ones (May 1973). This statement, and the results of the heterogeneous diversity measures, would therefore suggest that the old-growth sites are the least stable of the three forested sites. This is not likely the case. The increased diversity associated with instability (such as that seen around coral reefs) is thought to be the result of repeated mild to moderate environmental perturbations. The time between these disturbances is long enough to allow the establishment of communities, but short enough that equilibrium conditions are never reached (Connell 1978). Logging on the other hand, is a single, abrupt, catastrophic event. The brief explosion of diversity which follows is attributable, as stated earlier, to a rapid influx of specialists in the exploitation of disturbed habitats. Once the canopy begins to re-establish itself the original forest fauna begins the 47 process of recolonization. It is reasonable to conclude that the regeneration specialists disappear fairly early in this process. This, coupled with the fact that the forest species rely principally on walking for dispersal, (which means their return to the area is slow), leads to a transitional period of reduced diversity in the immature, and mature sites compared to either the regeneration, or old-growth sites. It seems, then, that little is gained from the calculation of any species richness or diversity measures without an understanding of the biology, and/or function of the various species involved. The diversity, and similarity measures also suggest that once the canopy is re-established carabid communities can recover to pre-harvest states. Yet, by examining the individual species of carabids it is evident that this is an over-simplification. An old-growth specialist like Z. matthewsii is just beginning to recover by sixty-five to eighty-five years (Fig. 6a & 6b). This type of information must be taken into account by the forest industry when determining the length of time between cuts. If the rotation time is too short for this beetle's population to recover completely, then there is a very real risk that this species wil l be lost. It is unique species such as Z. matthewsii that we must pay attention to. Why are they unique? What needs do they have that can only be satisfied in their particular environment? How large does such an area need to be in order to sustain these species? Though my study is concerned with beetles, these same questions will apply to any species unique to a habitat, regardless of the kingdom to which they might belong. Forestry is one of British Columbia's most important industries, and as such it is not likely to, nor is it desirable that it should, disappear. At the same time the practice of forestry cannot go on, as it has in the past, simply regarding forests as so much inventory waiting to be shipped to market. Good, objective reasons exist for making large changes in the management of our forests, and studies such as this one are necessary first steps to putting forestry management on a sound footing. 48 C O N C L U S I O N S 1) Pitfall traps, while having a number of limitations are the best technique we currently have to do large scale studies on carabid beetles. 2) Carabid community structure does vary with successional stage. a) Clearcut logging increases species diversity and richness of carabid beetles. Much of this increase is a result of winged, early colonizing, opportunistic species, that take advantage of the open habitat. Once the canopy has been re-established these species disappear. b) Some species found in old-growth forests are lost or have their numbers reduced when the habitats are cut. This process of loss may take a number of years to become evident, or may show up quite rapidly. c) Of those species which are reduced in number or lost, some recover once a canopy is re-established, while others appear unable to fully recover some sixty-five to eighty-five years later. Therefore, the process that leads to an increase in species diversity in the recently logged sites may ultimately cause a decrease in global diversity. 3) Further research on the biology of west coast carabid species would help considerably when trying to understand the underlying determinants of beetle distribution. 49 L I T E R A T U R E C I T E D Angelstam, P. 1992. Conservation of Communities - The Importance of Edges, Surroundings and Landscape Mosaic Structure. In Ecological Principles of Nature Conservation. Edited by L. Hansson. Elsevier Science Publishers Ltd. London, pp. 9-70. Baars, M . A . 1979. Catches in pitfall traps in relation to mean densities of carabid beetles. Oecologia (Berl.) 41, pp. 25-46. Bart, J., and Forsman, E D . 1992. Dependence of northern spotted owls Strix occidentalis caurina on old-growth forests in the western U.S.A. Biological Conservation 62 pp. 95-100. Blackwell, B .A. and Trofymow, J.A. 1993. An Overview of the Stand Characteristcs of the Intensively Studied Chronosequence Plots. In Proceedings of the Forest Ecosystem Dynamics Workshop February 10-11, 1993. Compiled by V. Marshall. Forestry Canada and B.C. Ministry of Forests, Victoria B.C. pp. 10-15. Borror, D.J., De Long, D .M. , Triplehorn, C A . 1981. An Introduction to the Study of Insects, fifth edition. Saunders College Publishing, Philadelphia, pp. 356-456. Boyle, T.J.B. 1991. Biodiversity of Canadian forests: Current status and future challenges. The Forestry Chronicle 68, pp. 444-453. Connell, J.H. 1978. Diversity in tropical rain forests and coral reefs. Science 199, pp. 1302-1310. den Boer, P.J. 1968. Zoologisch onderzoek op het Biologisch Station Wijster, 1959-1967. Misc. Papers Landbouwhogesch. Wageningen 2, pp. 161-181. (Original not consulted) den Boer, P.J. 1971. Stabilization of animal numbers and the heterogeneity of the environment: The problem of the persistence of sparse populations. Proceedings of Advanced Study Institute on Dynamics Numbers Population. (Oosterbeek 1970), (Edited by P.J. den Boer, G.R. Gradwell) Wageningen: Pudoc. pp. 77-97. (Original not consulted) Digweed, S.C. 1993. Selection of Terrestrial Gastropod Prey by Cychrine and Pterostichine Ground Beetles (Coleoptera: Carabidae). The Canadian Entomologist 125, pp. 463-472. Eyre, M.D. and Luff, M . L . 1994. Carabid species assemblages of North-East England woodlands. In Carabid Beetles: Ecology and Evolution. Edited by K . Desender et al. Kluwer Academic Publishers, Netherlands, pp. 277-281. 50 Fritz, E.C. 1989. Clearcutting: A Crime Against Nature. Eakin Press. Austin, Texas. Gold, W. 1985. Logging As It Was; A Pictorial History of Logging on Vancouver Island. Morriss Publishing Ltd. Victoria, British Columbia. Goodall, D.W. 1970. Statistical plant ecology. Annual Review of Ecology and Systematics 1, pp. 99-124. Gould, E. 1975. Logging; British Columbia's Logging History. Hancock House Publishers Ltd. Saanichton, British Columbia, Canada. Greenslade, P. J .M. 1964a. Pitfall trapping as a method for studying populations of Carabidae (Coleoptera). Journal of Animal Ecology 33, pp. 301-310. Greenslade, P. J .M. 1964b. The distribution, dispersal and size of a population of Nebria brevicollis (F.), with comparative studies on three other carabidae. Journal of Animal Ecology 33, pp. 311-333. Haila, Y . , Hanski, I.K., Niemela, J., Punttila, P., Raivio, S., and Tukia, H. 1994. Forestry and the boreal fauna: matching management with natural forest dynamics. Annales Zoologici Fennici 31, pp. 187-202. Halme, E., and Niemela, J. 1993. Carabid beetles in fragments of coniferous forest. Annales Zoologici Fennici 30, pp. 17-30. Halsall, N .B . and Wratten, S.D. 1988. The efficiency of pitfall trapping for polyphagous predatory Carabidae. Ecological Entomology 13, pp. 293-299. Hammond, H. 1993. Forest Practices: Putting wholistic forest use into practice. In Touch Wood: B.C. Forests at the Crossroads. Edited by K. Drushka, B. Nixon, R. Travers. Harbour Publishing Co. Ltd., Madeira Park, B.C. pp. 96-136. Harris, L .D. 1984. The Fragmented Forest. The University of Chicago Press, Chicago. Hatch, M . H . 1953. The Beetles of the Pacific Northwest: Parti: Introduction and Adephaga. University of Washington Press, Seattle Washington. Vol.16, pp. 1-340. Hengeveld, R. 1980. Polyphagy, Oligophagy and Food Specialization in Ground Beetles (Coleoptera, Carabidae). Netherlands Journal of Zoology 30 (4), pp. 564-584. Holliday, N.J. 1991. Species Responses of Carabid Beetles (Coleoptera: Carabidae) During Post-Fire Regeneration of Boreal Forest. The Canadian Entomologist 123, pp. 1369-1389. 51 Krebs, C.J. 1989. Ecological Methodology. Harper & Row, Publishers, New York. Lattin, J.D. 1993. Arthropod Diversity and Conservation in Old-Growth Northwest Forests. American Zoologist 33, pp. 578-587. Lenski, R.E. 1982a. The impact of forest cutting on the diversity of ground beetles (Coleoptera: Carabidae) in the southern Appalachians. Ecological Entomology 7, pp. 385-390. Lenski, R.E. 1982b! Effects of Forest Cutting on Two Carabus Species: Evidence for Competition for Food. Ecology 63, pp. 1211-1217. Leopold, A . 1933. Game Management. Scribners, New York. Lindroth, C H . 1961. The ground-beetles of Canada and Alaska. Opuscula Entomologica, Supplementum X X , pp. 1-200. Lindroth, C H . 1963. The ground-beetles of Canada and Alaska. Opuscula Entomologica, Supplementum X X I V , pp. 201-408. Lindroth, C H . 1966. The ground-beetles of Canada and Alaska. Opuscula Entomologica, Supplementum X X I X , pp. 409-648. Lindroth, C H . 1968. The ground-beetles of Canada and Alaska. Opuscula Entomologica, Supplementum XXXIII , pp. 649-944. Lindroth, C H . 1969. The ground-beetles of Canada and Alaska. Opuscula Entomologica, Supplementum X X X I V , pp. 945-1192. Lovejoy, T.E., Bierregaard Jr., R.O., Rylands, A.B. , Malcolm, JR., Quintela, C.E., Harper, L . H . , Brown Jr., K.S., Powell, A .H . , Powell, G.V.N. , Schubart, H.O.R., and Hays, M . B . 1986. Edge and Other Effects of Isolation on Amazon Forest Fragments. In Conservation Biology: The Science of Scarcity and Diversity. Edited by M.E . Soule, Sinauer Associates, Inc. Sunderland, Massachusetts, pp. 257-285. Ludwig, J.A., and Reynolds, J.F. 1988. Statistical Ecology, a primer on methods and computing. John Wiley & Sons, Inc., New York. Luff, M . L . 1968. Some effects of formalin on the numbers of Coleoptera caught in pitfall traps. Entomologist's Monthly Magazine 104, pp. 115-116. Luff, M . L . 1975. Some features influencing the efficiency of pitfall traps. Oecologia (Berl.) 19, pp. 345-357. 52 Luff, M . L . 1986. Aggregation of Some Carabidae in Pitfall Traps. In Carabid Beetles: Their Adaptations and Dynamics. Edited by P.J. den Boer, M . L . Luff, D. Mossakowski, andF. Weber. Gustav Fischer, Stuttgart, pp. 385-397. Mader, H.J. 1984. Animal Habitat Isolation by Roads and Agricultural Fields. Biological Conservation 29, pp. 81-96. Mader, H.J., Schell, C , and Kornacker, P. 1990. Linear Barriers to Arthropod Movements in the Landscape. Biological Conservation 54, pp. 209-222. Magurran, A .E . 1988. Ecological Diversity and Its Measurement. Princeton University Press, Princeton, New Jersey. Marshall, S.A., Anderson, R.S., Roughley, R.E., Behan-Pelletier, V. , and Danks, H.V. 1994. Terrestrial Arthropod Biodiversity: Planning a Study and Recommended Sampling Techniques. Entomological Society of Canada, 26 (1) Supplement, pp. 1-33. Marshall, V . G . , McE. Kevan, D.K., Mathews, Jr., J.V., and Tomlin, A.D. 1982. Status and research needs of Canadian soil fauna. Bulletin of the Entomological Society of Canada, Supplement 14, pp. 1-5. Martin, J.E.H. 1977. The Insects and Arachinds of Canada Part 1: Collecting, Preparing, and Preserving Insects, Mites, and Spiders. Agriculture Canada, Minister of Supply and Services Canada. May, R M . 1973. Stability and Complexity in Model Ecosystems. Princeton University Press, Princeton. McNeely, J.A. 1994. Lessons from the past: forests and biodiversity. Biodiversity and Conservation 3, pp. 3-20. Meidinger, D. and Pojar, J. (Editors). 1991. Ecosystems of British Columbia. Special Report Series no. 6. Research Branch, B.C. Ministry of Forests, Victoria, B.C. pp. xi i + 330. (Original not consulted) Miller, J.C. 1993. Insect natural history, multi-species interactions and biodiversity in ecosystems. Biodiversity and Conservation 2, pp. 233-241. Mitchell, B. 1963. Ecology of two carabid beetles, Bembidion lampros (Herbst) and Trechus quadristriatus (Schrank).II. Studies on populations of adults in the field, with special reference to the technique of pitfall trapping. Journal of Animal Ecology 32, pp. 377-392. 53 Neumann, U . 1971. Die Sukzession der Bodenfauna (Carabidae-Coleoptera, Diplopoda und Isopoda) in den forstlich rekultivierten Gebieten des Rheinischen Braunkohlenreviers. Pedobiologia 11, pp. 193-226. (Original not consulted) Niemela, J., Spence, J.R., and Spence, D.H. 1992. Habitat Associations and Seasonal Activity of Ground-Beetles (Coleoptera, Carabidae) in Central Alberta. The Canadian Entomologist 124, pp. 521-540. Niemela, J., Langor, D., and Spence, JR. 1993. Effects of Clear-cut Harvesting on Boreal Ground-beetle Assemblages (Coleoptera: Carabidae) in Western Canada. Conservation Biology 7, pp. 551-561. Noonan, G.R. 1991. Classification, Cladistics, and Natural History of Native North American Harpalus Latreille (Insecta: Coleoptera: Carabidae: Harpalini), Excluding Subgenera Glanodes and Pseudophonus. The Thomas Say Foundation Vol.XIII. Entomological Society of America. Lanham, Maryland. Odum, E P . 1971. Fundamentals of Ecology. Saunders, Philadelphia, Pennsylvania. Petruska, F. 1969. On the possibility of escape of the various components of the epigaeic fauna of the fields from the pitfall traps containing Formalin (Coleoptera). Acta Universitatis Palacklanae Olomucensis Facultas Rerum Naturalium Biologica 31, pp. 99-124. (Original not consulted) Pojar, J. 1993. Terrestrial Diversity of British Columbia. In Our Living Legacy: Proceedings of a Symposium on Biological Diversity. Edited by M . A . Fenger, E .H. Miller, J.F. Johnson and E.JR. Williams. Royal British Columbia Museum, Victoria, B.C. pp. 177-190. Pojar, J., Klinka, K. , and Demarchi, D.A. 1991. Chapter 6: Coastal Western Hemlock Zone. In Ecosystems of British Columbia. Edited by D. Meidinger and J. Pojar. B.C. Ministry of Forests, Crown Publications Inc., Victoria, B.C. pp. 95-111. Pollard, D.W.F., and Trofymow, J.A. 1993. An Introduction to the Coastal Forest Chronosequences. In Proceedings of the Forest Ecosystem Dynamics Workshop February 10-11, 1993. Compiled by V. Marshall. Forestry Canada and B.C. Ministry of Forests, Victoria, B.C. pp. 5-7. Primack, R.B. 1993. Essentials of Conservation Biology. Sinauer Associates Inc. Sunderland, Massachusetts. Rober, H , Schmidt, G. 1949. Untersuchungen iiber die raumliche und biotopmaBige Verteilung einheimischer Kafer (Carabidae, Silphidae, Geotrupidae, Necrophoridae). Natur und Heimat 9, pp. 1-19. (Original not consulted) 54 Rushton, S.P., Wadsworth, R.A., Cherrill, A.J. , Eyre, M.D. , and Luff, M . L . 1994. Modelling the consequences of land use change on the distribution of Garabidae. In Carabid Beetles: Ecology and Evolution. Edited by K. Desender et al. Kluwer Academic Publishers, Netherlands, pp. 353-360. Russell-Hunter, W D . 1979. A Life of Invertebrates. Macmillan Publishing Co., Inc. New York. Ryan, M.W., and Fraser, D.F. 1993. Changes in Plant Diversity in Douglas-fir Stands Following the Conversion of Old Growth to Second Growth. In Proceedings of the Forest Ecosystem Dynamics Workshop February 10-11, 1993. Compiled by V . Marshall. Forestry Canada and B.C. Ministry of Forests, Victoria, B.C. pp 16-20. Samways, M.J. 1993. Insects in biodiversity conservation: some perspectives and directives. Biodiversity and Conservation 2, pp. 258-282. Simberloff, D.S., and Wilson, E.O. 1970. Experimental Zoogeography of Islands. A Two-Year Record of Colonization. Ecology 51, pp. 934-937. Skuhravy, V . 1959. Die Nahrung der Feldcarabiden. Acta Societatis Entomologicae Cechosloveniae 56, pp. 1-18. (Original not consulted) Skuhravy, V . 1970. Zur Anlockungsfahigkeit von Formalin fur Carabiden in Bodenfallen. Beitraege zur Entomologie 20, pp. 371-374. (Original not consulted) Smith, S.M. 1990. The Greening of the Forest: Forest Pest Management into the 21st Century. Proceedings of the Entomological Society of Ontario 121, pp. 49-59. Soule, M.E . 1986a. Conservation Biology and the "Real World". In Conservation Biology: The Science of Scarcity and Diversity. Edited by M.E. Soule, Sinauer Associates, Inc. Sunderland, Massachusetts, pp. 1-12. Soule, M.E . 1986b. Patterns of Diversity and Rarity: Their Implications for Conservation. In Conservation Biology: The Science of Scarcity and Diversity. Edited by M.E . Soule, Sinauer Associates, Inc. Sunderland, Massachusetts, pp. 117-121. Southwood, T.R.E. 1978. Ecological Methods; with particular reference to the study of insect populations, second edition. Chapman and Hall, London. Southwood, T.R.E., Brown, V . K . , and Reader, P .M. 1979. The relationships of plant and insect diversities in succession. Biological Journal of the Linnean Society 12, pp. 327-348. 55 Spence, J.R., and Niemela, J.K. 1994. Sampling Carabid Assemblages with Pitfall Traps: The Madness and the Method. The Canadian Entomologist 126, pp. 881-894. Taglianti, A . V . , and De Felici, S. 1994. Ground beetle communities in Central Apennines beech woods. In Carabid Beetles: Ecology and Evolution. Edited by K . Desender et al. Kluwer Academic Publishers, Netherlands, pp. 71-78. Thiele, H.-U. 1977. Carabid Beetles in Their Environments. Springer-Verlag, Berlin. Thompson, W.A. , Pearse, P.H., Van Kooten, G.C., and Vertinsky, I., 1992. Rehabilitating the backlog of unstocked forest lands in British Columbia: a preliminary simulation analysis of alternative strategies. In Emerging Issues in Forest Policy. Edited by P.N. Nemetz. U.B.C. Press, Vancouver, B.C. pp. 99-130. Travers, O.R., 1993. Forest Policy: Rhetoric and Reality. In Touch Wood: B.C. Forests at the Crossroads. Edited by K. Drushka, B. Nixon, R. Travers. Harbour Publishing Co. Ltd., Madeira Park, B.C. pp. 171-224. van der Drift, J. 1951. Analysis of the animal community of a beech forest floor. Tijdeschrift vor Entomologie 94, pp. 1-168. (Original not consulted) Van Essen, S.J., 1994. A method to evaluate the condition of heathlands by using catches of carabid beetles. In Carabid Beetles: Ecology and Evolution. Edited by K . Desender et al. Kluwer Academic Publishers, Netherlands, pp. 383-386. Wiens, J.A. 1976. Population responses to patchy environments. Annual Review of Ecology and Systematics 7, pp. 81-120. Wilcove, D.S., McLellan, C.H., and Dobson, A.P. 1986. Habitat Fragmentation in the Temperate Zone. In Conservation Biology: The Science of Scarcity and Diversity. Edited by M.E . Soule, Sinauer Associates, Inc. Sunderland, Massachusetts, pp. 237-256. Wilson, E.O. 1988. Biodiversity. Edited by E.O. Wilson. National Academy Press, Washington, D.C. pp. 3-18. Winchester, N .N . , and Scudder, G.G.E., 1993. Methodology for Sampling Terrestrial Arthropods in British Columbia. Resources Inventory Committee. B.C. Wing, L .W. 1951. Practice of Wildlife Conservation. Wiley, New York. Zar, J.H. 1984. Biostatistical Analysis, second edition. Prentice Hall, Inc., Englewood Cliffs, New Jersey. 56 APPENDIX SITE LOCATIONS Location Successional Latitude Longitude IITM System Stage Zone Easting Northing Regeneration 48 33 51.9 123 38 55.4 10 452250 5379000 Victoria Immature 48 33 51.9 123 38 55.4 10 452400 5378800 Watershed Mature 48 34 15 123 39 45 10 451100 5379700 Old Growth 48 33 44.6 123 38 53.2 10 452600 5378500 Regeneration 48 39 25 123 45 51.4 10 443600 5389400 Koksilah Immature 48 39 41.2 123 46 10 10 443850 5389850 Mature 48 39 20 123 44 50 10 444500 5389500 Old Growth 48 39 30 123 45.50 10 444500 5389500 SITE DESCRIPTION Location Successional Aspect Slope Elevation Stage Regeneration 50° 15% 280 m. Victoria Immature 20° 40% 305 m. Watershed Mature 315° 11.4% 240 m. Old Growth 30° 40% 390 m. Regeneration 170° 15% 595 m. Koksilah Immature 170° 15% 710 m. Mature 210° 35% 590 m. Old Growth 180° 15% 630 m. (Data provided by the Canadian Forest Service) 

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.831.1-0086840/manifest

Comment

Related Items