Effects of Clearcut Logging on Organic Matter and Invertebrate Drift Exported From Headwater Streams in the Interior of British Columbia Catherine A. Mackay B.Sc., Simon Fraser University, 1997 Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Master o f Science in Natural Resources and Environmental Studies (Biology) The University o f Northern British Columbia August 2006 © Catherine A. Mackay, 2006 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Library and Archives Canada Bibliotheque et Archives Canada Published Heritage Branch Direction du Patrimoine de I'edition 395 Wellington Street Ottawa ON K1A 0N4 Canada 395, rue Wellington Ottawa ON K1A 0N4 Canada Your file Votre reference ISBN: 978-0-494-28409-4 Our file Notre reference ISBN: 978-0-494-28409-4 NOTICE: The author has granted a non­ exclusive license allowing Library and Archives Canada to reproduce, publish, archive, preserve, conserve, communicate to the public by telecommunication or on the Internet, loan, distribute and sell theses worldwide, for commercial or non­ commercial purposes, in microform, paper, electronic and/or any other formats. AVIS: L'auteur a accorde une licence non exclusive permettant a la Bibliotheque et Archives Canada de reproduire, publier, archiver, sauvegarder, conserver, transmettre au public par telecommunication ou par I'lnternet, preter, distribuer et vendre des theses partout dans le monde, a des fins commerciales ou autres, sur support microforme, papier, electronique et/ou autres formats. The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. L'auteur conserve la propriete du droit d'auteur et des droits moraux qui protege cette these. Ni la these ni des extraits substantiels de celle-ci ne doivent etre imprimes ou autrement reproduits sans son autorisation. In compliance with the Canadian Privacy Act some supporting forms may have been removed from this thesis. Conformement a la loi canadienne sur la protection de la vie privee, quelques formulaires secondaires ont ete enleves de cette these. While these forms may be included in the document page count, their removal does not represent any loss of content from the thesis. i *i Bien que ces formulaires aient inclus dans la pagination, il n'y aura aucun contenu manquant. Canada Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPROVAL Name: Catherine A. Mackay Degree: Master o f Science Thesis Title: Effects of Clearcut Logging on Organic Matter and Invertebrate Drift Exported From Headwater Streams in the Interior o f British Columbia Examining Committee: Chair: Dr. Robert Tait Dean o f Graduate Studies University o f Northern British Columbia r Supervisor: Dr. Mark Shrimpton, Associate Professor Natural Resources and Environmental Studies Program University o f Northern British Columbia Committee Member: Dr. JEllen Petticrew, Associate Professor Natural Resources and Environmental Studies Program University of^Njprthem Br^fsh Columbia Comm ittejS^em ben Dr. Brian Heise, Associate Professor Department o f Natural Resource Sciences Thompson Rivers University External Examiner: Dr. Lisa Poirier, Assistant Professor Natural Resources and Environmental Studies Program University o f Northern British Columbia Date Approved: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Abstract We examined streams within clearcuts 1-3 years and 5-7 years post-harvest, and within mature forest (control), to determine the effects of clearcut logging on headwater streams in the interior of British Columbia. We measured fine particulate organic matter (FPOM) concentration during high discharge, dissolved organic carbon (DOC) concentration and DOC composition using spectral ultraviolet absorbance (SUVA) during high and low discharge seasons, and invertebrate drift densities during high discharge. During high discharge, FPOM increased in response to 1-3 year old clearcuts and recovered 5-7 years post harvest, whereas DOC and SUVA did not differ among treatments. During low discharge, DOC did not differ among treatments but SUVA responded positively to new clearcuts indicating an increased proportion o f allochthonous DOC. Between seasons, DOC declined in all treatment groups and SUVA increased except in old clearcuts. Invertebrate drift density increased in response to both clearcut groups, and significant responses occurred in seventeen o f 52 biotic metrics. Aquatic, terrestrial, Ephemeroptera, Diptera, collector-gatherer and scraper densities responded positively to clearcuts, whereas percent shredders and taxa richness density responded negatively. Densities o f common taxa were positively related to FPOM in two cases and did not depend on DOC or harvest history. Our results indicated that invertebrate drift and organic matter exported from small streams were altered 1-7 years post-harvest, and that different responses occurred between the post harvest periods for both abiotic and biotic datasets. With respect to fisheries management, short-term gains in the quantity of resources that sustain the trophic base occurred. The shift in composition o f these resources indicates that stability of this export from headwater streams is uncertain. ii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents A bstract.......................................................................................................................................ii Table of Contents...................................................................................................................... iii List of Tables.............................................................................................................................. v List of Figures............................................................................................................................vi Acknowledgements................................................................................................................... ix Prologue........................................................................................................................................1 Chapter 1: Effects of clearcutting on fine particulate organic matter (FPOM) and dissolved organic carbon (DOC) within headwater streams in British Columbia...................................................................................................................................... 5 Abstract.................................................................................................................................... 5 Introduction............................................................................................................................. 6 M ethods................................................................................................................................. 11 Study Location................................................................................................................. 11 Experimental D esign....................................................................................................... 16 Data Collection................................................................................................................. 21 Data A nalysis................................................................................................................... 22 R esults...................................................................................................................................24 Stream Physical Characteristics..................................................................................... 24 Pooling Study A reas........................................................................................................ 25 Treatment Response........................................................................................................ 25 Seasonal R esponse.......................................................................................................... 26 Treatment x Seasonal Response..................................................................................... 26 Discussion............................................................................................................................. 29 Treatment Response........................................................................................................ 29 Seasonal R esponse.......................................................................................................... 31 Treatment x Seasonal Response..................................................................................... 33 Experimental Approach...................................................................................................36 Summary........................................................................................................................... 37 Chapter 2: Effects o f clearcutting on invertebrate drift within headwater streams in British Columbia, during two periods post harvest.............................................39 Abstract.................................................................................................................................. 39 Introduction........................................................................................................................... 41 M ethods.................................................................................................................................47 Study Location................................................................................................................. 47 Experimental D esign....................................................................................................... 50 Data Collection....................................................................................... 52 Data A nalysis....................................................................................................................55 iii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. R esults................................................................................................................................... 63 Biotic Metrics................................................................................................................... 63 Biotic Metrics: Effect (%) Compared Among and Between Treatments...................65 EPT Families and Genera............................................................................................... 76 Discussion............................................................................................................................. 79 Biotic M etrics................................................................................................................... 79 EPT Families and Genera.......................................................................................i...... 87 Experimental Approach................................................................................................... 8 8 89 Implications and Summary....................................................................................... Epilogue..................................................................................................................................... 91 References.................................................................................................................................94 Appendix................................................................................................................................. 107 iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Tables Table 1. Physical characteristics and sampling dates of study streams sampled in 2004 within the Bone Creek and Damfino Creek study areas.................................... 20 Table 2. Mean DOC concentration (mg/L), specific UV absorbance (SUVA 3 5 0 , L m ' 1 m g'1) and FPOM concentration (mg/L) by treatment and study area summarized for downstream and upstream sites among treatments. Standard errors o f the mean are given in parentheses......................................................... 27 Table 3. Physical characteristics and sampling dates of study streams sampled in 2004 within the Bone Creek and Damfino Creek study areas.................................... 54 Table 4. Definitions o f the 52 biotic metrics calculated based on the designations for each taxon (Appendix 1) identified from drift samples in Bone Creek and Damfino Creek. Biotic metrics are organized into three assessment categories: a) Densities, (b) Biotic indicators, and (c) Diversity...........................................57 Table 5. Structure matrix of the two factor solution for organic matter variables. Oblimin rotation with Kaiser normalization was used........................................ 63 Table 6 . Mean biotic metrics (SEM) for downstream sites at Bone and Damfino Creek study areas among treatments. Category A biotic metrics are summarized.... 64 Table 7. Mean biotic metrics (SEM) in downstream sites at Bone and Damfino Creek study areas among treatments. Category B metrics that are inherent from Table 6 are omitted, and all category C metrics are summarized. HBI = Hilsenhoff Biotic Index.......................................................................................... 65 Table 8 . Biotic metrics that differed for comparisons among (ANOVA) and between (contrasts) treatments o f downstream-upstream effect in Bone and Damfino Creek study areas. Mean effect represents the difference between sites as a percentage................................................................................................................ 6 6 Table 9. Representation of EPT families and genera among treatments in drift samples at Bone Creek and Damfino Creek study areas. Dots indicate presence of genera listed under each family............................................................................. 78 Table A l. List o f taxa and their designations as aquatic (A), terrestrial (T), clinger (CLG=Y); in functional feeding groups predator (PR), collector-gatherer (CG), collector-filterer (CF), scraper (SC), shredder (SH); their Hilsenhoff biotic index value (HBI); and their status as tolerant (TOL=T), and long-lived (Longlived=Y). Biomass formulas estimate dry mass (DM) in mg using body length (L) measurements.................................................................................................. 107 v Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Figures Figure 1. Small stream ecosystem showing energy flow, denoted by arrows. Detritus comprises coarse, fine particulate, and dissolved organic matter (CPOM, FPOM, DOM); rounded boxes show generalized groups o f biota (from Allan 1995).......................................................................................................................... 2 Figure 2. Location of study areas within British Columbia................................................14 Figure 3. Two examples of slash in proximity to streams in this study compared to two control streams; all photographs were taken in 2004. Top left photograph is stream 215 logged in 2003, and top right photograph is stream 206 logged in 1998; both streams are within Damfino Creek study area. Bottom left photograph is control stream 013 at Bone Creek study area, and bottom right photograph is control stream 203 at Damfino Creek study area........................ 15 Figure 4. Schematic diagram showing experimental design.............................................. 17 Figure 5. Hydrographs for (a) Bone Creek and (b) West Kettle River showing mean daily discharges. (Water Survey o f Canada, 2006)........................................... 19 Figure 6. (a) Two discriminant functions describe stream abiotic characteristics and do not predict group membership to the three treatments (Wilks’ X=0.45, p=0.19). (b) Mean crown closure at downstream sites does not differ among treatments (p=0.12); error bars represent SEM. Control n=8, old clearcuts n=6, new clearcuts n=8........................................................................................................... 26 Figure 7. Columns show mean effect for (a) FPOM, (b) DOC and (c) specific absorbance (SUVA 3 5 0 ) among treatments during high flow; error bars denote SEM. Positive effect indicates an increased value, whereas negative effect indicates a decrease from upstream to downstream sites. Treatment means with common letter do not differ significantly (p>0.05). (Control n=8, Old n=6, New n=8)........................................................................................................ 27 Figure 8. Mean effect for (a) DOC concentration and (b) specific absorbance (SUVA 350 ) among treatments during low flow season; error bars denote SEM. Positive effect indicates an increased value, whereas negative effect indicates a decrease from upstream to downstream sites. Treatment means with common letters do not differ significantly (p>0.05)...........................................................28 Figure 9. Mean seasonal differences at downstream (impacted) sites for DOC concentration at (a) Bone Creek, (b) Damfino Creek; and for specific absorbance, SUVA 350 at (c) Bone Creek and (d) Damfino Creek. Error bars denote SEM. DOC decreased during low flow in every treatment indicated by negative values, whereas SUVA 350 tended to increase from high to low flow. Bone Creek: Control n=4, Old n=2, New n=3. Damfino Creek: Control n=4, Old n=4, New n=4.................................................................................................. 28 Figure 10. (a) Mean DOC concentration and (b) mean specific absorbance (SUVA 350 ) during high flow and low flow sampling. Error bars represent SEM. Treatment means with a common letter do not differ significantly (p>0.05). (DOC n=39; SUVA350 n=37)................................................................................ 29 vi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 11. Mean seasonal effect among treatments for (a) DOC concentration and (b) specific absorbance (SUVA 350 ); error bars denote SEM. DOC concentration decreased from high to low flow indicated by negative effect values. SUVA 350 increased in control and new, and decreased in the old clearcut treatment. Control n= 8 , Old n= 6 , New n=7........................................................................... 29 Figure 12. Schematic diagram showing the flow o f energy from allochthonous sources and light to heterotrophs and macroinvertebrates............................................... 35 Figure 13. Locations of Bone Creek and Damfino Creek study areas within the southern interior o f British Columbia.................................................................................. 50 Figure 14. Schematic diagram illustrating the experimental design....................................51 Figure 15. Densities of (a) total invertebrates, (b) aquatic invertebrates, and (c) terrestrial invertebrates found in control, new and old clearcut drift samples in Bone and Damfino Creek study areas. Abundance densities are on the left, biomass densities are on the right. Control mean effects are open columns (n= 8 ), old clearcuts are hatched (n= 6 ), new clearcuts are grey (n= 8 ), and error bars represent SEM. A positive effect indicates that drift densities increased from the upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05).............................................................................................69 Figure 16. Densities of (a) Ephemeroptera, (b) Plecoptera, (c) Trichoptera, and (d) Diptera found in control (n= 8 ), old (n= 6 ) and new (n= 8 ) treatment drift samples in 2004. Abundance densities are on the left, biomass densities are on the right. A positive effect indicates drift densities increased from the upstream to downstream sites. Error bars represent SEM. Means that share a letter do not differ significantly (p>0.05).............................................................70 Figure 17. Densities o f (a) EPT and (b) EPT/D found in control (n= 8 ), old (n= 6 ) and new (n= 8 ) treatment drift samples in Bone and Damfino Creek study areas. Error bars represent SEM. Abundance densities are on the left, biomass densities are on the right. A positive effect indicates that drift densities increased from the upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05).............................................................................................71 Figure 18. Densities o f (a) predators, (b) collector-gatherers, (c) collector-filterers, (d) scrapers, and (e) shredders in control (n= 8 ), old (n= 6 ) and new (n= 8 ) drift samples in Bone and Damfino Creek. Abundance densities are on the left, biomass densities are on the right. Positive effect indicates that drift densities increased from upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05). Error bars represent SEM.....................................72 Figure 19. Percentage of total invertebrates made up of (a) aquatic invertebrates, (b) terrestrial invertebrates, (c) Ephemeroptera, (d) Plecoptera, (e) Trichoptera, and (f) Diptera in control (n= 8 ), old (n= 6 ) and new (n= 8 ) treatment drift samples in Bone and Damfino Creek study areas. Error bars represent SEM. A positive effect indicates that drift densities increased from the upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05)................................................................................................................... 73 vii with permission of the copyright owner. Further reproduction prohibited without permission. Figure 20. Percentage of groups by abundance, followed by a biotic index in control (n=8), old (n=6) and new (n=8) treatment drift samples in 2004. Groups are: (a) predators, (b) collector-gatherers, (c) collector-filterers, (d) scrapers, (e) shredders, (f) clingers, (g) non-insects, (h) long-lived, and (i) intolerant. Biotic index is (j) Hilsenhoff Biotic Index. Positive effect indicates that drift densities increased from upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05). Error bars represent SEM 74 Figure 21. Mean effect among control (n=8), old (n=6) and new (n=8) treatment drift samples in 2004 of: (a) total taxa richness; taxa diversity density o f the following groups: (b) total taxa, (c) Ephemeroptera, (d) Plecoptera, (e) Trichoptera, and (f) clingers; and biotic indices measuring: (g) heterogeneity, reciprocal of Simpson’s index and (h) evenness, Smith and W ilson’s index. Positive effect indicates that metrics increased from upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05). Error bars represent SEM.........................................................................................................75 Figure 22. FPOM as PCA Component 2 was significantly related to (a) Chloroperlidae density at Bone Creek and (b) Ameletus density at Damfino Creek study area. ..................................................................................................................................79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Acknowledgements I acknowledge Brian Heise, Ellen Petticrew, and Mark Shrimpton for their input, guidance and reviews that greatly improved this thesis. Brian Heise initiated this project and inspired my interest in the topic. Mark Shrimpton’s support was invaluable and allowed me to undertake the degree at UNBC. Thank you to Lisa Poirier for her meticulous review as external examiner. The Forest Science Program provided funding and a number of additional organizations provided in-kind contributions or assisted me during this project: Thompson Rivers University, National Water Research Institute, Department of Fisheries and Oceans, Forrex, Ministry o f Forests, Ministry of Environment, Canfor-Vavenby, Meeker Log and Timber, Tolko Industries and Weyerhaeuser. I am grateful to several folks in particular for their advice and assistance: Max Bothwell, Sharon Brewer, Denise and Tim Clark, Eric Medina, Jacque Sorensen, Dana Taylor, and Viva Wolf. The support of my husband, Shon, was critical to pursuing this degree. Thank you for maintaining the lifestyle to which I ’m accustomed and for encouraging me! I am grateful to the Greenhalgh’s and Mackay’s for expressing their interest in my topic. Carol and Bill Greenhalgh lead the way in graduate studies and inspired my curiosity for ecology at a young age. Thank you to Lome, Mary and Warren Mackay for facilitating my mobile lab. ix Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Prologue The focus of this study was to determine if streamside clearcut logging affected organic matter and invertebrate drift exported from non-fish bearing, subalpine streams in the interior o f British Columbia. Among freshwaters, forested headwater streams have the closest link between terrestrial and aquatic ecosystems, the highest degree of canopy closure, relatively low primary production, and cool water temperatures (Vannote et al. 1980). Small streams have an important element of large woody debris that allows retention of suspended particles and that mediates discharge (Murphy and Meehan 1991). The result o f these characteristics is that headwater streams export unique resources that subsidize downstream reaches (Vannote et al. 1980; Wipfli and Gregovich 2002). Decline o f allochthonous inputs, increased primary productivity, increased water temperatures and decreased large woody debris occur in the initial years following clearcut logging adjacent to small streams (Brown and Krygier 1970; Webster et al. 1983; Bilby and Bisson 1992; Macdonald et al. 2003a). Consequently, forestry activities alter the characteristics that influence small streams’ subsidy to larger streams, yet non-fish bearing streams have the least amount of protection in terms of a no-harvest riparian buffer (Young 2000; Cummins and Wilzbach 2005). Detritus forms the base o f small stream food webs (Fig. 1) and is categorized based on particle size and solubility as coarse particulate organic matter (CPOM >lmm), fine particulate organic matter (FPOM 0.45 pm - 1 mm), and dissolved organic matter (DOM <0.45 pm). DOM comprises the largest pool of organic carbon in freshwaters, an estimated 45-50 % organic carbon by mass (Allan 1995). We used dissolved organic carbon (DOC) as a measurement o f DOM and use the terms interchangeably, as is common in the literature. Allochthonous sources of organic matter are leachate of leaf 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. litter, groundwater, surface flow, and subsurface flow, while autochthonous sources include extracellular release from primary producers, and metabolic by-products of organisms (Lock 1990; Allan 1995). Heterotrophic consumers such as bacteria and fungi are an important means of energy transfer in the food web through their use of DOM. The microbial loop is a pathway whereby energy is essentially recycled among bacteria and successively larger heterotrophs that both consume and exude DOM (Allan 1995). Predators invertebrates, fish Primary Producers Plant Inputs leaf litter, organic debris Other inputs rain, upstream, overland, subsurface bryophytes, algae, cyanobacteria C >l Detritus CPOM, FPOM, DOM Invertebrate Consumers i> herbivores, detritivores Decomposers/ Heterotrophs Microbial bacteria, fungi, protozoa .Loop \ Figure 1. Small stream ecosystem showing energy flow, denoted by arrows. Detritus comprises coarse, fine particulate, and dissolved organic matter (CPOM, FPOM, DOM); rounded boxes show generalized groups o f biota (from Allan 1995). Invertebrates in small streams subsist on detritus directly, or on consumers of detritus. It is useful to categorize macroinvertebrates into functional feeding groups as shredders, collector-gatherers, collector-filterers, scrapers, and predators (Merritt and Cummins 1996). Shredders consume leaf litter and play an important role in breaking down CPOM to FPOM and DOM, after fungi and bacteria decomposers initially colonize the material. Decomposers supply shredders with nutrient value that may surpass that of their substrate (Cummins 1974). Collector-gatherers and collector-filterers directly consume FPOM and DOM, while scrapers consume biofilm, that is a gel matrix of algae, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. bacteria, fungi, and detritus that forms on substrates (Giller and Malmqvist 1998). Predators, including groups of macroinvertebrates and fish, consume all o f the former groups of consumers. Reduction of terrestrial litter inputs (Wallace et al. 1997) and reductions in forest cover (England and Rosemond 2004) can affect headwater stream biota by reducing energy transfer through bottom up forces in the food web. In addition to the amount o f litter inputs, the quality of organic matter affects energy transfer in the food web by influencing microbial activity (Stelzer et al. 2003) and macroinvertebrates (Hawkins et al. 1982), and predators may rely on a specific composition of invertebrate drift (Bilby and Bisson 1992). The type of organic matter and invertebrates available to predators affects trophic energy transfer because autochthonous derived organic matter and the biota that consume it generally have a higher nutritional quality than allochthonous matter due to lower carbon: nitrogen ratios in the former (Murphy and Meehan 1991). Under the legislation o f the Forest Practices Code of British Columbia Act, implemented in 1995, proximity of logging adjacent to streams was determined by stream categories such as interior versus coastal streams, and stream classifications based primarily on fish presence and bank-full width (Province o f B.C. 1995). In 2004, the province implemented the Forest and Range Practices Act, and regulations governing forest harvesting are currently in transition from the Forest Practices Code to the new legislation. The essence of this transition is a shift from basing decisions on broad categories of streams to placing responsibility on forestry professionals to ensure that specific criteria are met as a result o f their planning and operations. The goal o f this 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. research is to provide scientific knowledge in a forest type that currently lacks research, and thereby provide tools to strengthen decision-making for resource managers. Although it is known that clearcut logging can reduce overall organic matter inputs and change their composition (Bilby and Bisson 1992), and alter invertebrate drift (Culp and Davies 1983), the majority of research to date has focussed on interior and coastal fish bearing streams. We identified a knowledge gap for the Englemann spruce subalpine fir (ESSF) biogeoclimatic zone. The ESSF covers 12 % of British Columbia and is the highest elevation forested zone in the southern and central interior of the province (DeLong and Meidinger 2003). Published research in this forest type include clearcut logging effects on hydrology in Colorado (Troendle and King 1985), hydrology research in British Columbia (Winkler et al. 2005), and effects of harvesting on benthic invertebrates in British Columbia (Heise 2001). We chose to examine a dataset in this forest type that also lacks research: the combination of invertebrate drift and organic matter. These variables directly contribute to trophic resources exported from headwater streams. In addition to assessing quantities o f these elements, we assessed composition of the drift community, and quality o f DOC by measuring the spectral ultraviolet absorbance (SUVA) as an estimate o f the proportion of allochthonous DOC. Chapter 1 addresses the abiotic variables, FPOM and DOC, and Chapter 2 examines effects of clearcut logging on invertebrate drift. By examining these variables in headwater streams, this research will expand knowledge for a forest type that is ecologically and economically important, but has previously been little studied. 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 1: Effects of clearcutting on fine particulate organic matter (FPOM) and dissolved organic carbon (DOC) within headwater streams in British Columbia. Abstract Fine particulate organic matter (FPOM) and dissolved organic carbon (DOC) form the majority of organic matter transported in headwater streams. We examined effects of clearcutting on FPOM (mg I / 1) and DOC, and effects of clearcutting on seasonal shifts in DOC. We measured DOC concentration, and estimated its composition by spectral UV absorbance (SUVA). We selected streams within Englemann spruce subalpine fir (ESSF) forests in British Columbia in three harvest history treatment groups: “new” 1-3 years post harvest, “old” 5-7 years post harvest, and control streams within mature forests. At each stream, we compared differences between paired sites for FPOM and DOC, and seasonal change relative to high flow for DOC. During high flow, FPOM increased in response to new clearcuts, and recovered 5-7 years post-harvest. DOC concentration and SUVA showed similar trends but were not significant. During low flow, SUVA responded positively to new clearcuts indicating increased allochthonous produced DOC. Seasonally, DOC concentration declined in every treatment and SUVA increased in new clearcuts and controls, but declined in old clearcuts. We suggest that increased photoreactive DOC combined with reduced microbial activity lead to a seasonal increase in the labile pool of DOC in old clearcuts. In new clearcuts, photoreactivity combined with elevated nutrient availability lead to a seasonal reduction in labile DOC. These results indicate increased FPOM, and elevated DOC occurred in new clearcuts that recovered in 5 years post harvest. Responses in DOC composition were only evident 5-7 years post-harvest during the low flow sampling period. 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Introduction Due to the “fish-centric” nature o f riparian forest management, small streams that are non-fish bearing often have the least protection in terms o f a no harvest riparian buffer (Young 2000; Cummins and Wilzbach 2005). However, fish such as salmonids are seasonally food limited and may receive an important subsidy o f organic matter from headwater reaches, especially when considering their cumulative export (Wipfli and Gregovich 2002). It is important to consider cumulative impact because small streams exist in the highest numbers among the continuum of freshwaters and have the closest link between terrestrial and aquatic ecosystems (Vannote et al. 1980). Headwater streams are generally detrital based food webs that rely on a net import of allochthonous organic matter derived from the terrestrial ecosystem. A high degree of canopy closure results in relatively low primary production and cool water temperatures. Small streams also have an important element of large woody debris that allows retention of suspended particles and mediates discharge (Toews and Moore 1982). The result of the combination of these elements is that headwater streams export a unique component of organic matter to subsidize downstream trophic webs. Streamside timber harvest has varied effects on small stream ecosystems. The main effects include increased nutrients (Gregory et al. 1987), increased primary production (Webster et al. 1983), increased stream temperatures (Brown and Krygier 1970; Macdonald et al. 2003a), and declines o f allochthonous inputs (Bilby and Bisson 1992). These responses are predominant in the first decade post harvest and extend until successional vegetation becomes established. In the longer term, reduced recruitment of large woody debris alters stream morphology, leading to declines of woody debris dams, 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. decreased streambed stability and decreased channel complexity (Gregory et al. 1987). Annual water yields and peak flows generally increase following harvest and can persist for up to three decades post harvest in interior subalpine forests of Colorado (Troendle and King 1985). In combination with changes in stream morphology, increased water yields can lead to increased transport of organic matter in the initial years following harvest, and for decades post harvest (Webster et al. 1983; Troendle and King 1985). Small streams in northern interior forests may respond differently to disturbance (Moore and Wondzell 2005). Yet, among research conducted in the Pacific Northwest Region, which includes the entire province o f British Columbia, little research has investigated effects of forestry in the interior o f the province. Macdonald et al. (2003a, 2003b) reported increased temperature and suspended sediment in headwater streams o f the subboreal spruce biogeoclimatic zone and Winkler et al. (2005) reported hydrologic effects of logging in the montane spruce zone in the southern interior of BC. The Englemann spruce subalpine fir (ESSF) biogeoclimatic zone covers 12 % of the province and is the highest elevation forested zone in the southern and central interior of British Columbia (DeLong and Meidinger 2003). Englemann spruce (Picea englemanni) is one o f the most important commercial species in the interior due to its wood characteristics, and excellent source of pulp (Parish et al. 1996). Despite its ecological and economic importance, research on small streams and their response to forest harvesting is lacking for this forest type. Removal o f riparian vegetation can affect stream ecosystems through organic matter by bottom up processes in the food web. Biota among all trophic levels either directly or indirectly rely on energy input from detritus. Headwater streams quickly 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. process the largest size category of detritus, coarse particulate organic matter (CPOM), into the smaller particle categories by mechanical action and biotic functions (Winterboum and Townsend 1991). Fine particulate organic matter (FPOM) is defined as 0.45 um - 1 mm particle diameter, and dissolved organic matter (DOM) is the category smaller than 0.45 pm (Wotton 1990). Invertebrates and microorganisms directly consume FPOM that is a by-product o f fungi conditioning and invertebrate consumption of coarse particles (Suberkropp 1998). Consumers and predators indirectly take up FPOM while consuming primary producers and prey. The combination o f FPOM and DOM comprises the majority of organic matter transport in headwater streams, with DOM representing more than 70 % of annual energy flux in streams (Winterboum and Townsend 1991). DOM comprises the largest pool of organic carbon in freshwaters, and is an estimated 45-50 % organic carbon by mass (Allan 1995). We used dissolved organic carbon (DOC) as a measurement o f DOM and use the terms interchangeably, as is common in the literature. DOM is taken up by bacteria, zooplankton and protozoans, and is recycled by these biota through the microbial loop that is thought to form a large part of aquatic food webs (Allan 1995). Fish and macroinvertebrates indirectly consume DOM while preying on biota, and DOM may form a substantial part o f their diet. For example, DOM was shown to provide thirty percent o f the diet of Limnephilidae caddisfly larvae, a family whose members are generally detritivores (Wotton 1990). Components of DOM play an additional role in the biotic community by attenuating ultraviolet radiation (UVR) that can have harmful effects on fish and invertebrates (Bothwell et al. 1994; Brooks et al. 2005). 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DOM substances fall into two general categories: large, complex compounds including humic acids that are refractory and not easily digested, and compounds with low molecular weight such as carbohydrates and free amino acids that are easily assimilated by biota (Wotton 1990). Refractory compounds are generally derived from allochthonous sources o f DOM that enter streams from rainfall, run-off, groundwater, and detritus. Low molecular weight compounds tend to be autochthonous, produced by extracellular release from algae, and metabolic waste from stream biota (Allan 1995). In addition to influencing bioavailability of compounds, source o f DOM influences the reactivity of compounds when exposed to light, and ability of DOM to attenuate UVR. Allochthonous produced DOM compounds that are large and refractory react on exposure to light to produce low molecular weight, labile compounds, whereas autochthonous DOM tends to be non-photoreactive (Tranvik and Bertilsson 2001). High molecular weight compounds attenuate UVR to a greater degree than low molecular weight components of DOM (Thomas 1997). Thus, allochthonous produced DOM provides a lower quality food resource, tends to be more photoreactive, and provides greater attenuation of UVR than autochthonous produced DOM. DOM comprises the largest carbon pool in stream food webs, and concentration of dissolved organic carbon (DOC) is used as a measure of DOM (Thomas 1997). Absorbance is measured at various wavelengths in the UV spectrum (280-400 nm) to indicate composition of DOC, and UVR exposure through the water column (Moore 1989; Bothwell 2000; Weishaar et al. 2003; Cory et al. 2004). Spectral UV absorbance (SUVA) is defined as UV absorbance measured in m '1 divided by DOC concentration measured in mg L"1, and indicates amount of aromatic carbon (Weishaar et al. 2003). 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Aromatic carbons, defined as compounds that contain a benzene ring, are photo- and chemically-reactive, and are primarily associated with humic substances (Thomas 1997). Allochthonous DOC generally contains higher concentrations of UV absorbing compounds, and more aromatic carbons than autochthonous DOC (Tranvik and Bertilsson 2001). Therefore, SUVA can be used to estimate the relative content of allochthonous versus autochthonous DOC. Measuring SUVA at 350 nm indicates both source of DOC and UV exposure in the mid-range o f the UVR spectrum. In terms of organic matter, removal o f riparian vegetation may affect stream biota by altering transfer o f energy in the food web, or by altering the UVR attenuating effects o f DOC. In a stream survey of dissolved organic matter concentrations including 600 streams in British Columbia, ESSF watersheds had among the lowest values, ranging from 1 to 3 mg/L (Bothwell 2000). In northern subalpine forests, DOC levels reach minima during low flow (Burney 1990), and at this time stream biota are most susceptible to the effects of UV radiation (Kelly and Bothwell 2002). Removal of riparian vegetation may affect energy transfer by altering the source of DOC, or by allowing increased photolytic reactions through increased solar radiation. These two effects influence each other as photolytic reactions require DOC primarily made up of allochthonous produced substances. Our goal was to investigate responses of organic matter to clearcutting within a decade post harvest and to determine if seasonal differences occurred, and if seasonal shifts in organic matter transport responded to clearcutting. We selected the time frame for several reasons. First, the Province of British Columbia implemented the Forest Practices Code in 1995, and in order to compare streams managed under the same 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. umbrella of legislation and regulations, we selected streams harvested after 1995. Second, we selected a range o f years with varying reports o f response and recovery for hydrologic and organic matter transport in headwater streams. Third, we investigated a relatively narrow range o f years within each clearcut group to minimize variability within treatments. Therefore, we compared undisturbed “control” streams in mature ESSF forests to streams in “new clearcuts” harvested up to three years prior to sampling, and streams in “old clearcuts” harvested up to seven years prior to sampling. We posed three questions with respect to our comparisons among “control”, “old clearcuts” and “new clearcuts”. First, we evaluated responses of FPOM and DOC to clearcutting during high seasonal discharge. Second, we investigated the seasonal variation in DOC concentration and composition, and third we determined if clearcutting altered seasonal variation in DOC concentration and composition. Methods Study Location We conducted a comparative survey using paired sites located on each selected stream within the ESSF biogeoclimatic zone in the southern interior of British Columbia. Our requirements for selected watersheds within this forest type were that each stream be small, non-fish bearing and that “new” and “old” treatment streams be within clearcuts ranging from one to ten years post harvest. Similar streams within undisturbed forest served as control streams. We also chose to examine two subzones within the ESSF biogeoclimatic zone. O f five candidate study areas identified in conjunction with tenure holders in the region, field reconnaissance determined that two of these areas, Bone Creek and Damfino Creek, met the criteria. 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In 2004, we sampled twenty-two streams within the Bone Creek and Damfino Creek study areas (Fig. 2). Ten o f the streams were within the Bone Creek study area, located approximately 255 km northeast of Kamloops and situated in the ESSF wet cold subzone Thompson Plateau variant (ESSF wc2; Lloyd et al. 1990). The ESSF wc2 is characterized by 361 mm mean precipitation during the growing season, 782 cm mean annual snowfall, and 1.1 0 C mean annual temperature. Bone Creek is a tributary of the North Thompson River. A waterfall located 150 metres upstream from the mouth of Bone Creek prevents fish passage, leaving the remainder o f the watershed absent of fish (FRBC 2001). The remaining 12 streams were located in the Damfino Creek study area. Damfino Creek is located approximately 260 km southeast o f Kamloops within the ESSF dry cold subzone, Okanagan Highlands variant (ESSF del; Lloyd et al. 1990). The ESSF del is characterized by 261 mm precipitation during the growing season, 635 cm mean annual snowfall and 2.0 0 C mean annual temperature (Lloyd et al. 1990). Streams in this study area were within two watersheds, Damfino Creek and Two John Creek, which are separated by approximately ten kilometres. Damfino Creek flows into the Kettle River, and Two John Creek flows into the West Kettle River; both rivers are tributaries o f the Columbia River. A waterfall located 561 m from the mouth o f Damfino Creek prevents fish passage to the upstream reaches, and barriers restrict fish migration in the upper reaches o f Two John Creek where sample streams were located (FRBC 2000). All streams selected in this study fall under the S6 stream classification (Province o f BC 1995), which includes non-fish bearing streams no greater than 3 m bank-full width. Cutblocks selected in this study ranged in size from 20 to 95 ha and contained up 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to 24 identified S6 streams within one block. Best management practices prescribed retention o f the most windfirm 5 % of codominant conifers within 20 m of the channels of two streams for every 40 ha o f harvest area, and full retention o f non-merchantable vegetation within 5 m (Province o f BC 1995). All treatment streams were nearly to completely void of conifers in their riparian zones, and where conifers remained they were early serai and sparse. A portion of streams in each treatment had nonmerchantable vegetation left in the riparian zone that comprised deciduous species such as alder (Alnus tenufolia) and rhododendron (Rhododendron albiflorum). The primary logging method used within Damfino study area streams was conventional grapple skidding, while high lead yarding was common in Bone Creek study area due to steeper topography. Although this study was not designed to quantify or evaluate the effects of slash, organic matter left in or adjacent to streams is considered an important influence on stream responses to clearcut logging. Our observations were that slash presence was variable among clearcut streams ranging from very little slash present immediately adjacent to the channel, to high amounts of slash that covered the stream channel. Figure 3 shows examples of slash presence representative o f the intermediate amount o f slash, not completely covering the stream but present in the riparian zone and the channel. The specific influence of slash was not tested in this study as the objective was to evaluate effects o f the end result from clearcut harvesting, regardless o f site specific differences in logging operations. 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Southeastern British Columbia K W E ^T } .—4. f lake -' \ ' wells \ ' f Um'Vt, ’ Voi# k Index Map of BC .*1-’‘J•A6'**.J . |SJ . i * f- ■o ■ «® ,/ ' -Z >.. ; ^ :j | ■■■"' '%■•.?.c■" ./o ■-% S*£ '■ ;;r o ■ QLACI&1* '** . ^ ^ ‘i "tH ^^StOK E _\ ''|T ^ •' ■;v s -■ -\MArfwnt* w •■ :. \ t # ^ , IV V -" \ r «-Orl \ J U a rftliE ff/ - ^_ * ' :T\OA M ■/ \V j .^0^^ f£pf "V*' K j Uffl# «■ Vz , *4 ~ ..'Li- ■' v/ j i"KXMioofe,s. ^H o eV c^ 'T ' ;i m Legend 1 L”“ * »] 1 \ ^ ^ - ' y ' v . }.i;t e i ' A I W W O W ,1 -— f |V vewJpN^’'S'* ii* ■. bymtJy ,....-. .Lj|ff Coi^NM'' L~ \ ;%f ■'Oyanl^-^fe<» 1JM * KRUWf: 1 ^ Edg^ 1 ■ ?I J 1!; : < .■ » 64 km. — • tOWrtl ' Provincial B oundary City or Com m unity JV C B one C reek S tudy A rea :L* ^ \K;*■•'\okM^O\. \ "H w*-' \ \ .f' □ Damfino C reek S tu d y A rea x'.^;..pgNT^ro^' ■ 120VV Major Highway :W AfO*I fl V^AOR-ANO f fiHWFTON W ater , 0UVEW - .■ / /y us w naw Map cent«r: 50*49* N, 116*59* W Ministry of Sustainable Resource Management Figure 2. Location o f study areas within British Columbia. 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3. Two examples o f slash in proximity to streams in this study compared to two control streams; all photographs were taken in 2004. Top left photograph is stream 215 logged in 2003, and top right photograph is stream 206 logged in 1998; both streams are within Damfino Creek study area. Bottom left photograph is control stream 013 at Bone Creek study area, and bottom right photograph is control stream 203 at Damfino Creek study area. 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Experimental Design Selected streams fell into one of three treatment categories: “control” streams within undisturbed mature forest, “new” streams within clearcuts 1-3 years post harvest, and “old” streams within clearcuts 5-7 years post harvest. The minimum distance that a stream flowed through its associated cutblock was 300 metres at Bone Creek and 250 metres at Damfino Creek. Therefore, this minimum distance separated the downstream site from the upstream forest edge, and an additional 50 metres separated the forest edge from the upstream site (Fig. 4). The same distances separated control stream paired sites located at elevations and physical characteristics representative of treatment streams. Streams served as replicate sampling units to avoid issues of pseudoreplication (Hurlbert 1984). We considered differences between paired sites of control streams as the reference level, and compared the reference level to differences between paired sites of clearcut treatments. This is similar to a before-after control-impact design, and is effective to detect responses to disturbance (Underwood 1994). We addressed potentially confounding variables by judicious site selection and timing of sampling. We maintained a consistent slope distance between paired sites in order to reduce variability associated with effects that may occur as a function o f reach length. We excluded streams if they had insufficient flow, if they flowed through a known disturbance upstream from the selected clearcut, and if they contained a riparian buffer of timber. To avoid the additional influence of roads, we excluded streams crossed by active or decommissioned roads larger than skid trails within or above the paired sites. We selected reference streams with a range o f physical characteristics representing those of treatment streams. 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Experimental Design Paired sites: Reference stream, top Treatment stream, bottom Legend Upstream site Downstream site Cutblock *V Stream Forest vegetation Figure 4. Schematic diagram showing experimental design. Timing o f sampling influences organic matter export rates due to seasonal and diel periodicity, and discharge pulses. Sampling dates corresponded with high and low periods according to Water Survey of Canada archived data for the study areas (Fig. 5). We commenced sampling at the same time each day, with downstream sites in the morning followed by upstream sites in the afternoon. We did not sample within 24 hours of heavy rainfall. Treatments were sampled alternately within each sampling period, as opposed to sampling all control streams then all clearcut streams, to avoid bias from seasonal fluctuation through our sampling timeframe. From June 18 - July 23, 2004 we sampled paired sites on 12 streams at Damfino Creek, and from August 12 - 29 we sampled paired sites on 10 streams at Bone Creek. 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. We sampled an additional four unpaired sites consisting o f two “old” and two “control” treatments at Bone Creek because paired sites were not available that met our criteria. During the low flow period, we did not sample any upstream sites in the Bone Creek study area due to time constraints, and data collection was not possible at stream 002 due to access restriction from a landslide. At Damfino Creek, we did not conduct low flow sampling at the upstream site o f stream 210 that was entirely dry. Each stream was assigned an identification number in chronological order of sampling with the leading digit representing the study area. All physical attributes were measured during the high flow period (Table 1). Water temperature data were successfully collected at 18 sites for periods ranging from one to seven days surrounding the sample date o f each stream. Water temperature data were collected for reference of physical stream characteristics and were not entered into analysis, whereas the remaining physical data were used in analysis. 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40 mean /-------- -- mm —•max 30 o 0) 20 0) o> (0 .c o V) 10 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 120 — mean - - min —■max T" O 0) <0 E 0) o> k_ <0 s: o v> 5 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Time (Month) Figure 5. Hydrographs for (a) Bone Creek and (b) West Kettle River showing mean daily discharges. (Water Survey o f Canada, 2006). 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Nov Dec Stream ID a Bone Creek 005 012 013 025 021 023 002 027 010 011 Canopy closure Treat­ ment Year logged Control Control Control Control N ew N ew New New Old Old 2003 2003 2002 2001 1999 1998 25 78 94 52 0 0 42 Damfino Creek Control 203 208 Control 209 Control Control 213 New 201 New 205 N ew 207 215 New 206 Old 210 Old 211 Old 202 Old 2003 2003 2003 2003 1998 1998 1998 1997 77 5 73 84 (% ) b 59 0 0 63 17 55 57 92 18 39 45 Discharge Bank-full (L sec'1) width (m) Stream gradient Aspect (%) Min (m) CO Max CC) 9.8 8.9 5.7 7.9 9.9 6.1 n.d. 14.8 12.3 7.6 9.9 11.8 12.3 n.d. 9.5 9.2 9.2 11.6 11.8 4.6 8.4 6.5 n.d. 12.0 8.6 16.0 n.d. 5.5 7.3 6.1 n.d. 12.0 11.2 19.9 n.d. 6.4 7.4 6.2 n.d. 16.7 10.0 8.5 n.d. 5.46 12.38 2.39 4.28 1.31 0.59 0.63 3.42 0.43 1.66 0.47 2.72 2.13 0.61 0.85 0.68 0.94 1.03 2.50 1.00 35 34 32 32 41 31 57 65 58 33 W E SE NE S S E SW N NE 1184 1131 1297 1671 1536 1845 1453 1543 1262 1171 4.16 4.29 4.11 1.17 19.04 3.77 5.63 1.47 1.41 0.18 0.92 2.69 1.63 0.71 1.56 1.04 0.86 1.02 0.76 1.62 1.01 1.30 0.76 0.86 4 18 23 12 7 14 14 16 9 24 27 8 SW 1566 1592 1820 1682 1547 1660 1730 1652 1873 1825 1830 1695 s NW NE SE N SE NE E N W w Water Temp.d Elevation m edianc 0° bo Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1. Physical characteristics and sampling dates o f study streams sampled in 2004 within the Bone Creek and Damfino Creek study areas. Date sampled High Low Flow Flow 14-Aug 24-Aug 19-Aug 27-Aug 13-Aug 13-Aug 25-Aug 29-Aug 12-Aug 26-Aug 4-Oct 4-Oct 4-Oct 26-Jun 9-Jul 11-Jul 21-Jul 18-Jun 29-Jun 10-Jul 23-Jul 30-Jun 13-Jul 14-Jul 25-Jun 31-Aug 31-Aug 31-Aug 31-Aug 30-Aug 31-Aug 31-Aug 31-Aug 31-Aug 31-Aug 31-Aug 31-Aug 4-Oct 4-Oct 4-Oct 5-Oct 4-Oct 4-Oct n.d. Note: Attributes measured during high flow sampling period with multiple measurements at each site. Stream average values are presented with the exception o f canopy closure, which are given for downstream sites for comparison o f logged versus old growth, (n.d. = no data). ■ All streams are unnamed; therefore, each stream was assigned a number with the leading digit indicating study area B Canopy closure represents percentage o f cover measured by densiometer 1 m above ground at downstream sites c Represents the median elevation between downstream and upstream paired sites d Water temperaure logged hourly over at least 24 hours during high flow sampling date o f each stream 20 Data Collection We recorded physical characteristics at each site during high flow sampling including canopy closure, discharge, bank-full width, stream gradient, aspect, and elevation. At two locations for each site we used a densitometer held 1 metre above ground, facing four directions to determine canopy closure (Lemmon 1956). We measured discharge by recording the time to fill a large container o f known volume at least six times throughout sampling and obtaining the average (Gore 1996). We measured bank-full width at three locations, and gradient and aspect at each site using a clinometer and compass. We obtained elevation and UTM coordinates with GPS at each site location, and verified elevation using topographical maps. Temperature loggers were placed in streams at each site surrounding the high flow sampling period and recorded hourly water temperature over at least 24 hours. We collected fine particulate organic matter (FPOM) samples during the high discharge period only, and dissolved organic matter (DOM) samples during two seasons. At each downstream and upstream site we triple rinsed three 11 L carboys using stream water, and filled each carboy using a sieve to remove particles greater than 1 mm in diameter. Using vacuum filter apparatus with hand pumps, we filtered the water using pre-ashed, pre-weighed glass fibre filters of nominal pore size 0.7 pm (Whatman GFF) in triplicate. From each carboy, we filtered four litres o f water or an amount sufficient to leave a visible layer, and retained each filter in a pre-ashed tinfoil envelope within a sealed plastic bag. At the end of each day, we froze the samples and kept them frozen until analysis. In the laboratory, we processed the filters and used gravimetric analysis as 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. described in Wallace and Grubaugh (1996) to determine ash-free dry mass (mg L '1) of FPOM. We retained the remaining filtrate of each sample in 60 mL amber glass bottles and kept samples refrigerated for subsequent DOC analysis. We used a total carbon analyzer (Model V SCH, Shimadzu Corporation; Kyoto, Japan), using the non-purgible organic carbon method, to obtain DOC concentrations in parts per million or mg L’1. We measured the absorbance of each sample at 350 nm (25° C) using a Cary 50 spectrophotometer (Varian Incorporated; Palo Alto, California, USA). The spectral ultraviolet absorbance (SUVA 3 5 0 ) o f each sample was calculated using absorbance divided by DOC concentration (Weishaar et al. 2003). Data Analysis We used discriminant analysis to test for natural patterns among streams using mean discharge, bank-full width, stream gradient, aspect and elevation of each stream as independent variables. If physical characteristics did not predict group membership among treatments, then we assumed that responses of organic matter were due to logging or season rather than natural variability among subject streams. The physical stream variables were log transformed in order to achieve equal population covariances, to reduce the influence of outliers, and to meet the assumption o f multivariate normality required for discriminant analysis (Tabachnick and Fidell 1989). For each response variable, FPOM concentration, DOC concentration, and SUVA 3 5 0 , we used site means to determine effect as the difference between paired sites relative to the upstream reference site (Equation 1). A negative effect indicated the variable decreased after flowing through the clearcut or between paired sites, whereas a 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. positive effect indicated the variable increased. Control stream effects served as the base condition to compare against treatment stream effects. ds-us E ffe c t = us x 100 Equation 1 Where ds = downstream site average, us = upstream site average. All data were checked for normality and homoscedasticity. FPOM effect required a Logio(x+1) transformation to meet the assumption of normality; the remaining variables did not require transformations. Data are presented untransformed for ease of interpretation. To assess the appropriateness o f pooling data from the two study areas, we used a two-way ANOVA with treatment and watershed as fixed factors. If there was not a significant interaction between watershed and treatment (p>0.05), then analysis proceeded using pooled data and single factor ANOVA was used to detect differences among treatments for the remaining variables with a priori contrasts. The contrasts were chosen to detect differences between the following groups: i) control and old clearcuts, ii) control and new clearcut streams, and iii) old and new clearcut streams. We first assessed seasonal differences in DOC concentration and SUVA350 using a paired t-test. This approach evaluated seasonal differences by comparing two measurements taken at identical sites. We used all upstream and downstream sites that had measurements for both seasons (n=36). Additional sample size was possible for this analysis because the unpaired downstream sites were included in the paired t-test but not in downstream-upstream effect (Equation 1) analysis. The effect between sampling periods gave differences between paired site measurements repeated during high and low flow, relative to high flow values (Equation 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2). In this second approach, we used a subset of the paired t-test data due to Bone Creek sites lacking upstream low flow data. The purpose o f this approach was to detect if seasonal response depended on treatment for DOC concentration and SUVA 3 5 0 . We used a two-way ANOVA with study area and treatment as fixed factors. If study area did not contribute to the general linear model, we then used a one way ANOVA with treatment as the factor and a priori contrasts as specified above to determine if season effect depended on treatment. LF-H F xlOO E ffe c t HF Equation 2 Where HF = high flow site average, LF = low flow site average. Results Stream Physical Characteristics Discriminant analysis found that discharge, bank-full width, stream gradient, aspect and elevation did not predict group membership to the three treatment types (Wilks’ f = 0.45, p=0.19). This supported our assumption that effects found to differ among treatment type are due to logging or season rather than natural variability, with respect to the physical attributes measured (Fig. 6a). Average canopy closure at downstream sites o f control streams was 61 %, and was reduced to 32 % and 37 % on average at old and new clearcut sites respectively (Fig. 6b). A priori contrasts did not detect any differences between treatment groups (p>0.05). ANOVA also did not detect a difference in canopy closure among treatment streams (F 2 ,i9 = l .9, p=0.18). Canopy closure values varied highly within treatments and overlapped between control and clearcut streams. This was due to numerous clearcut streams being highly shaded with dense early successional vegetation, and several control streams having an understory 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. comprising herbaceous species that were less dense than that of early successional vegetation. Pooling Study Areas Table 2 summarizes DOC concentration, SUVA 350 , and FPOM concentrations by treatment and study area for upstream and downstream sites. The two way ANOVA analyses showed that study area (Bone Creek or Damfino Creek) did not significantly contribute to the model used to discern treatment responses in FPOM, DOC and SUVA 350 effects (p=0.16, 0.10, 0.85), nor were there significant interactions between treatment and study area for the three variables (p=0.15, 0.22, 0.50). Therefore, we pooled data from both study areas and compared mean effect among treatments including all 2 2 streams. Treatment Response During the high flow period, FPOM differed among treatments (F2 ,i9 = 3 .6 , p=0.048). FPOM responded positively to new clearcuts compared to control streams (p=0.032) and old clearcut streams (p=0.036, Fig.7). Contrasts and ANOVA did not detect differences in DOC (F2 ,i9 = 1 .7 , p=0.20) or SUVA 350 (F2,i8=0.69, p=0.52) among treatments (Fig. 7). Data were missing for one control sample for SUVA 350 measurements. Effects for streams 002 and 023, both within new clearcuts, were extreme at 381 % and 196 % respectively. Excluding these outliers changed the mean o f the new treatment from 75.7 % (SEM = 49.8 %) to 4.75 % (6.21 %) and the result of the ANOVA test to F2 ,17= 0 .008, p=0.992. One of these streams, 023, was also an outlier for FPOM effect; however, exclusion of this stream did not alter the significances for FPOM. During the low flow period, DOC did not differ among treatments (Fig. 8, ANOVA: F2,io=0.52, p=0.61; contrasts: p>0.05). SUVA350 differed among treatments 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (F2 ,io= 5 .9 , p=0.027), with increased values between new clearcut sites (p=0.009) and pooled clearcut sites (p=0.022) compared with controls (Fig. 8). Seasonal Response There were seasonal differences for DOC concentration (t3g= 4.2, p<0.001) and SUVA 350 (t36=-2.6, p=0.012). In Figure 9 it is clear that DOC declined in each treatment within both study areas in the low flow compared to high flow season. SUVA 350 seasonal differences followed similar trends in each study area, with seasonal decline in old clearcut streams only (Fig. 9). Mean DOC concentration decreased by 19 % from high to low flow, and mean SUVA3so increased by 19 % at low flow (Fig. 10). Treatment x Seasonal Response Season effect did not vary among treatments for DOC (Fig. 11, ANOVA: F 2,2o= 0.98, p=0.40; contrasts: p>0.05). Although ANOVA did not detect differences in SUVA35 o among treatments (F2,2o=3.4, p=0.057), a priori contrasts showed that SUVA35 o seasonal effect in old clearcuts differed from both control streams (p=0.029) and new clearcut streams (Fig. 11, p=0.041). CM •2 1 oc O ^ % x° * ♦ A 100 ♦ Control x O ld o N ew (a) c 0 ♦ c c E u (0 25 5 — 1— -3 -1 1 3 Discriminant Function 1 Control Old New Figure 6 . (a) Two discriminant functions describe stream abiotic characteristics and do not predict group membership to the three treatments (Wilks’ A.=0.45, p=0.19). (b) Mean crown closure at downstream sites does not differ among treatments (p=0.12); error bars represent SEM. Control n= 8 , old clearcuts n= 6 , new clearcuts n= 8 . 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2. Mean DOC concentration (mg/L), specific U V absorbance (SU VA 350, L m ' 1 mg'1) and FPOM concentration (mg/L) by treatment and study area summarized for downstream and upstream sites among treatments. Standard errors o f the mean are given in parentheses._______ Damfino Creek Bone Creek C ontrol Old N ew Control Old N ew 3 .0 9 (0.77) 2 .1 2 (0.72) 0 .6 9 (0.09) 0 .9 9 (0.05) 1.24 (0.58) 4 .9 2 (0.72)a 4 .7 7 (0 .1 1)a 0.90 (0.01 )a 0 .8 5 (0.02)a 1.75 (0.75)a 5 .0 9 (1 .1 1 ) 3 .5 3 (0.68)b 1.08 (0.11) 1.42 (0.11 )b 5 .2 8 (0.33) 5 .0 9 (0.44) 0.87 (0.08) 0.91 (0.05) 0 .3 3 (0.11) 5.30 (0.50) 5.11 (0.41) 1.01 (0.68) 0.92 (0.12) 0.32 (0.06) 5.53 (0.51) 4 .4 9 (0.36) 0.88 (0.11) 1.03 (0.96) 0.41 (0.12) 2 .8 9 (0.46) 3.81 (0.24)a - - 0.81 (0.13) 1.08 (0.07)a 3 .7 4 (1 .6 0 ) 0 .9 6 (0.18) 5 .1 0 (0 .3 4 ) 4 .5 8 (0.32) 0 .9 3 (0.10) 1 .1 4 (0 .1 6 ) 0 .6 9 (0.36) 5.96 (0.40) 5.24 (0.66)b 0 .9 4 (0 .1 4 ) 0 .9 4 (0.14)b 0 .3 6 (0.10) 5.30 (0.53) 4 .3 7 (0.13) 0.84 (0.10) 0.87 (0.07) 0.31 Downstream Sites DOC: High flow DOC: Low flow SUVA: High flow SUVA: Low flow FPOM : High flow 2 .7 0 (0.69) DOC: High flow DOC: Low flow SUVA: High flow SUVA: Low flow FPOM: High flow - - - 0 .9 0 (0.11) 3.50 (1.51 )a 0 .7 4 (0 .2 1 ) O O Upstream Sites Note: T h ere a re four s tre a m s in e a c h cate g o ry e x c e p t w h e re n o ted b y a n=2; b n=3. 350 (a) FPOM C oncentration - 300 C?250 -| 0s * ~200 - a 150 2]100 50 - 0 -50 125 - _L Control 1 _i_ Old New 125 (b) DOC C oncentration 100 - 100 ^75o' ^75 '-■'50 H rso 0 25 UJ 0 A o ,£ 2 5 - T Hi 0 A A /J __________ I ____ -25 -25 -50 - (c) Specific A bsorbance Control Old -50 J New Control Old New Figure 7. Columns show mean effect for (a) FPOM, (b) DOC and (c) specific absorbance (SUVA 3 5 0 ) among treatments during high flow; error bars denote SEM. Positive effect indicates an increased value, whereas negative effect indicates a decrease from upstream to downstream sites. Treatment means with common letter do not differ significantly (p>0.05). (Control n= 8 , Old n= 6 , New n= 8 ). 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50 -| 50 (a) DOC concentration (b) Specific ab so rb an ce _25 - .25 o 0 St UJ-25 A 50 -1 Control X XL. Old New JL B 13 0 £ m-25 A 50 -* Control New Old Figure 8 . Mean effect for (a) DOC concentration and (b) specific absorbance (SUVA 3 5 0 ) among treatments during low flow season; error bars denote SEM. Positive effect indicates an increased value, whereas negative effect indicates a decrease from upstream to downstream sites. Treatment means with common letters do not differ significantly (p>0.05). 1.6 J at ", (a) Bone Creek 1.6 -| 0 .8 - J E 0.0 o> E 0 .0 T I -0 .8 - 1 - 0 . 8 -I r 6 - 2 .4 - O 0 - 3 .2 - ^ 0 .6 Control Old 6 -2 .4 - O 0 -3 .2 -4 .0 New E 0 .4 0.1 ^-0.3 u. I 0.2 -| ^ 0.1 jC-0.1 £ - 0.1 - 5 0-2 3 -0 2 - -0 .3 New O) L * 0.2 i Old E ' o .5 O) ~ 0 .3 ll Control _ 0 .6 -, (d) Damfino Creek (c) Bone Creek £ 0 .5 E 0 .4 T U.-1.6 it-1 .6 - -4 .0 - (b) Damfino Creek 0 .8 - Control Old New - I 1. ■ 0.3 j Control T Old New Figure 9. Mean seasonal differences at downstream (impacted) sites for DOC concentration at (a) Bone Creek, (b) Damfino Creek; and for specific absorbance, SUVA350 at (c) Bone Creek and (d) Damfino Creek. Error bars denote SEM. DOC decreased during low flow in every treatment indicated by negative values, whereas SU VA 350 tended to increase from high to low flow. Bone Creek: Control n=4, Old n=2, New n=3. Damfino Creek: Control n=4, Old n=4, New n=4. 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.0 5.0 o> 1.0 Ll 4.0 2* 3.0 O 2.0 w O.o 0.0 High Flow Low Flow High Flow Low Flow Figure 10. (a) Mean DOC concentration and (b) mean specific absorbance (SUVA 3 5 0 ) during high flow and low flow sampling. Error bars represent SEM. Treatment means with a common letter do not differ significantly (p>0.05). (DOC n=39; SUVA 3 5 0 n=37). 1 50 n (a) DOC concentration _25 ^25 55 o t3 0 (b) Specific a b so rb an ce _JL B 0 st St ^ -2 5 m-25 ~ r~ -50 J Control Old -50 New Control Old New Figure 11. Mean seasonal effect among treatments for (a) DOC concentration and (b) specific absorbance (SUVA350); error bars denote SEM. DOC concentration decreased from high to low flow indicated by negative effect values. SUVA 3 5 0 increased in control and new, and decreased in the old clearcut treatment. Control n= 8 , Old n= 6 , New n=7. Discussion Treatment Response FPOM concentration increased in streams flowing through new clearcuts compared with control and old clearcut streams. This is in agreement with results from previous work. In North Carolina, disturbed streams contained higher particulate organic matter (POM) initially following clearcut logging, and recovered within four years (Webster et al. 1983). Macdonald et al. (2003a) found increased levels o f suspended 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sediments following harvesting activities in the sub-boreal spruce zone in central BC that recovered to pre-harvest conditions within three years following harvest. Our results were also consistent with data collected a year previous to this study in Bone Creek (Greenhalgh and Heise 2004). Our findings indicate that FPOM in the water column increases 1-3 years post harvest in response to removal of riparian vegetation, but returns to reference levels in the period 5-7 years following harvest. DOC concentrations we observed fall in the lower range of comparable headwater streams reported in the literature. Reported values include a peak flow average of 7.0 mg/L (Colorado, Hood et al. 2005), a range o f 5.3-17.6 mg/L (Colorado/Montana, Brooks et al. 1999) and 3.2-7.2 mg/L (Quebec, Eckhardt and Moore 1990). In the current study, DOC high flow treatment means fell between 2.89-5.96 mg/L, and low flow means were 2.12-5.24 mg/L. Biota within streams containing DOC in this range are most vulnerable to UV exposure through either increased UVR or decreased DOM (Kelly and Clare 2001; Kelly et al. 2003). Subalpine aquatic food webs may be particularly susceptible to global or regional increased UVR because light penetration increases with elevation due to atmospheric thinning, reduced topographic shading and longer daylength (Williamson et al. 1996). The results of this study support a prior survey (Bothwell 2000) that small streams in the ESSF contain DOC at threshold concentrations in terms of detriment to biota if UVR increased, or if DOM input decreased. Although differences were not significant during high flow, DOC concentration tended to increase within streams flowing through new clearcuts (Fig. 7b). SUVA 350 showed a similar non-significant trend during high flow. Although low flow DOC concentration effects did not differ among treatments, SUVA 350 effect differed indicating 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. elevated allochthonous input in response to new clearcuts. Our findings that DOC concentration did not respond, or only slightly responded to disturbance are comparable to other findings from headwater streams. Kreutzweiser et al. (2004) found no change in DOC concentrations, and increased DOC export that recovered two years post harvest in selective harvesting after up to 89 % removal of basal area. Moore and Jackson (1989) found 1-4 year old cutblocks had little effect on DOC export. Our SUVA 3 50 results concur with the response reported by Hood et al. (2006) that as DOC concentrations increased during high run-off, SUVA 254 increased 35 % in clearcut watersheds, and only 10 % in reference watersheds. An investigation of two headwater streams in Arkansas found higher refractory pools o f DOC in an agriculturally impacted stream compared to an undisturbed stream (Ziegler and Brisco 2004). Burney (1990) proposed two pathways that cause increased DOC concentrations during rising water; flush o f the accumulation of microbes and organic matter under snow pack, and run-off from mineral soil-derived refractory compounds. The first pathway dominates snow-fed streams and the amount of labile DOC derived is positively related to amount of heterotrophic activity during winter (Brooks et al. 1999). We suggest that elevated organic matter adjacent to new clearcuts influenced the DOC flush during run-off by supplying increased allochthonous compounds compared with control streams. Seasonal Response The allochthonous influence on DOC during high flow in new clearcuts was also found during low flow, but this terrestrial signal was greater during low flow. DOC concentrations decreased between high and low flow while the portion o f refractory compounds (SUVA 350 ) increased. Seasonal reductions in DOC concentration agree with 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. many previous studies in headwater streams and are attributed to highest discharge periods providing maximum inputs of DOC (Tate and Meyer 1983; Moore 1989; Brooks et al. 1999; Kiffney et al. 2000; Hood et al. 2003). During high discharge periods, DOC is derived primarily from terrestrial sources leading to high refractory composition unless mediated by heterotrophic activity under snow pack. Previous work on seasonal shifts in DOM composition found differing results depending on dominating sources of DOM. Hood et al. (2005) attributed a seasonal increase in SUVA values to a spring flush of labile DOC from accumulated microbial activity under snow pack in Colorado forested subalpine streams. In Arkansas, Brisco and Ziegler (2004) found the opposite seasonal trend in an undisturbed headwater stream that they attributed to increased algal growth during summer and photochemical breakdown of allochthonous DOC. In the current study, we suggest that a high degree o f labile DOC input flushed during spring lead to a relative reduction in labile DOC during low flow sampling after depletion of the source. Autochthonous production was limited in our control streams due to high shade and so the importance of this pathway and photochemical reactions were limited. Old and new clearcut streams appeared to function differently from each other and from control streams throughout the season. New clearcuts tended to show a positive response in SUVA 350 during both sampling periods. Old clearcut streams were the only treatment where DOC seasonally shifted to a more labile pool indicated by decreased SUVA 350 effect between high and low flow. Next, we will discuss reasons for this differing response among treatments. 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Treatment x Seasonal Response Increased light and nutrient availability are two known responses to clearcut logging in the initial years following harvest. We suggest these factors acted together resulting in an opposite seasonal shift in old clearcuts, and alternate functioning within new clearcuts. Humic derived DOC exposed to light generally enhances the labile pool of DOC through photolysis (Tranvik and Bertilsson 2001; Brisco and Ziegler 2004). Increased heterotrophic activity follows photolysis by providing bioavailable DOC unless a further limiting factor exists. An initial nutrient flush followed by nutrient limitation is associated with removal of riparian vegetation (Gregory et al. 1987). An investigation of ESSF wc forest soils in close proximity to our study locations revealed that forest soil nitrogen concentrations more than doubled in the first year following clearcut logging, and then declined 4-5 years post harvest to less than reference conditions (Feller 1997; Hannam and Prescott 2003). In old clearcut streams, exposure to light during the summer may have broken down photoreactive compounds and lead to increased labile compounds. However, in combination with decreased nutrient availability, the labile pool of DOC likely accumulated throughout the summer due to reduced consumption by heterotrophic microbes compared with control streams. Increased nutrient availability within new clearcuts provided a different scenario that lead to a decreased portion of refractory DOC. Ziegler and Brisco (2004) showed that elevated nutrients in a disturbed headwater stream supported more microbes and resulted in a larger pool of refractory DOC than a comparable forested stream. The authors reasoned that two processes lead to this result: microbial degradation o f terrestrial organic matter released humic substances (generally aromatic carbons), and elevated microbial uptake o f labile DOC increased the portion o f refractory compounds. In the 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. present study, the refractory portion o f DOC was elevated in new clearcuts in both seasons sampled, but particularly during low flow. DOC concentration tended to decrease seasonally more in new clearcuts than in control and old clearcut streams (Fig. 1la). Seasonal effects in new clearcuts differed from old clearcuts but not from control streams (Fig. 1 lb). Thus, we suggest that increased nutrients in new clearcuts, likely acting synergistically with photolytic production of labile DOC compounds, supported increased microbial activity throughout the summer. Although we did not examine microbial biomass, we suggest that microbial uptake of the labile pool would lead to the seasonal increase in the refractory pool o f DOC compounds in new clearcuts. Heavily shaded control streams lacked the further addition of labile DOC through photolysis, but also lacked the additional consumption o f labile DOC through elevated microbial biomass. Thus, the net effect of new clearcuts on aromaticity of DOC appeared similar to control streams. Although canopy closure was similar between clearcut groups, and vegetation had not yet established significantly in either group, our results showed different temporal responses to forest harvesting. Generally, three main silviculture activities were completed following winter logging in our locations. In the initial year following logging we found that slash remained on most sites. Debris was generally burned and site preparation such as mounding occurred to expose mineral soil for tree planting the next summer. These silviculture activities take place within the time frame o f our new clearcut treatment, and can take place immediately adjacent to S5 and S6 streams, which do not require a riparian reserve or machine free zone. Organic debris generally causes negative effects such as increased suspended sediment, although it is noted to mediate 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. temperature increases associated with logging (Jackson et al. 2001). We suggest that in new clearcuts organic debris caused elevated levels of FPOM and DOC and supplied photoreactive DOC compounds that in turn produced labile DOC when exposed to light. Organic debris would also supply increased nutrient availability to allow elevated biomass o f stream heterotrophic biota that in turn depleted the supply o f labile DOC (Fig. 12). This response may trigger a bottom up response in the stream food web if consumers o f the heterotrophs are limited either by nutrient availability or by their prey. R efractory DOC C om pounds Labile DOC O rganic D ebris M acro in v erteb rates In c rea se d N utrients 1 1/ In cre a se d H eterotrophs Figure 12. Schematic diagram showing the flow o f energy from allochthonous sources and light to heterotrophs and macroinvertebrates. An alternative or additional explanation to the changes in DOC is a photochemical pathway reported for autochthonous compounds. Algal derived compounds can condense on exposure to light thereby reducing the labile pool o f DOC (Tranvik and Bertlisson 2001; Brisco and Ziegler 2004). This would also explain seasonal increased SUVA 350 in new clearcut streams if a smaller but important portion of DOC was derived from algae. We did not measure the influence o f stream temperature, a factor shown to positively affect microbial production in headwater streams (Peters et al. 1987). In order to confirm or refute these pathways, a more complete dataset would be required such as measurements o f nutrient levels, UVR penetration, and microbial 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. activity in clearcut streams. Analytical methods such as stable isotope analysis and fluorescence index analysis would assist in a further examination o f the relative influences of algal, deciduous or coniferous vegetation on DOC. Experimental Approach There were several assumptions made in the design of this experiment. Hydrologic data for the headwater streams in this study do not exist and our discharge measurements were point data collected during high flow sampling only. Discharge is an important parameter that influences total export o f organic matter in response to clearcut harvesting. In North Carolina, increased particulate organic matter (POM) standing crop occurred in small streams 16 and 30 years after clearcut harvesting (Silsbee and Larson 1983; Stone and Wallace 1998). Several authors have suggested that recovery from disturbance in northern streams may take longer than for temperate streams at lower latitude due to hydrology (Moore and Wondzell 2005), and differences in temperature regime (Macdonald et al. 2003b). In similar forest types to our study, increased discharge persists for at least 15 years in the montane spruce in the southern interior of BC (Winkler et al. 2005) and 30 years in the ESSF in Colorado (Troendle and King 1985). This response may cause increased annual export of organic matter even if concentrations do not differ among treatments. However, even intensive sampling of organic matter concentrations combined with hydrologic data underestimates export due to under representation of storms where most of POM transport occurs (Cuffney and Wallace 1988; Webster et al. 1990). Continuous hydrologic data in our treatments streams would have enabled the determination of minimum annual export rates. 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. We assumed a high degree o f natural variability existed among our streams, making it difficult to use the reference stream approach. To address this issue we used mean differences between paired sites on control streams as a reference for the base condition. These differences were not always equal to zero, most notably there was a slight increase in DOC concentration within control streams from upstream to downstream during low flow, accompanied by a shift to more labile carbon composition of DOC. During high flow smaller differences in the same directions (+/-) occurred within control streams. It is expected that DOC concentrations increase longitudinally from the headwaters downstream, with cumulative inputs from detritus, and through contact with the upper soil horizon. Another factor during high flow sampling is diel fluctuations in DOC. We sampled downstream sites between 8 am and noon, and upstream sites between noon and 4 pm, to maintain consistency across treatments. These are two considerations that may account for differences in organic matter between paired sites on reference streams. Sampling bias was constant among treatments because we compared differences between two sites per stream, with similar physical attributes among all streams. Summary We found that FPOM transport increased initially in response to clearcutting, but recovered with 5 years. DOM transport did not change in response to clearcutting, but seasonally dependent composition of DOM was altered in clearcuts, and did not recover to pre harvest conditions within 7 years post harvest. Our results indicate that pathways of DOC utilization were altered in clearcuts. This was only evident during the low flow sampling period. The responses seen in this study were not necessarily detrimental to 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. stream biota, but could lead to a shift in community composition among trophic levels. The research has applicability to the ESSF forest, which has lacked investigation in the interior of the Pacific Northwest Region to date. It is a first step to quantify levels of organic matter transported in small streams, and to identify responses to forestry disturbance up to eight years post harvest. Future research should investigate whether responses occur in the longer term. Further investigation is warranted to determine if the responses we found translate to changes in annual export o f organic matter and if stream biota respond to the alterations we reported. 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 2: Effects of clearcutting on invertebrate drift within headwater streams in British Columbia, during two periods post harvest. Abstract Invertebrate drift exported from headwater streams subsidizes downstream food webs and alterations to invertebrate drift can affect fish production. We investigated effects of clearcut logging on density and composition of drift in headwater streams within Englemann spruce - subalpine fir (ESSF) forests in British Columbia in three treatment groups: “new” 1-3 years post harvest, “old” 5-7 years post harvest, and reference streams (non harvested). We compared mean differences between paired sites on streams for 52 biotic metrics. We used logistic regression and multiple linear regressions for common taxa to determine if disturbance predicted their presence/absence, and if fine particulate organic matter (FPOM) and dissolved organic carbon (DOC) predicted their densities. Seventeen o f the 52 biotic metrics differed among treatments. Total invertebrate densities (abundance and biomass) increased due to clearcutting, but showed no signs of recovery as new and old clearcuts differed significantly from reference streams. Both aquatic and terrestrial invertebrate drift increased in steams flowing through clearcuts. Taxa within the orders Ephemeroptera and Diptera drove the increased drift. Scrapers increased in old clearcuts, whereas collector-gatherers contributed to response in both clearcut groups. Percent shredders responded negatively to both clearcut groups, and taxa richness weighted for abundance in each sample responded negatively to new clearcuts. Site location (upstream o f or within clearcut) did not predict presence/absence o f common families and genera, but FPOM concentration was positively related to density o f family Chloroperlidae and the genus Ameletus. Our results indicate that 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. clearcut logging increased quantity and altered composition of invertebrate drift up to 7 years post harvest. We suggest that the response was primarily due to increased densities of the invertebrate benthic standing stock in clearcut streams. Changes in composition were likely due to increased benthic production that favoured multivoltine taxa, taxa that directly or indirectly benefited from increased primary production, and taxa that tend to drift. Implications for fisheries management are that short-term gains may occur, but long term stability in resources exported from headwater streams are uncertain. 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Introduction Due to the “fish-centric” nature of riparian forest management, small streams that are non-fish bearing are often the least protected in terms o f a no harvest riparian buffer (Young 2000; Cummins and Wilzbach 2005). Fish such as salmonids, however, are seasonally food limited and may receive an important subsidy o f organic matter from headwater reaches, especially when considering cumulative export (Wipfli and Gregovich 2002). It is important to consider cumulative impact because small streams exist in the highest numbers among the continuum of freshwaters and have the closest link between terrestrial and aquatic ecosystems (Vannote et al. 1980). Headwater streams are generally detrital based food webs that rely on a net import of allochthonous organic matter derived from the terrestrial ecosystem. A high degree of canopy closure results in relatively low primary production and cool water temperatures. Small streams have an important element o f large woody debris that allows retention of suspended particles and also mediates discharge. The result of these characteristics is that headwater streams export a unique component o f organic matter, including invertebrate drift, to subsidize downstream trophic webs. Invertebrates enter the drift passively due to disturbance that causes habitat deterioration, for example increased suspended particles (Culp and Davies 1983), and actively for reasons that include predator avoidance, response to resource depletion, or as part of their life cycle (Brittain and Eikeland 1988). Water temperatures are closely linked to life cycle cues and can affect invertebrate drift through altering emergence times (Vannote and Sweeney 1980). Drift is density dependent, reported by Waters (1965) to be the subset of the benthos that exceeds its carrying capacity. Thus, increased 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. invertebrate production would lead to increased drift rates. Reduced habitat quality can stimulate invertebrates to enter the drift both actively (Kohler 1984) and passively (Culp and Davies 1983). Resource depletion likely stimulates increased drift and can cause short-term spikes in drift rates (Siler et al. 2001). Drift is positively correlated with, and has shown the same trends as, benthic assemblages in response to disturbance (Waters 1965; Siler et al. 2001). However, in response to food depletion or altered habitat quality, drift rates may increase and cause a decline in benthic densities (Brittain and Eikeland 1988). Research on the effects o f streamside clearcutting on invertebrate drift is lacking in comparison to the effects on invertebrate benthic densities. Yet, drift may be more closely related to fish production than benthos because a wide range of fish species subsist primarily on invertebrate drift (Hynes 1970). Although a small percentage of the benthic community (<1 %) is suspended in the drift at one time, daily drift rates over an area are much greater than the benthic standing crop o f that area (Giller and Malmqvist 1998). Wipfli and Gregovich (2002) reported that export o f drift and detritus can support 100-2000 young-of-the-year salmonids per kilometre o f a fish bearing reach that receives water from headwater streams. In addition to the quantity of invertebrates drifting, the type o f invertebrates in the drift may also influence fish resources. Terrestrial invertebrates that drop in from adjacent vegetation are known to be an important resource to drift feeding fish due primarily to their larger size, and make up 40-70 % of salmonid diet in summer months (Nakano et al. 1999; Kawaguchi and Nakano 2001). However, most animals require the low carbon:nitrogen ratio that autochthonous sources and certain species of deciduous riparian vegetation provide (Murphy and Meehan 1991). 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Therefore, a shift from dominant allochthonous to autochthonous production may provide a higher quality food source for invertebrates and fish, if biota were facultative in their food sources. If logging affects the density or composition o f invertebrates in the drift of headwater reaches, the response can carry through to affect fish production (Bilby and Bisson 1992; Hetrick et al. 1998). Since drift is a subset of the benthic community, effects of clearcut logging on benthic invertebrates are useful as comparisons to the effects on drift; however, the mechanism of the effect and characteristics o f drift require consideration. The main effects o f streamside forest harvesting on small streams include decreased shade leading to elevated water temperatures (Brown and Krygier 1970) and increased primary productivity (Webster et al. 1983). In the initial years following logging, nutrient concentrations increase (Gregory et al. 1987) and allochthonous inputs decline (Bilby and Bisson 1992). Longer-term effects include increased water yields (Troendle and King 1985; Chamberlin et al. 1991), increased suspended particles (Macdonald et al. 2003a), and altered stream morphology due to a decline in large woody debris recruitment (Hicks et al. 1991). Increased light and increased autochthonous resources generally lead to increased benthic macroinvertebrate densities following clearcutting. For example, invertebrate densities increased three to five years following clearcutting in hardwood forests (Noel et al. 1986; Silsbee and Larson 1983; Stone and Wallace 1998). Studies in the Pacific Northwest reported increased benthic invertebrate densities in clearcuts within five years post harvest (Newbold et al. 1980; Fuchs et al. 2003; Kiffney et al. 2003; Hernandez et al. 2005). Conversely, Culp and Davies (1983) 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. found decreased benthic invertebrate densities, and increased drift rates in a coastal system within three years post harvest. While macroinvertebrate assemblages generally increase following clearcutting, changes to community composition are more difficult to interpret. One method of tracking community changes is by categorizing organisms based on their food source and method of obtaining it. Taxa are categorized among five functional feeding groups: predators, collector-gatherers, collector-filterers, scrapers and shredders (Merritt and Cummins 1996). Two o f these groups, scrapers and shredders, rely on the two main energy resources directly affected by forest harvest; light and organic matter. In headwater streams scrapers subsist primarily on aufwuchs, made up primarily of algae, and are influenced by increased light, and shredders consume leaf litter (Merritt and Cummins 1996). Collector-gatherers and collector-filterers subside on fine particles that are products o f detritus, algae, and metabolic wastes. Therefore, collectors are affected by a shift in food base that may occur through increased primary production in the short­ term, and increased detrital processing rates due to composition of early successional vegetation. Predators subsist on all non-predator groups and, therefore, the same factors that influence the remaining functional feeding groups indirectly affect predators (Hawkins et al. 1982). When invertebrate densities increased in response to clearcutting, scraper and collector-gatherer densities drove the response in the first 2-5 years post harvest (Webster et al. 1983; Gurtz and Wallace 1984; Hernandez et al. 2005), and shredder densities drove increased invertebrate densities in the longer term, 16 years (Stone and Wallace 1998) and 35-45 years post harvest (Hernandez et al. 2005). The 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. factors controlling the time frame in which scrapers decline and shredders begin to increase remains unclear. Biotic metrics are also useful measures to track changes in composition and diversity. For example, densities of groups such as orders, families and functional feeding groups, taxa richness, indices of taxa heterogeneity and evenness, and consideration of organisms’ reported tolerances to disturbance can be useful biotic metrics. Benthic bioassessment protocols commonly use many of these biotic metrics to assess stream health by comparing similarities between disturbed streams and reference conditions (Barbour et al. 1999). Alternatively, a multimetric approach evaluates a set of biotic metrics measured in streams along a gradient o f disturbance to assess stream health relative to the degree of disturbance (Karr and Chu 1999). Biotic metrics are not commonly used to describe invertebrate drift as drift does not represent the entire macroinvertebrate community and, therefore, cannot directly represent stream health. However, we used biotic metrics to detect relative differences in drift densities and composition. Among investigations of the effects of forestry in the Pacific Northwest Region, most research to date has focussed on southern and coastal areas. Some exceptions include findings in sub-boreal spruce forests in north central British Columbia. Fuchs et al. (2003) reported increased benthic densities, and Macdonald et al. (2003a, 2003b) reported increased temperature and suspended sediment in headwater streams. In montane spruce forest in the southern interior of B.C., Winkler et al. (2005) reported hydrologic effects of logging. In Englemann spruce subalpine fir (ESSF) forests, Heise (2001) reported altered benthic invertebrate functional feeding group representation and 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. decreased diversity up to five years post harvest in some but not all study streams. The ESSF biogeoclimatic zone covers 12 % of the province and is the highest elevation forested zone in the southern and central interior o f British Columbia (DeLong and Meidinger 2003). Englemann spruce (.Picea englemanni) is one of the most important commercial species in the interior due to its wood characteristics, and excellent source o f pulp (Parish et al. 1996). Small streams and their responses to clearcut logging in ESSF forests have been studied very modestly considering their ecological and economic importance. Our goal was to investigate invertebrate drift to determine if clearcut logging adjacent to headwater streams altered the export o f prey to downstream fish bearing reaches. We also tested the hypothesis that, after any initial responses, invertebrate drift would recover within the first decade post harvest. We selected this time frame for several reasons. First, the Province of British Columbia implemented the Forest Practices Code in 1995 and in order to compare streams managed under the same umbrella of regulations, we selected streams harvested after 1995. Second, we selected a range of years with varying reports o f response and recovery in headwater streams. Increased nutrients coupled with increased light cause elevated primary and secondary production within the first several years following harvest (Murphy and Meehan 1991), but it becomes unclear if recovery of the invertebrate community may occur in subsequent years within the first decade following harvest. Third, we investigated a relatively narrow range o f years within each clearcut group to minimize variability within treatments. Therefore, we compared undisturbed (“control”) streams in old growth ESSF forest to streams in “new clearcuts” 1-3 years post harvest, and streams in “old clearcuts” 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5-7 years post harvest. We posed three questions with respect to our comparisons among “control”, “old clearcuts” and “new clearcuts”. First, we evaluated response o f overall invertebrate drift to clearcutting during high seasonal discharge. Second, we investigated responses of the drift composition by utilizing a range o f biotic metrics at varying taxonomic levels, and presence/absence of common families and genera. Third, we investigated whether fine particulate organic matter (FPOM) and dissolved organic carbon (DOC) levels were associated with drift rates o f the common taxa that we identified. Methods Study Location We conducted a comparative survey using paired sites located on each selected stream within the ESSF biogeoclimatic zone in the southern interior o f British Columbia. Our requirements for selected watersheds within this forest type were that each stream be small, non-fish bearing and within clearcuts ranging from one to ten years post harvest for the treatment streams. Similar streams within undisturbed forest served as control streams. We also chose to examine two subzones within the ESSF biogeoclimatic zone. O f five candidate study areas identified in conjunction with tenure holders in the region, field reconnaissance determined that two of these areas, Bone Creek and Damfino Creek, met the criteria. In 2004, we sampled twenty-two streams within Bone Creek and Damfino Creek (Fig. 13). Ten o f the streams were within the Bone Creek study area, located approximately 255 km northeast of Kamloops and situated in the ESSF wet cold subzone Thompson Plateau variant (ESSF wc2; Lloyd et al. 1990). The ESSF wc2 is 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. characterized by 361 mm mean precipitation during the growing season, 782 cm mean annual snowfall, and 1.1 °C mean annual temperature. Bone Creek is a tributary o f the North Thompson River, which supports numerous salmonid species. A waterfall is located 150 metres upstream from the mouth o f Bone Creek that prevents fish passage, leaving the remainder o f the watershed absent o f fish (FRBC 2001). The remaining twelve streams were within the Damfino Creek study area located approximately 260 km southeast o f Kamloops within the ESSF dry cold subzone, Okanagan Highlands variant (ESSF del; Lloyd et al. 1990). The ESSF del is characterized by 261 mm growing season precipitation; 635 cm mean annual snowfall and 2.0 0 C mean annual temperature (Lloyd et al. 1990). Sites in this study area were sampled within two watersheds: Damfino Creek and Two John Creek. Two John Creek is located ten kilometres south of Damfino Creek within the same biogeoclimatic zone and variant. A waterfall located 561 m from the mouth o f Damfino Creek prevents fish passage to the upstream reaches, and barriers restrict fish migration in the upper reaches of Two John Creek where sample streams were located (FRBC 2000). Damfino Creek flows into the Kettle River, and Two John Creek flows into the West Kettle River, which then flows into the Kettle River. The Kettle River is a tributary o f the Columbia River and supports numerous salmonid species. All streams selected in this study fall under the S6 stream classification, which includes non fish bearing streams no greater than 3 m bank-full width under British Columbia’s Forest and Range Practices Act. Cutblocks selected in this study ranged in size from 20 to 95 ha and contained up to 24 identified S6 streams within one block. Best management practices prescribed retention o f the most windfirm 5 % o f codominant 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. conifers within 20 m o f the channels of two streams for every 40 ha of harvest area, and full retention o f non-merchantable vegetation within 5 m (Province of BC 1995). All treatment streams were nearly to completely void o f conifers in their riparian zones, and where conifers remained they were early serai and sparse. A portion o f streams in each treatment had non-merchantable vegetation left in the riparian zone that comprised deciduous species such as alder (Alnus tenufolia) and rhododendron (.Rhododendron albiflorum). The primary logging method used within the Damfino study area streams was conventional grapple skidding, while high lead yarding was common in the Bone Creek study area due to steeper topography. Although this study was not designed to quantify or evaluate the effects o f slash, organic matter left in or adjacent to streams is considered an important influence on stream responses to clearcut logging. Our observations were that slash presence was variable among clearcut streams ranging from very little slash present immediately adjacent to the channel, to high amounts of slash that covered the stream channel. The specific influence of slash was not tested in this study, as the objective was to evaluate effects of the end result from clearcut harvesting, regardless of site-specific differences in logging operations. 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 120W 1 18 W Southeastern British Columbia Index Map of BC Legend W ater Major Highway Provincial B oundary .QW'p 1 City o r Com m unity C n "y B one C reek S tudy A rea D amfino C reek S tu d y A rea 120 W 64 km. Map cantor: $0*49* N. 118*59* W Ministry of Sustainable Resource Management Figure 13. Locations o f Bone Creek and Damfino Creek study areas within the southern interior o f British Columbia. Experimental Design The selected streams fell into one o f three treatment categories: “control” streams, “new“ clearcuts (1-3 years post harvest), and “old” clearcuts (5-7 years post harvest). The minimum distance that a stream flowed through its associated cutblock was 300 metres at Bone Creek and 250 metres at Damfino Creek. Therefore, this minimum distance separated the downstream site from the upstream forest edge, and an additional 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50 metres separated the forest edge from the upstream site (Fig. 14). The same distances separated control stream paired sites located at elevations and physical characteristics representative of treatment streams. Streams served as replicate sampling units to avoid issues of pseudoreplication (Hurlbert 1984). Experimental Design P aired sites: R e fe re n ce stre a m , top T reatm en t stre a m , bottom Legend Upstream site Downstream site Cutblock Stream Forest vegetation Figure 14. Schematic diagram illustrating the experimental design. We addressed potentially confounding variables by judicious site selection and timing of sampling. We maintained a consistent slope distance between paired sites in order to reduce variability associated with effects that may occur as a function o f reach length. We excluded streams if they had insufficient flow to use appropriate drift sampling methods (described below), if they flowed through a known disturbance upstream from the selected clearcut, and if they contained a riparian reserve zone, defined as an undisturbed area o f timber adjacent to the stream. To avoid the additional influence 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. o f roads, we excluded streams if they contained active or decommissioned roads larger than skid trails within the paired sites. We selected reference streams with a range of physical characteristics comparable to treatment streams. Timing of sampling influences drift rates due to seasonal and diel periodicity, and discharge pulses. Due to the influence of hydrology on drift, we did not sample during or following heavy rainfall. We commenced sampling at the same time each day with downstream sites in the morning followed by upstream sites in the afternoon. We began sampling at the period o f high flows and, therefore, maximum drift in order to maximize sampling efficiency. The onset of maximum flows at Bone Creek begins later than Damfino Creek due to its glacial source and higher latitude. Therefore, we began sampling at Bone Creek following completion o f sampling at Damfino Creek. To avoid bias from seasonal fluctuations, we alternated the sampling among treatments at each study area, as opposed to sampling all control streams then all clearcut streams. Data Collection We recorded physical characteristics at each site including canopy closure, discharge, bank-full channel width, stream gradient, aspect and elevation (Table 3). At two locations for each site we used a densiometer held 1 m above ground facing four directions and averaged the readings to determine canopy closure (Lemmon 1956). We measured discharge by recording the time to fill a large container o f known volume at least three times and obtaining the average (Gore 1996). Bank-full channel widths were measured at three locations and averaged; gradient and aspect were measured using a clinometer and compass. Elevation and UTM coordinates were taken by GPS at each site location and elevations were verified using topographical maps. Temperature loggers 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. were placed in streams at each site and temperature was recorded hourly over at least twenty-four hours. We collected dissolved organic carbon (DOC) and fine particulate organic matter (FPOM) samples preceding invertebrate drift sampling at each site, and at an additional sampling period during low flow. Methodologies for these variables and spectral ultraviolet absorbance (SUVA 350 ) are described in Chapter 1. We sampled invertebrate drift in triplicate at each site using a drift net with collecting bucket and standard mesh size of 254 pm. Samples were stored in 70 % ethanol in Whirl-Pak® bags and later sorted and identified in the lab under 10-8 Ox magnifications. Total body lengths o f organisms were measured using an ocular micrometer and biomass was calculated using length-weight regressions from the literature (Smock 1980; Sample et al. 1993; Burgherr and Meyer 1997; Benke et al. 1999), or average dry mass per individual for taxa not represented in the literature (see Appendix 1). For these organisms, a sample of 20-30 individuals taken from numerous samples was dried at 60 °C for at least 24 hours, and average dry weight per individual was calculated as an estimation of biomass. All organisms were identified to family except the following taxonomic groups that were not further identified: Collembola, Isopoda, Branchiopoda, Cladocera, Copepoda, Ostracoda, Nematoda, Nematomorpha, Platyhelminthes, Oligochaeta. The orders Ephemeroptera, Plecoptera, Trichoptera (E,P,T; mayflies, stoneflies and caddisflies, respectively) were further identified to genus using Merritt and Cummins (1996), Stewart and Stark (2002), and Wiggins (1996). 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Stream ID 3 Bone Creek 005 012 013 025 021 023 002 027 010 011 Treat­ ment Control Control Control Control N ew N ew N ew New Old Old Damfino Creek 203 Control 208 Control 209 Control 213 Control 201 New 205 New 207 New 215 N ew 206 Old 210 Old 211 Old 202 Old Year logged - 2003 2003 2002 2001 1999 1998 - 2003 2003 2003 2003 1998 1998 1998 1997 Canopy closure Discharge (L sec'1) Stream gradient 5.46 12.38 2.39 4.28 1.31 0.59 0.63 3.42 0.43 1.66 0.47 2.72 2.13 0.61 0.85 0.68 0.94 1.03 2.50 1.00 35 34 32 32 41 31 57 65 58 33 W E SE NE S S E SW N NE 4.16 4.29 4.11 1.17 19.04 3.77 5.63 1.47 1.41 0.18 0.92 2.69 1.63 0.71 1.56 1.04 0.86 1.02 0.76 1.62 1.01 1.30 0.76 0.86 4 18 23 12 7 14 14 16 9 SW (% )b 25 78 94 52 0 0 42 59 0 0 77 5 73 84 63 17 55 57 92 18 39 45 Aspect Bank-full width (m) Elevation m edianc Min (°C ) Max (°C ) 1184 1131 1297 1671 1536 1845 1453 1543 1262 1171 9.8 8.9 5.7 7.9 9.9 6.1 n.d. 14.8 12.3 7.6 9.9 11.8 12.3 n.d. 9.5 9.2 9.2 11.6 11.8 1566 1592 1820 1682 1547 1660 1730 1652 1873 1825 1830 1695 4.6 8.4 6.5 n.d. 12.0 8.6 16.0 n.d. 5.5 7.3 6.1 n.d. 12.0 11.2 19.9 n.d. 6.4 7.4 6.2 n.d. 16.7 10.0 8.5 n.d. (m) (%) s NW NE SE N SE NE E N W W 24 27 8 Water Temp.d OO OO Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3. Physical characteristics and sampling dates o f study streams sampled in 2004 within the Bone Creek and Damfino Creek study areas. Date Sampled (2004) 14-Aug 24-Aug 19-Aug 27-Aug 13-Aug 13-Aug 25-Aug 29-Aug 12-Aug 26-Aug 26-Jun 9-Jul 11-Jul 21-Jul 18-Jun 29-Jun 10-Jul 23-Jul 30-Jun 13-Jul 14-Jul 25-Jun Note: Attributes measured during high flow sampling period with multiple measurements at each site. Stream average values are presented with the exception of canopy closure, which are given for downstream sites for comparison o f logged versus old growth. " All streams are unnamed; therefore, each stream was assigned a number with the leading digit indicating study area ° Canopy closure represents percentage o f cover measured by densiometer 1 m above ground at downstream sites I Represents the median elevation between downstream and upstream paired sites II Water temperaure logged hourly over at least 24 hours during sampling date o f each stream 54 Data Analysis B io tic M e tr ic s All organisms were categorized as aquatic or terrestrial and were assigned to the functional feeding groups o f predators, collector-gatherers, collector-filterers, scrapers and shredders based on Merritt and Cummins (1996). Semi-voltine organisms were considered long-lived. Using a database prepared for the Pacific Northwest by Robert Wisseman and Leksa Fore (SalmonWeb 2001), taxa were assigned a Hilsenhoff tolerance value (Hilsenhoff 1982), and designations as intolerant to organic pollution and clinger taxa. Clinger taxa not listed in this source were classified according to Merritt and Cummins (1996). The Hilsenhoff Biotic Index (HBI) and intolerant designations are regionally specific and have not been tested for our study area, but have been used previously for bioassessment in the central interior o f BC (Bennett and Hewgill 2002; Croft 2003). Terrestrial taxa and indistinct taxa were excluded from calculation of HBI. Indistinct taxa are defined as those not identified to the lowest levels previously stated due to immature or damaged specimens. Indistinct taxa were included in metrics that sum total invertebrates and in sums of orders and families depending on their level of identification. Diversity metrics included indistinct taxa if they were identified to the family level and no other family member was collected at that site. All taxa in Bone Creek and Damfino Creek drift samples and their designations are listed in Appendix 1. We quantified the density o f invertebrates by abundance and biomass of organisms per volume o f water in each replicate for the following groups: total invertebrates, aquatic invertebrates, terrestrial invertebrates, Ephemeroptera, Plecoptera, Trichoptera, Diptera, sum o f EPT, sum of EPT divided by D, and the five functional 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. feeding groups (Table 4a). Percentages of invertebrates categorized as aquatic, terrestrial, E,P,T,D, clinger, non-insect, long-lived, intolerant, and each functional feeding group were calculated as abundance o f each group divided by abundance of total invertebrates, and the Hilsenhoff Biotic Index (HBI) was calculated for each replicate (Table 4b). We used additional biotic metrics defined by Karr and Chu (1999) and Krebs (1999) to detect responses in diversity of the drift community (Table 4c). There were 52 variables in total and we refer to the group as biotic metrics. 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4. Definitions o f the 52 biotic metrics calculated based on the designations for each taxon (Appendix 1) identified from drift samples in Bone Creek and Damfino Creek. Biotic metrics are organized into three assessment categories: a) Densities, (b) Biotic indicators, and (c) Diversity. a) Densities The following metrics each represent two variables: abundance divided by volume o f water (# organisms m'3), and biomass divided by volume o f water (mg m'3). Invertebrate density: The average density o f total invertebrates collected in the three samples at each site. Aquatic Taxa density: The average density o f aquatic taxa collected in the three samples at each site. Terrestrial Taxa: The average density o f terrestrial taxa collected in the three samples at each site. Ephemeroptera density: The average density o f Ephemeroptera (E) collected in the three samples at each site. Piecoptera density: The average density o f Plecoptera (P) collected in the three samples at each site. Trichoptera density: The average density o f Trichoptera (T) collected in the three samples at each site. Diptera density: The average density o f Diptera (D) collected in the three samples at each site. EPT density: The average density o f the sum o f all Ephemeroptera, Plecoptera, and Trichoptera individuals collected in the three samples at each site. (Resh and Jackson 1993) EPT/D abundance: The average density o f the sum o f all Ephemeroptera, Plecoptera, and Trichoptera divided by Diptera individuals collected in the three samples at each site. (Resh and Jackson 1993) Predators: The average density o f predators collected in the three samples at each site. Collector-gatherers: The average density o f collector-gatherers collected in the three samples at each site. Collector-filterers: The average density o f collector-filterers collected in the three samples at each site. Scrapers: The average density o f scrapers collected in the three samples at each site. Shredders: The average density o f shredders collected in the three samples at each site. 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4 (Continued). b) Biotic indicators. % Aquatic: The abundance o f aquatic individuals divided by the total number o f individuals in each sample, averaged among three samples per site. % Terrestrial: The abundance o f terrestrial individuals divided by the total number o f individuals in each sample, averaged among three samples per site. % Ephemeroptera: The abundance o f Ephemeroptera divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Plecoptera: The abundance o f Plecoptera divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Trichoptera: The abundance o f Trichoptera divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Diptera: The abundance o f Diptera divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Clingers: The abundance o f clingers divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Non-insects: The abundance o f non-insect individuals divided by the total number of individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Long lived individuals: The abundance o f long-lived individuals divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Intolerant: The abundance o f intolerant individuals divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Predators: The abundance o f predators divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Collector-gatherer: The abundance o f collector-gatherers divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Collector-filterer: The abundance o f collector-filterers divided by the total number of individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Scrapers: The abundance o f scrapers divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) % Shredders: The abundance o f shredders divided by the total number o f individuals in each sample, averaged among three samples per site. (Karr and Chu 1999) H ilsenhoff Biotic Index (HBI): The sum o f abundance weighted by assigned tolerance to organic pollution value for each taxon, averaged among three samples per site. (Hilsenhoff 1982). 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4 (Continued). c) Diversity. Measures o f diversity, evenness and heterogeneity among taxa in drift samples.___________ Total taxa richness: The total number o f distinct taxa counted in the three samples at each site. Distinct taxa includes individuals identified only to family/order if there were no individuals identified to a lower level for that taxon (Resh and Jackson 1993, Karr and Chu 1999). Total taxa diversity density: Cumulative number o f distinct taxa in three samples at each site divided by total number individuals counted in the three samples. (Karr and Chu 1999, Krebs 1999). Ephemeroptera taxa diversity density: Cumulative number o f distinct Ephemeroptera taxa in three samples at each site divided by total number individuals counted in the three samples. (Karr and Chu 1999, Krebs 1999). Plecoptera taxa diversity density: Cumulative number o f distinct Plecoptera taxa in three samples at each site divided by total number individuals counted in the three samples. (Karr and Chu 1999, Krebs 1999). Trichoptera taxa diversity density: Cumulative number o f distinct Trichoptera taxa in three samples at each site divided by total number individuals counted in the three samples. (Karr and Chu 1999, Krebs 1999). Clinger taxa diversity density: Cumulative number o f distinct clinger taxa in three samples at each site divided by total number individuals counted in the three samples. (Karr and Chu 1999, Krebs 1999). Reciprocal o f Simpson’s index: Describes heterogeneity using cumulative numbers o f taxa in three replicates at each site. The index was calculated using Ecological Methodology software (Krebs 1999). 1/D = recip ro cal o f S im p so n ’s index P, = P ro p ortion o f species i in th e co m m u n ity Smith and W ilson’s index of evenness: Describes evenness o f the distribution among taxa based on variance in abundance o f each taxon. This index was calculated using cumulative numbers o f taxa in three replicates at each site. (Krebs 1999). The index was calculated using Ecological Methodology software (Krebs 1999). 2 arctan- £ lo g e(«,) - £ l o g e(«/) / Evar = S m ith and W ilso n ’s in d ex o f ev en n ess n, = N u m b er o f in dividuals in tax o n i in sam p le n, = N u m b er o f in dividuals in tax o n j in sam p le s = N u m b er o f tax a in entire sam ple.___________ 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For each biotic metric, we determined effect as the difference between paired sites weighted for the sum of the metric within each stream (Equation 3). Effect ranges from -1 0 0 % to + 1 0 0 %, and the effect formula achieves equal effects whether there is a positive or negative difference between sites over the same order o f magnitude. For example, if there was a near zero value (0 . 0 0 0 1 individuals m '3) at the upstream site, and a value of 0.5 individuals m ' 3 at the downstream site, effect approaches 100 %. In the opposite situation, with 0.5 individuals m ' 3 at the upstream site and nearly zero at the downstream site, effect approaches -100 %. A positive effect indicates the metric increased after flowing through the clearcut, and a negative effect indicates the metric decreased after flowing through the clearcut. Control stream effects served as the base condition to compare against treatment stream effects. ds-us E ff e c t-d s + us x l0 0 Equation 3 Where ds = downstream site metric value and us = upstream site metric value. All data were checked for normality and homoscedasticity. Where there were deviations among the data, transformations such as log (x+1), Ln (x+1), and arcsin (sqrt(x)) were applied both to the raw data prior to calculating effects, and to the effect variables (Zar 1984). The transformation, however, did not improve deviations from normality or homoscedasticity. Therefore, we continued biotic metric effect analysis on the basis that ANOVA is robust to minor deviations from these assumptions. To assess the appropriateness o f pooling data from the two study areas, we used a two-way ANOVA with treatment and study area as fixed factors. If there were no significant results for study area and no interactions between study area and treatment (p>0.05), we pooled the data and used single factor ANOVAs to detect differences 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. among treatments for the remaining variables with a priori contrasts. Contrasts were chosen to detect differences between the following groups: i) control and old clearcuts, ii) control and new clearcut streams, and iii) old and new clearcut streams. The method of comparing differences between sites spanning a disturbance using control streams as the base condition is effective for detection o f responses to disturbance (Underwood 1994). E P T F a m il ie s a n d G en e r a The effectiveness of the preceding analysis was reliant on summing organisms at a level that was common between study areas, and standardizing metrics for each stream using “effect” . At lower taxonomic groups, it was inappropriate to use the same approach for two reasons. Many families and genera were unique to each area making it impossible to pool study areas. Additionally, three quarters of the taxa occurred in less than 20 % o f the replicates. There were many empty cells at the family and genus levels and applying the effect formula would result in either zero change for these streams or 100 % if only one of the pair was zero. Therefore, presence/absence data were more appropriate for analysis of treatment responses. We excluded families and genera that occurred as less than 1 % o f total invertebrates on average in each study area. We selected all clearcut streams in Bone Creek (n=12) and Damfino Creek (n=16) and compared presence/absence data for each selected taxon as dependent variables in logistic regressions. Site location was entered as the categorical predictor variable with an indicator contrast that compared downstream to upstream as the reference condition (SPSS® Version 11.0, SPSS Inc., 2005). The minimum number of cases to independent variables recommended is 1 0 and there are no 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. assumptions o f normality, linearity or homoscedasticity in logistic regression (Hosmer and Lemeshow 2000). In the preceding chapter, we reported that FPOM increased in response to new clearcuts, and DOC responded between high and low flow sampling periods in old clearcuts. To investigate whether these effects transferred to stream biota, we used multiple linear regression analyses to assess the ability of FPOM, DOC and SUVA 350 to predict abundance o f common taxa. We conducted this analysis on lowest taxonomic levels because differing requirements among taxa of higher levels tend to mask specific responses to disturbance (Karr and Chu 1999). To reduce number o f variables and to address correlations between variables, we first subjected DOC, SUVA 3 5 0 , and FPOM to principal components analysis (PCA) using SPSS® (Version 11.0, SPSS Inc. 2005). Suitability o f the data for PCA was assessed with the following results (Tabachnick and Fidell 1989). Cases included all sites (n=44) resulting in a ratio of cases to variables of 14 that exceeded the minimum recommendation of five. Several coefficients greater than 0.3 occurred in the correlation matrix indicating suitable correlations among variables. The Kaiser-Meyer-Oklin value was 0.519, and measures of sampling adequacy exceeded 0.500 for each variable. Bartlett’s Test o f Sphericity was significant (p<0.001). We extracted two components with eigenvalues greater than 0.9 based on examination of the scree plot. The components individually explained 59.5 % and 31.5 % o f the variance for a cumulative 91 % o f the variance. We used oblimin rotation to improve interpretability and the rotated solution indicated a simple structure with DOC and SUVA loading highly on the first component, and FPOM highly on the second (Table 5). DOC and SUVA are 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. closely related as raw variables, and their linear combination into a component allowed their use with FPOM as independent variables in multiple linear regressions. Table 5. Structure matrix o f the two factor solution for organic matter variables. Oblimin rotation with Kaiser normalization was used. Component 1 DOC 0.92 0.90 Variable DOC SUVA FPOM % of Variance explained 59.50% Note: Only loadings above 0.4 are displayed Component 2 FPOM 0.93 31.40% Results Biotic Metrics We identified 78 distinct taxonomic groups within 11 classes and within 11 orders o f insects in drift samples from Bone and Damfino Creek study areas (Appendix 1). Diptera made up 38 % of invertebrate abundance in drift samples on average, the majority of these being Chironomidae larvae (14 %) and Diptera adults (13 %). Collembola made up 16 %, Plecoptera 12 %, Arachnida 9 %, Ephemeroptera 7 %, Crustacea 5 %, and the remaining orders of insects and phyla of worms made up less than 5 % of the drift samples. Mean total invertebrate drift density was 22 individuals m '3 (SEM=3 m"3, n=44), and mean biomass was 5.10 mg m'3 (SEM=0.95 mg m '3, n=44) among all sites in both study areas. Tables 6 and 7 summarize the means o f biotic metrics among treatments in each study area. 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6. Mean biotic metrics (SEM) for downstream sites at Bone and Damfino Creek study areas among treatments. Category A biotic metrics are summarized._______________________________ Control B o n e Creek Old New Control D am fino Creek Old New T o tal In v e r te b ra te s 23 (9) 23 (4) 3 4 (1 2 ) 1 0 (4 ) 23 (8) 1 8 (4 ) A qu. In v e r te b ra te s 1 5 .0 (4 .0 ) 20.1 (3.9) 29.0 (9.5) 5.7 (1.6) 13.4 (3.8) 11.2 (2.7) T err. In v e r te b ra te s 7.8 (6.1) 3.2 (0.3) 4.4 (2.8) 4.4 (2.5) 9.9 (5.9) 7.1 (2.2) E p h e m e r o p te ra 2.4 (1.1) 6.7 (1.1) 1.8 (0.8) 0.8 (0.1) 0.8 (0.4) 0 .5 (0.2) Biotic Metric Abundance 3 P le c o p te ra 2.0 (0.7) 1.4 (0.8) 1.6 (0.4) 1.1 (0.3) 1.8 (0.9) 2.8 (1.7) T ric h o p te ra 0.6 (0.1) 2.3 (1.2) 2.0 (0.6) 0.3 (0.2) 0.3 (0.2) 0 .2 (0.1) D ip tera 6.8 (3.6) 9.4 (1.9) 22.6 (12.2) 4.2 (2.5) 10.3 (3.3) 8.1 (1.8) EPT 5.0 (1.7) 10.4 (3.1) 5.1 (1.6) 2.1 (0.5) 2.8 (1.0) 3.6 (1.7) E P T /D (no u n its) 1.1 (0.4) 1.1 (0.01) 0.5 (0.2) 1.4 (0.7) 0.3 (0.1) 0.9 (0.4) P re d a to rs 3.3 (0.8) 4.0 (0.6) 11.0 (5.5) 2.5 (0.9) 5.1 (1.7) 3 .5 (0.1) C o lle c to r-g a th e re rs 5.9 (1.5) 10.8 (2.0) 12.9 (4.2) 1.4 (0.5) 3.4 (1.6) 2.6 (0.9) C o llecto r-filte re rs 0 .9 (0.8) 0.2 (0.2) 0.3 (0.1) 0.9 (0.7) 2.4 (1.7) 1.4 (0.6) S c ra p e rs 0.3 (0.1) 1.3 (0.4) 0.6 (0.3) 0.07 (0.05) 0.08 (0.08) 0.05 (0.03) S h re d d e rs 1.9 (0.7) 1.6 (1.0) 1.9 (0.5) 0.3 (0.1) 0.8 (0.6) 2.2 (1.8) T o tal In v e r te b ra te s 1.8 (0.69) 1 2 .5 (1 1 .2 ) 9.8 (5.8) 2.2 (0.8) 11.7 (5.9) 5.9 (1.1) A qu. In v e r te b ra te s 1.3 (0.5) 1.1 (0.2) 8.9 (6.0) 1.3 (0.5) 6.7 (1.9) 2.8 (1.4) T err. In v e r te b ra te s 0.5 (0.2) 1 1 .4 (1 1 .0 ) 0.8 (0.5) 0.9 (0.4) 8.1 (6.3) 3.1 (1.4) E p h e m e r o p te ra 0.4 (0.2) 0.2 (0.04) 0.4 (0.4) 0.2 (0.1) 0.1 (0.1) 0.3 (0.3) ng nr3 P le c o p te ra 0 .1 4 (0 .0 4 ) 0 .1 0 (0 .0 9 ) 0 .1 6 (0 .0 9 ) 0.10 (0.02) 0.27 (0.17) 1 .1 9 (1 .1 2 ) T ric h o p te ra 0.39 (0.32) 0 .1 3 (0 .0 3 ) 7.04 (5.57) 0.36 (0.20) 0.93 (0 .7 5 ) 0.92 (0.52) D ip tera 0 .1 5 (0 .0 7 ) 0.47 (0.15) 0.99 (0.49) 0.32 (0.15) 0.94 (0.36) 0.48 (0.21) Biomass b EPT 0 .9 6 (0.45) 0.47 (0.09) 7.64 (6.02) 0.62 (0.29) 1.34 (0.71) 2 .1 2 (1 .4 4 ) E P T /D (no u n its) 8.3 (3.6) 1.6 (0.2) 11.6 (5.8) 2.0 (0.9) 1.6 (0.7) 0.4 (0.4) P re d a to rs 0 .1 3 (0 .0 6 ) 0.21 (0.06) 0.61 (0.25) 0.49 (0.23) 1.55 (0.93) 0.23 (0.16) C o lle c to r-g a th e re rs 0 .4 8 (0.19) 0.44 (0.02) 2 .4 0 (1 .7 3 ) 0.27 (0.13) 0.59 (0.26) 0.21 (0.13) C o llecto r-filte rers 0 .0 3 (0.02) 1 (1 )g g n r3 1.02 (0.83) 0 .1 7 (0 .1 4 ) 0.39 (0.31) 0 .1 6 (0 .0 5 ) S c ra p e rs 0 .0 9 (0.06) 0.11 (0.03) 0.30 (0.22) 0.03 (0.02) 0.1 (0.1) 8 (8) gg m"3 4.43 (4.30) 0 .1 0 (0 .0 4 ) 0.95 (0.67) 1.6 (1.5) S h re d d e rs 0 .3 5 (0.23) 0 .1 6 (0 .1 5 ) Note: T h e re a r e 4 s t r e a m s in e a c h tr e a tm e n t e x c e p t B o n e C r e e k old c le a rc u ts , n = 2 . a N u m b e r o f in d iv id u a ls m '3; b m g m '3 e x c e p t w h e r e n o te d . 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 7. Mean biotic metrics (SEM) in downstream sites at Bone and Damfino Creek study areas among treatments. Category B metrics that are inherent from Table 6 are omitted, and all category C metrics are summarized. HBI = Hilsenhoff Biotic Index.__________________________ Biotic Metric Control B one Creek Old New Control Damfino Creek Old New % clingers 2.42 (0.66) 3.21 (2.13) 5.32 (1.45) 5.24 (4.17) 1.07 (0.85) 0.35 (0.31) % n o n -in se cts 27.13 (9.78) 10.60 (4.25) 13.60 (5.35) 18.33 (3.52) 25.58 (6.48) 14.49 (5.05) % long-lived 0.25 (0.13) 2.21 (1.03) 1.35 (0.64) 1.56 (0.92) 0.79 (0.67) 0.00 (0) % intolerant 2.31 (0.85) 5.08 (2.91) 3.07 (1.03) 0.57 (0.35) 0.92 (0.67) 0.24 (0.14) HBI 4.12 (0.68) 4.81 (0.02) 4.70 (0.18) 2.97 (0.44) 3.53 (0.38) 2.80 (0.80) 25 (3) 1 7 (3 ) 21 (2) 1 8 (3) 1 7 (3) 17 (2) 0.21 (0.10) T ax a rich n ess T ax a div. den sity 0.10 (0.03) 0.25 (0.03) 0.16 (0.06) 0.24 (0.04) 0.30 (0.08) E div. den sity 0.27 (0.16) 0.09 (0.02) 0.14 (0.09) 0.23 (0.02) 0.41 (0.22) 0.25 (0.25) P div. den sity 0.12 (0.04) 0.47 (0.13) 0.43 (0.20) 0.21 (0.03) 0.56 (0.19) 0.28 (0.24) T div. d en sity 0.46 (0.10) 0.75 (0.25) 0.45 (0.11) 0.93 (0.11) 0.33 (0.24) 0.53 (0.27) C linger div. density 0.19 (0.05) 0.53 (0.07) 0.35 (0.11) 0.33 (0.05) 0.63 (0.22) 0.39 (0.21) H e te ro g e n e ity a 6.30 (1.46) 4.86 (0.33) 4.91 (1.84) 6.58 (1.49) 5.45 (0.73) 4.92 (1.62) E vennessb 0.34 (0.04) 0.52 (0.01) 0.45 (0.07) 0.52 (0.04) 0.61 (0.08) 0.45 (0.11) Note: There are 4 streams in each treatment except Bone Creek old clearcuts, n=2. Div. Density= Diversity Density, units in # taxa per individual. a Reciprocal of Simpson's Index;bSmith and Wilson's Index of Evenness. Biotic Metrics: Effect (%) Compared Among and Between Treatments There were no significant interactions between treatment and study area (p>0.05), and study area did not significantly contribute (p>0.05) to the general linear model for a subset of the first fourteen biotic metric effect variables. This indicated that study area did not influence biotic metric effect variables and it was appropriate to pool data from the two study areas for further analysis. O f the 52 response variables compared among and between treatments using single factor ANOVA and a priori contrasts, ANOVA detected significant differences among treatments for 11 metrics, and a priori contrasts detected differences between treatments in an additional six metrics. Table 8 presents the mean effects (%) among treatments for each of these variables, and gives p values for each test (a=0.05). The 17 biotic metrics showed several different trends, and are presented in Table 8 according to 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. having a positive or negative response to both clearcut groups, new clearcuts or old clearcuts. Table 8. Biotic metrics that differed for comparisons among (ANOVA) and between (contrasts) treatments o f downstream-upstream effect in Bone and Damfino Creek study areas. Mean effect represents the difference between sites as a percentage. Trend group Biotic Metric Mean Effect (SEM) New Control Old Positive resp on se to clearcuts Total Invertebrates (Abd) -34 (9) 15(8) Total Inverterbates (Biom) -31 (10) 27 (25) Aquatic Invertebrates (Abd) -15(7) 28 (9) Diptera (Abd) -30(11) 21 (15) Diptera (Biom) -10(18) 61 (16) Collector-gatherer (Abd) -12(9) 67 (8) Terrestrial Invertebrates (Biom) -32 (16) 33(22) ANOVA Contrasts (p values) p value Ctrl-New Ctrl-Old Old-New 27 (9) 45(13) 33(11) 43 (13) 38 (14) 41 (12) 19(23) 0.000 0.005 0.002 0.002 0.019 0.000 0.083 0.000 0.002 0.001 0.001 0.040 0.001 0.080 0.002 0.019 0.004 0.016 0.007 0.000 0.041 0.338 0.419 0.727 0.259 0.349 0.091 0.641 Negative resp on se to clearcuts Shredders (%) 26 (14) -56 (9) -27 (19) Total Taxa diversity density -7 (8) -18(8) -31 (8) 0.015 0.114 0.030 0.040 0.007 0.350 0.291 0.290 Positive resp on se to old clearcuts Ephemeroptera (Abd) -22 (9) 67(16) -5 (30) 6(11) 77(11) -25 (35) Ephemeroptera (Biom) Collector Gatherer (%) 1 (5) 61 (13) 13(11) Scraper (Biom) -35 (20) 73 (11) 27 (40) Clinger diversity density -15(10) 36 (23) -7(14) Collector-gatherer (Biom) -14(14) 69 (11) 24 (22) 0.026 0.028 0.002 0.098 0.088 0.014 0.550 0.310 0.388 0.139 0.716 0.123 0.009 0.052 0.001 0.045 0.038 0.004 0.033 0.009 0.003 0.373 0.074 0.090 Positive respon se to new clearcuts Aquatic inverterbate (Biom) -16(18) 19(23) 43(16) 0.086 0.029 0.210 0.388 Old clearcuts » New clearcuts 8(14) Ephemeroptera (%) 0.104 0.409 0.138 0.037 60(19) -18(31) Note: Level of significance is considered p = 0.05 Abd = abundance, Biom = biomass; Control n=8, Old n=6, New n=8. We now describe our findings for each set of biotic metrics following the order they were introduced in Table 4. When referring to the significance for biotic metric comparisons among treatments, we are referring to effect (Equation 3) and not to the raw biotic metrics presented previously in Tables 6 and 7. Note that the level o f significance is set at cx=0.050, and the actual p values are listed in Table 8 for each biotic metric (effect) that significantly differs in at least one comparison. 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Invertebrate abundance, invertebrate biomass, and aquatic invertebrate abundance significantly differed between controls and each clearcut group (Fig. 15). Aquatic invertebrate biomass differed only between control and new clearcuts. Old and new clearcut treatment effects did not differ for any of these variables. The magnitude o f each effect for total invertebrate and aquatic invertebrate densities increased from control, to old, to new clearcuts. Terrestrial invertebrate abundance density did not differ significantly among or between treatments. Terrestrial biomass density tended to be higher in both clearcut groups, and the largest effect occurred in old clearcuts, which differed significantly from the control group (Fig. 15). Ephemeroptera abundance and biomass density effects were both over 60 % in old clearcuts compared to nearly or less than 0 in controls and new clearcuts (Fig. 16). Ephemeroptera density differed between controls and old clearcuts for abundance (p<0.050) and biomass (p=0.052), and differed between old and new clearcuts for both measures (p<0.050). Plecoptera and Trichoptera densities did not differ among treatments, but Diptera densities (abundance and biomass) differed between controls and both clearcut groups. Old and new clearcut effects were very similar, approximately 50%, compared to control effects o f nearly 0. The combined measures o f EPT and EPT/D did not differ significantly among treatment groups (Fig. 17). Predator, collector-filterer, and shredder densities did not differ significantly among or between treatments (Fig. 18). Collector-gatherer abundance density differed significantly between controls and both clearcut groups, whereas biomass density differed only between control and old clearcuts (Fig. 18). Neither variable differed significantly between clearcut groups. Scraper biomass differed between control and old 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. clearcuts, with an effect of approximately 75 % in old clearcuts compared with -4 0 % in the control group, whereas scraper abundance density did not differ. Percentage o f aquatic and terrestrial invertebrates and percentage of the orders Plecoptera, Trichoptera, and Diptera did not differ among or between treatments (Fig. 19). Percent Ephemeroptera differed between old and new clearcuts, with the largest effect in old clearcuts. Among percentages of functional feeding groups, collectorgatherers differed between control and old clearcut groups, and between new and old clearcuts (Fig. 20). The greatest effect occurred in old clearcuts (61 %), with controls and new clearcuts showing a near 0 effect. Percent shredders differed between control and both clearcut groups, and did not differ between clearcut groups (Fig. 20). Contrary to all effects described thus far, percent shredders declined in both clearcuts compared to the control group. Effects o f the remaining biotic indicators, percentage o f clingers, non­ insects, long-lived taxa and intolerant taxa, and the Hilsenhoff Biotic Index, did not differ among or between treatments. Total taxa diversity density responded negatively to new clearcuts but there was not a significant difference between effect in old clearcuts and controls (Fig. 21). Effect in old and new clearcuts also did not differ significantly. Clinger diversity density responded positively to old clearcuts and did not differ between control and new clearcuts, or between clearcut groups (Fig. 21). The remaining effects for diversity, heterogeneity, and evenness measures did not differ among or between treatments. 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABUNDANCE BIOMASS 1001 a) Total Invertebrates _ 50- o St 0 ■ B B m -50 H -100 1 0 0 -i b) Aquatic Invertebrates 50 - X -50 - ab -100 J 100 _ 50 £ O st= o c) Terrestrial Invertebrates X A 3-1 T m -50 -100 Control Old New a ~T Control ab Old New Figure 15. Densities o f (a) total invertebrates, (b) aquatic invertebrates, and (c) terrestrial invertebrates found in control, new and old clearcut drift samples in Bone and Damfino Creek study areas. Abundance densities are on the left, biomass densities are on the right. Control mean effects are open columns (n=8), old clearcuts are hatched (n=6), new clearcuts are grey (n=8), and error bars represent SEM. A positive effect indicates that drift densities increased from the upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05). 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 n ABUNDANCE a) Ephemeroptera BIOMASS 50 - T -50 - ab -100 -J 1 0 0 -i b) Plecoptera 50 - X T -50 - -100 J 100 n c) Trichoptera 50 - a a a Old New -50 -100 -* 1 0 0 -i _ 50 - O 0 sP d) Diptera 1= ^ = r m -50 -I -100 J Control B B Old New T a Control Figure 16. Densities o f (a) Ephemeroptera, (b) Plecoptera, (c) Trichoptera, and (d) Diptera found in control (n=8), old (n=6) and new (n=8) treatment drift samples in 2004. Abundance densities are on the left, biomass densities are on the right. A positive effect indicates drift densities increased from the upstream to downstream sites. Error bars represent SEM. Means that share a letter do not differ significantly (p>0.05). 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BIOMASS ABUNDANCE 100 a) EPT X _ 50 d O £ 0 , L_ r_ J H 1H r - J^ L—,j.., A A A m -50 J . -100 -J 100 1 ' b) E P T /D 50 T 1 A A < -100 J T T Old New a — | O £ HI T T Control Old New Control Figure 17. Densities o f (a) EPT and (b) EPT/D found in control (n=8), old (n=6) and new (n=8) treatment drift samples in Bone and Damfino Creek study areas. Error bars represent SEM. Abundance densities are on the left, biomass densities are on the right. A positive effect indicates that drift densities increased from the upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05). 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 ABUNDANCE BIOMASS a) Predators —. 50 O 0 £ ^ -50 T T- _L T A a -100 100 b) Collector-gatherers 50 - ri= ^ = r ab -50 ■ -100 100 c) Collector-filterers 1 1 -50 -100 J 1 0 0 -i d) Scrapers 50 - -50 - -100 -I 100 ■r a b ab e) Shredders T -50 ■ 100 C ontrol Old N ew C ontrol Old N ew Figure 18. Densities o f (a) predators, (b) collector-gatherers, (c) collector-filterers, (d) scrapers, and (e) shredders in control (n=8), old (n=6) and new (n=8) drift samples in Bone and Damfino Creek. Abundance densities are on the left, biomass densities are on the right. Positive effect indicates that drift densities increased from upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05). Error bars represent SEM. 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 100 -t a) Aquatic invertebrates 50 - b) Terrestrial Invertebrates 50 - 1 -50 - -50 - -100 J -100 -I 100 50 - -50 - 1 0 0 -I c) E phem eroptera _ 50 - O it o m -50 AB -100 -100 J 100 -. d) Plecoptera 100 -, e) Trichoptera _ 50 - f) Diptera 50 ’ T 0s * O it LU -50 - -100 J Control Old T 0 1 A A A - Control Old New -50 - Figure 19. Percentage o f total invertebrates made up o f (a) aquatic invertebrates, (b) terrestrial invertebrates, (c) Ephemeroptera, (d) Plecoptera, (e) Trichoptera, and (f) Diptera in control (n=8), old (n=6) and new (n=8) treatment drift samples in Bone and Damfino Creek study areas. Error bars represent SEM. A positive effect indicates that drift densities increased from the upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05). 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 100 a) Predators 50 . 50 0 0 m -50 -50 o !t f) Clingers JL T A A A -100 -I 100 b) Collector-gatherers % _ 50 o ■ g) Non-Insects T j: o 0 m -50 -50 -100 -100 100 _ 50 VP 100 c) Collector-filterers x A -100 100 -| d) Scrapers i) Intolerant 50 - 50 1 n . it Hi I T 0 -100 vS> h) Long-lived ‘ A -50 - A ^ A 50 JL o 0 £ *-50 100 I A f r sn-1—1 i i t i f— nu _- I^ -50 - A A -100 100 _ 50- u 0 -100 100 -| e) Shredders 1 A j) Hilsenhoff Biotic Index 50 - i B ~l~ i E s im . 0 Ol -50 100 -J Control 1 A ' ..... j : ..... ■50 ■ -50 - Old -100 -I Control New Old New Figure 20. Percentage o f groups by abundance, followed by a biotic index in control (n=8), old (n=6) and new (n=8) treatment drift samples in 2004. Groups are: (a) predators, (b) collectorgatherers, (c) collector-filterers, (d) scrapers, (e) shredders, (f) clingers, (g) non-insects, (h) longlived, and (i) intolerant. Biotic index is (j) Hilsenhoff Biotic Index. Positive effect indicates that drift densities increased from upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05). Error bars represent SEM. 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50 -I b) Taxa Diversity Density a) Total Taxa Richness 25 AB 0 -25 -25 - -50 -50 - c) E Diversity density B J 50-i d) P Diversity Density 25- -20 o -45 -25 -70 -50 -I 50 A T f) Clinger Diversity Density e) T Diversity density 50 25 - 25 o o 3= m-25 4 0 -25 - T B T AB -50 -J -50 - 50 i g) Reciprocal of Simpson’s Index 501 h) Smith and Wilson’s Index _25 o o st= m -25 25 - r o 1 I A A -25 - A A x .. A -50 -50 ^ N ew C ontrol Old Control Old N ew Figure 21. Mean effect among control (n=8), old (n=6) and new (n=8) treatment drift samples in 2004 of: (a) total taxa richness; taxa diversity density o f the following groups: (b) total taxa, (c) Ephemeroptera, (d) Plecoptera, (e) Trichoptera, and (f) clingers; and biotic indices measuring: (g) heterogeneity, reciprocal o f Simpson’s index and (h) evenness, Smith and Wilson’s index. Positive effect indicates that metrics increased from upstream to downstream sites. Means that share a letter do not differ significantly (p>0.05). Error bars represent SEM. 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. EPT Families and Genera Among the orders Ephemeroptera, Plecoptera and Trichoptera, we identified 18 families and 34 genera overall with differing representation between study areas (Table 9). In Bone Creek Ameletidae, Baetidae, Heptageniidae, Chloroperlidae, Nemouridae, Peltoperlidae, Ameletus, Baetis, Cinygmula, Zapada, and Yoraperla occurred as greater than 1 % of total invertebrate abundance. In Damfino Creek, Ameletidae, Chloroperlidae, Nemouridae, Ameletus, Haploperla, Suwallia, and Podmosta occurred as greater than 1% o f total invertebrate abundance on average (Table 9). Two families, Ameletidae and Baetidae, were represented by one genus each (Ameletus and Baetis, respectively) and were, therefore, redundant in the analysis due to the family densities being identical to the respective genus. We conducted 13 logistic regressions that all showed site location (upstream of or within clearcuts) did not predict presence/absence o f common taxa. There were two analyses where standard error of the beta coefficient for site location was greater than 2.0, indicating problems such as all cases having the same value or complete separation. Examination o f these taxa, Nemouridae and Zapada in Bone Creek, revealed two important points. First, their presence/absence data were identical as Zapada made up the majority of the family. Second, presence/absence data showed nearly all cases had the same value, as Zapada occurred in all clearcut sites except one new and one old downstream site. The remaining 11 logistic regressions were not significant with Damfino Creek p values for overall model fit ranging from 0.614 - 1.00, and Bone Creek p values ranging from 0.244 - 0.557. 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. To meet the assumptions of linear regression, we examined standardized residuals to assess outliers, and excluded cases with values between -3.3 and 3.3 (Tabachnick and Fidell 1989). This occurred only in regressions that were not significant, and excluding outliers did not change their results. We assessed scatterplots of residuals and found the assumptions o f normality, linearity, and variances evenly distributed along the range of the dependent variables were met in all except two regressions. Residuals for Nemouridae in Damfino Creek were not normally distributed. We applied logarithmic and square root transformations recommended for abundance data (Rosenberg and Resh 1993), but they did not improve the distributions and therefore we proceeded with untransformed variables. Tolerance values approached 1 showing singularity between independent variables. O f the 15 linear regressions using DOC and FPOM (PC A components 1, 2) entered together, two models were significant. In Bone Creek, Chloroperlidae abundance was significantly related to organic matter that explained 31 % of its variance (F2,i7=3.744, p=0.045). Within the model, DOC was not significant but FPOM contributed significantly to the model (t=2.634, p=0.017). The model did not improve appreciably by removing DOC (adjusted R2=0.31, p=0.012). Organic matter predicted abundance o f Ameletus in Damfino Creek (F2,2i=3.568, p=0.046), with FPOM contributing significantly to the model (t=2.431, p=0.024). The model improved by removing DOC, and FPOM explained 25 % of the variance in Ameletus (p=0.014). There were positive relationships between the independent variable representing FPOM concentration, and abundance of Chloroperlidae and Ameletus in Bone Creek and Damfino Creek study areas, respectively (Fig. 22). 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 9. Representation o f EPT families and genera among treatments in drift samples at Bone Creek and Damfino Creek study areas. Dots indicate presence of genera listed under each family. mean % E Ameletidae Ameletus Baetidae Baetis Ephemerellidae Attenella Drunella Heptageniidae Cinygmula Epeorus (Iron) Epeorus (Ironopsis) Rhithrogena P Chloroperlidae Haploperla Suwallia Sweltsa Leuctridae Despaxia Nemouridae Podmosta Zapada Peltoperlidae Yoraperla Soliperla Perlodidae o f drift 1.8 1.8 2.6 2.6 0.3 0.05 0.02 11 1.0 0.2 0.1 0.01 1.0 0.2 0.4 0.1 0.01 Brachycentridae Micrasema Glossosomatidae Glossosoma Hydropsychidae Parapsyche Limnephilidae Chyranda Cryptochia Desmona Eocosmoecus Homophylax Onocosmoecus Psychoglypha Spagnophylax Philopotamidae Wormaldia Rhyacophilidae Himalopysche Rhyacophila Uenoidae • • • • • • • • 0.1 0.1 • • • • • • • • • • • • • • • • • • • • • • • • • 9.4 9.3 0.04 • 0.1 0.01 0 • • • 0.8 0 • 0.1 0.1 0 • • • 0.1 0.1 • • • • • • • • • • • • • • 0 0 • 0 0 0.5 0.1 0.5 0.01 0.4 • 0 0.8 0.7 0.3 0.3 • 0.2 0.02 0 0 0 3.5 1.5 1.3 0 New PS US 0 0 0.3 0.1 0.2 07 0 0 0.2 0.03 0.1 0 0.2 0 Damfino Creek Control Old PS US PS US 0.03 0 0.03 • • mean % o f drift 1.6 1.6 • 1.9 1.0 0.03 Taenioptergydae Moselyana • 3.9 0.03 3.2 02 0.2 Allomyia • New PS US 0.5 0.5 Setvena T Apataniidae BONE CREEK Control Old PS US PS US 0 0 • • • • • • • 0.5 0.1 0.2 0.03 0 0 0.01 0.1 0.01 0.03 0.03 0.1 0 0.1 • 0.2 0.2 • • • • • • • • • • • • • • • 0 0 • • • Neothremma Note: Mean % ot drift calculated using a b u n d an c e s a t e ac h site within study a re a s , treatm ents, and locations. uS = upstream , D S=dow nstream site location. T h e re a re 4 stream s in e ac h category e x ce p t B one C reek old clearcuts w here n=2. Indistinct sp ecim en s included in family num bers, adults (terrestrial) w ere excluded. 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • 2.0 E 1U ' =tt ]aj 0.5 - a5 £ §• 0 .5 - 0.0 1.5 -0.5 0.5 0.0 -0.5 1.5 RCA Comp. 2 (FPOM) 0 0.5 PCA Comp. 2 (FPOM) Figure 22. FPOM as PCA Component 2 was significantly related to (a) Chloroperlidae density at Bone Creek and (b) Ameletus density at Damfino Creek study area. Discussion Biotic Metrics Mean drift of invertebrates in this study was 10-34 individuals m '3 and is in the mid-range o f densities reported for other small streams throughout the world (0.02-143 individual m '3, see Giller and Malmqvist 1998). In North America, our drift rates were comparable to an average 29 individuals m"3 found by Siler et al. (2001) in hardwood forested streams, and to 1-22 invertebrates m'3 reported by Wipfli and Gregovich (2002) in Alaska. Assuming discharge and drift were constant through 24 hours, our streams exported 652-3813 mg o f invertebrates stream-1 day 1 on average among treatment groups, compared to that reported in Alaska of 163 mg stream _1 day _1 (Wipfli and Gregovich 2002). We only measured drift rates during the day in a season when drift was expected to be high coinciding with high discharge, whereas the other researchers sampled drift throughout the day and year. Although drift diel patterns are reduced in non-fish bearing streams, a pattern o f increasing drift in late afternoon and evening is still expected (Smock 1996). Therefore, our densities represent minimum estimates of invertebrate drift during a high drift seasonal period. However, if invertebrate drift rates 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. that we reported were representative for the summer period, these headwater streams have the potential to subsidize downstream food webs considerably in the season when fish are generally food limited. Total invertebrate drift responded positively to clearcuts within 1-3 years post harvest and did not recover within seven years post harvest. Our finding that overall drift increased is similar to the benthic invertebrate density responses to clearcut logging for the same time frame (Newbold et al. 1980; Webster et al. 1983; Noel et al. 1986; Stone and Wallace 1998; Hernandez et al. 2005). Increased drift corresponds directly with the findings of Wilzbach et al. (1986) who reported increased drift densities in clearcuts seven years post harvest in third order streams in Oregon. Our findings of the composition changes, discussed later, support that elevated primary production lead to increased macroinvertebrate production and increased drift during the seven year post harvest period that we examined in this study. In addition, the disturbed habitat likely favoured production of short-lived multivoltine taxa that further contributed to increased macroinvertebrate production (Gregory et al. 1987). An additional factor in our study is that alder (Alnus sp.) was a common deciduous species that was either left intact, or quickly regenerated adjacent to streams in both clearcut types. This affects invertebrate production in two ways. Alder has a faster processing rate than coniferous species or other deciduous species and can therefore supply more quality food to the stream food web (Cummins et al. 1989). Secondly, increased drift densities are associated with alder in the riparian zone. Wipfli and Musselwhite (2004) found that amount o f red alder in the riparian zone was positively related to aquatic taxa in the drift, and Allan et al. (2003) reported that more terrestrial invertebrates tended to drop from red alder than hemlock and spruce vegetation. Both aquatic and terrestrial invertebrates drove increased drift in 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. clearcuts, and the percentage of both groups was not altered in response to clearcut treatments. Terrestrial invertebrates are generally larger than aquatic invertebrates (Allan 1995), and this could explain why we detected a response in biomass o f terrestrial invertebrates without detecting a response in abundance. Edwards and Huryn (1995) also found that although biomass o f terrestrial invertebrates in the drift differed in response to land use, abundance of terrestrials did not. Drift feeding fish select terrestrial invertebrates due to their larger size and their differential availability compared to aquatic invertebrates on a seasonal and diel basis (Nakano et al. 1999). Terrestrial invertebrates comprise an important part of adult salmonid diets (Bridcut 2000; Kawaguchi and Nakano 2001; Allan et al. 2003); however, Bilby and Bisson (1992) attributed increased autochthonous inputs to increased production of juvenile coho salmon in headwater streams seven years following clearcutting. Wilzbach et al. (1986) found a strong correlation between trout growth and drift densities and reported short-term relative growth rates of trout in clearcut streams greatly exceeded that of reference streams. In the period up to seven years post harvest in the present study, clearcuts appeared to cause more aquatic invertebrates and larger terrestrial invertebrates to drift, both of which could lead to increased fish production. Diptera abundance responded positively to both clearcut treatments, but maximum effect occurred in new clearcuts. Diptera biomass density showed the highest effect in old clearcuts. This effect is parallel to that of terrestrial invertebrates and the order Diptera was highly influenced by adults in the drift. While abundance of Diptera responded positively 1-3 years post harvest, a positive response o f Ephemeroptera was delayed until 5-7 years following clearcutting. Both abundance and biomass of 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ephemeroptera had maximum effects in old clearcuts. These two groups of insects are reported to respond positively to disturbance in benthos, in part because they comprise short-lived taxa that are able to exploit disturbance (Wallace and Gurtz 1986, Gregory et al. 1987). For example, Noel et al. (1986) found that Ephemeroptera and Diptera drove 2-4 fold increases of benthic invertebrate density within three years post harvest. Three multivoltine taxa drove increased densities found by Newbold et al. (1980) 1-5 years following harvest in Northern Californian headwater streams. A shredder stonefly, Nemoura, showed the highest increased density and drove the elevated total benthic invertebrate density in combination with increases to Chironomidae (Diptera) and a collector-gatherer mayfly, Baetis. Because these taxa are able to colonize rapidly, they are able to increase densities rapidly and are therefore favoured in a disturbed habitat. Although the authors noted the shift to multivoltine taxa, they primarily attributed increased densities to a shift in food base resulting from increased light, nutrients and temperature following logging. These results correspond with Wallace and Gurtz (1986) who reported that initial increases and subsequent declines o f Baetis density coincided with changes in primary production. Percent shredders responded negatively to clearcuts while two other functional feeding groups, collector-gatherers and scrapers, responded positively to clearcuts. The response o f shredders was negative in the first 1-3 years, and no recovery was found seven years post harvest. This finding agrees with Gurtz and Wallace (1984), who found declines in benthic shredder abundance five years post harvest. In the longer term, Stone and Wallace (1998) and Hernandez et al. (2005) found that shredder densities drove a positive response in overall invertebrate benthic densities in clearcuts compared to reference streams 16 and 35-45 years post harvest, respectively. In the present study, 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. increased non-shredder densities are partly responsible for the decrease in percent shredders, and overall shredder densities were not significantly altered in either clearcut treatment. The post harvest increase of shredders reported in benthic invertebrate research did not occur within seven years post harvest in the present study. In several studies where invertebrate benthic densities increased in response to clearcutting, scraper and collector gatherer densities drove the response in the first 2-5 years post harvest (Webster et al. 1983, Gurtz and Wallace 1984, Hernandez et al. 2005). Each study attributed the response of these trophic groups to elevated primary production due to increased light, temperature and nutrients in clearcut streams. Our finding that the scraper effect only differed significantly from controls in clearcut streams 5-7 years and not 1-3 years post harvest is contrary to the findings o f Gurtz and Wallace (1984), who found that scraper benthic densities declined five years post harvest. A possible explanation is that increased drift o f scrapers resulted from degradation o f their habitat in the old clearcut treatment. Siler et al. (2001) found that percent o f benthos in the drift increased for detritivores in a habitat where detritus was removed; however, overall drift densities were higher in the reference conditions regardless o f percent o f benthos in the drift. In the present study, there was the tendency for scraper densities to increase in both clearcut groups, with effects of 73 % in old clearcuts and 27 % in new clearcuts, compared with -35 % in control streams. Scrapers made up the smallest percentage of the drift among functional feeding groups. Therefore, our ability to detect differences among treatments may have depreciated due to a number of zero counts among sites for this variable. Collector-gatherer was the only functional feeding group that significantly differed in abundance, biomass, and percent of invertebrates measures calculated within 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the drift samples. For each variable, maximum effect occurred in old clearcuts. This trend differed from total invertebrate, aquatic invertebrate, and Diptera abundance, but was parallel to Ephemeroptera abundance. The majority o f Ephemeroptera taxa identified in our drift samples were collector-gatherers (6 o f 11), including two common genera, Ameletus and Baetis, and the remainder were scrapers. Unlike Ephemeroptera abundance, response of collector-gatherers was significant for both clearcut groups. This illustrates that by examining the functional feeding groups we detected a more specific response than was possible using an order level biotic metric. We found that fine particulate organic matter increased in new clearcuts but recovered in old clearcuts, and this would specifically affect collector-gatherers. An additional consideration is that trophic groups have differing propensities to enter the drift. Collector-gatherers tend to be swimmers and sprawlers that more actively drift, whereas scrapers tend to be clingers and passively enter the drift (Merritt and Cummins 1996). Among the taxa we collected, collector-gatherers and collector-filterers had the least number of clinger taxa in their groups. Therefore, a change in the benthic density o f collector-gatherers may be more readily detected in the drift than other groups. Clinger diversity weighted for number of clingers in a sample significantly responded to old clearcuts. The standard errors in this variable were extremely high. In 13 of 44 sites clinger taxa did not occur and percent o f clingers was low, ranging from 0.35 - 5.25 % among treatments (Table 7). It appears that our ability to detect differences in clinger diversity density may be limited due to their low occurrence and number o f zero counts in the data. Benthic clinger diversity is predicted to decline with increased levels of disturbance (Karr and Chu 1999). Our result that diversity of clingers 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. increased in old clearcuts complements our findings that total taxa diversity density recovered in that treatment. The density of taxa diversity showed a negative response to new clearcuts (Fig. 21b). Decreased taxa richness occurred in benthic bioassessment with increasing degree o f disturbance (Fore et al. 1996; Karr and Chu 1999), and due to effects of clearcut logging (Quinn et al. 2004). Collier (1995) found that taxa richness was positively correlated with shade but not with proportion of native forest cover in the riparian zone. This corresponds with our finding assuming that shade increased in old clearcuts with regrowth of early successional vegetation. Conversely, Hernandez et al. (2005) did not find differences in taxa richness between five year old clearcuts and control streams. In the present study, taxa richness without weighting for number of organisms sampled did not differ among treatments (Fig 21). Gotelli and Colwell (2001) emphasize several notes of caution in using diversity density, and suggest careful consideration to determine the variable that best determines sampling intensity, for example time, number of organisms, or volume of water in the case of drift samples. Despite our efforts to achieve consistent sampling intensity by adjusting length o f sampling time according to discharge o f the streams, volume of the drift sample differed by up to an order of magnitude higher among the samples collected from the study sites. There was a wide range in numbers of organisms collected in samples, and the larger volume samples did not necessarily have the highest number of organisms. Examination o f these data guided us to use number o f organisms in each sample as the determination o f sampling intensity, because the number of taxa identified tends to increase with number o f individuals in a sample (Krebs 1999). Our results showing decreased taxa diversity density indicates decreased stream health 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. initially following clearcutting, and recovery with respect to this metric 5-7 years post harvest (Karr and Chu 1999). The importance o f considering taxa richness within each order, Ephemeroptera, Plecoptera, and Trichoptera was emphasized by Karr and Chu (1999) because taxa in these groups respond differently to disturbance, thus combining them masks their more specific responses. We did not detect significant responses for diversity density in each of these orders (Fig. 21). Ephemeroptera, Plecoptera, and clinger diversity density effects showed a trend of positive response in old clearcuts. Trichoptera diversity density showed a similar trend to taxa diversity density, with most significant negative effects in new clearcuts, and intermediate values in old clearcuts. Our ability to detect differences in these biotic metrics may have been hampered by the patchy distribution of EPT among drift samples. Plecoptera occurred as 12 % of the drift, Ephemeroptera as 7 %, and Trichoptera as 4 % of the drift on average among all sites. Even by pooling replicate samples at each site, 10 o f 44 sites did not contain Ephemeroptera, three did not contain Plecoptera, and nine did not contain Trichoptera. This would cause numerous zero counts through the data set, and would decrease the likelihood of detecting a difference among treatments. Even though we did not detect significant differences, the results of two differing trends among the orders supports our previous arguments that differing responses to clearcut logging occurred in 1-3 year old versus 5-7 year old clearcuts. Other measures of the community composition such as the indices of heterogeneity and evenness did not respond to clearcut treatments and mean effects showed similar values among treatments (Fig. 21 g,h). A related index, Shannon diversity, was greater in benthos o f clearcut streams 35-45 years post harvest, but similar to reference streams five years post harvest (Hernandez et al. 2005). Newbold et al. 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (1980) found that Shannon diversity decreased in stream 1-5 years following logging. We chose the reciprocal o f Simpson’s index rather than Shannon diversity to describe heterogeneity in order to weight common species more than rare species in our analysis (Krebs 1999). Reciprocal o f Simpsons (1/D) varies from 1 to the number of species in a sample and is interpreted as the number of equally common species in a community. Like 1/D, Shannon diversity (H’) increases with increasing number of species in the community. Unlike benthic samples, drift samples do not represent the entire community o f invertebrates and it appears the responses to logging do not alter the heterogeneity or evenness of the community within the drift. EPT Families and Genera Site location did not predict presence/absence o f families and genera among clearcut sites. Examining the presence/absence of EPT taxa between site locations among treatments (Table 9) revealed no clear patterns with the possible exception o f Trichoptera taxa in Bone Creek. More Trichoptera taxa were identified in downstream than upstream samples o f both old (ds=5, us=3) and new (ds=8, us=4) clearcuts compared to five taxa identified in downstream sites and six taxa in upstream sites of control streams. However, in control streams different taxa were present depending on site location in all three orders, but especially in Trichoptera o f both study areas. This variation may be dependent on differences between site location such as elevation, or time of day that site locations were sampled, or may be due to natural variability among sites. In any case, high variability reduced our ability to detect differences in presence/absence o f taxa dependent on clearcutting. Abundance of Chloroperlidae at Bone Creek, and Ameletus at Damfino Creek were positively related to FPOM concentration. We found that FPOM concentrations 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. responded positively to new clearcuts. Considering that abundances of some common taxa were dependent on FPOM concentrations, this may in part explain the highest effect of invertebrate drift in new clearcuts. Ameletus are collector-gatherers and are linked to FPOM as a primary food source. One explanation for the relationship not occurring in Bone Creek is that FPOM concentrations were generally higher in that study area whereas in Damfino Creek FPOM concentrations may have been in the range that would be more limiting as a resource. Both genera of Chloroperlidae that occurred in drift samples were predators. Murphy and Hall (1981) reported that predator benthic density increased in response to clearcuts five and more years post harvest, and attributed the effect to increased prey through elevated primary production. Hawkins et al. (1982) reported a significant correlation between benthic densities o f predators and collectorgatherers in riffles. Another possible explanation is that drift density of Chloroperlidae, which are clingers, was related to FPOM concentration physically causing invertebrates to drift. This phenomenon was reported by Culp and Davies (1983), where negative pressure created by increased suspended particles caused invertebrates to enter the drift. Experimental Approach There were limitations to the present study. This is exemplified by several metrics that had distinct negative or positive values in control streams. Before-After Control-Impact (BACI) design and comparative surveys also have their limitations in aquatic research due to high natural variability among individual streams. In this design we used paired sites on streams upstream and within disturbance, but also established a base condition using paired sites on undisturbed streams. We recognize that differences between paired sites may occur independent of disturbance due to factors that may depend on elevation such as amount o f light, water temperatures, and amount of canopy 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. closure. In addition, drift has diel patterns that, although subtle, still prevail in non-fish bearing streams (Smock 1996). We chose to consistently sample downstream sites in the morning and upstream sites in the afternoon. The effect o f terrestrial and aquatic invertebrate contributions to the drift reflect this timing o f our sampling given known diel differences in these variables. In all treatments, percent aquatics in downstream sites exceeded upstream sites, and percent terrestrials in upstream sites exceeded that of downstream sites. Drift of aquatic invertebrates tends to peak just before dawn and decrease throughout the day, opposite to the diel drift patterns of terrestrial taxa (Nakano et al. 1999). In control streams, the percentage of scrapers in the drift was higher at upstream than downstream sites, while the remaining functional feeding group percentages were either higher at downstream sites or were similar between sites. This may be the result o f more light reaching streams at higher altitudes due to less canopy closure on approaching the alpine. This may also lead to increased productivity at upstream sites that accounts for the observed higher densities of invertebrates at upstream versus downstream sites o f control streams (Fig. 15). Implications and Summary Logging can increase the trophic resources for drift-feeding fish such as salmonids if they are food limited (Bilby and Bisson 1992; Hetrick et al. 1998). There are several ways that invertebrate responses to logging adjacent to headwater streams may achieve this. Increased light, temperature and nutrients following clearcut logging stimulates increased benthic invertebrate production. The taxa that drive increased benthic densities following clearcutting are those that tend to drift (Newbold et al. 1980; Noel et al. 1986). Although allochthonous resources may be reduced, lignin-based leaf litter from coniferous trees have a higher ratio of C:N and require further breakdown by 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. microbes to provide the same food quality to macroinvertebrates as autochthonous resources supply immediately (Murphy and Meehan 1991). This shift may transfer up the food web positively because at least one salmonid species, coho salmon, appears to consume primarily autochthonous food sources (Bilby and Bisson 1992). The drift densities we reported indicate that headwater streams in ESSF forests in British Columbia export significant amounts o f invertebrate drift to downstream fish habitat. Drift more accurately estimates prey availability than benthos and is positively correlated with fish production (Wilzbach et al. 1986). Our results indicated that clearcut logging up to seven years post harvest caused increased macroinvertebrate drift densities. By not altering the contribution o f aquatic versus terrestrial invertebrates, clearcut logging did not adversely alter the composition o f the drift with respect to fish production. Our results showed that other changes in community composition occurred. Some alterations were detected 1-3 years post harvest, while others onset 5-7 years following harvest. A number of alterations persisted through the 5-7 year post harvest period. We have used a number o f biotic metrics to assess relative differences in invertebrate drift and numerous metrics enabled detection o f differences among treatments. We caution against considering the metrics independently but do make inferences based on the group of responses that we detected. It is apparent from these responses that composition o f the invertebrate community was altered up to seven years post harvest. Although short-term gains in number and biomass of invertebrates drifting in headwater streams are apparent from these data, clearcut logging altered the composition of invertebrate drift that persisted at least seven years post harvest. Both responses are concerns for fisheries management, and the longer-term effects of clearcut logging on export from headwater streams requires further consideration. 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Epilogue There are several linkages between the responses we found in FPOM and DOC, and those o f invertebrate drift. Total invertebrate drift responded positively to both post­ harvest periods but was greatest in new clearcuts, partly following the response of FPOM that increased in new clearcuts. Collector-gatherers directly consume FPOM and this functional feeding group made up the majority o f invertebrates in drift samples. In addition to this connection, collector-gatherer taxa were shown to derive at least 20 %, and up to 100 % of their carbon from bacteria (Hall and Meyer 1998). Bacteria are known to respond rapidly to changes in DOC (Kreuztweiser and Capell 2003), in particular labile DOC o f low molecular weight that is generally autochthonous (Allan 1995). Although the number o f collector-gatherers responded positively to both “new” and “old” clearcut groups, positive responses in biomass and percent only occurred in old clearcuts. The latent positive responses of this group to old clearcuts may be in part, indirectly related to altered DOC composition. The only other group to show a significant positive response was the scraper functional feeding group in response to old clearcuts; this suggests that scrapers responded to increased primary productivity that occurred in old clearcuts. Scraper densities and primary production responses normally occur in the initial years post harvest, rather than the delayed response that we observed. Our DOC results correspond with this finding. We inferred that increased autochthonous production in old clearcuts through the summer season caused the observed seasonal decline of SUVA in old clearcuts. The response in scraper densities indicated increased autochthonous production during the high flow season. 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Results from the two approaches also showed different responses. Responses in organic matter were dominant in the 1-3 year post harvest period, whereas biotic metric responses occurred in both periods with a greater number of responses in the 5-7 year period. Removal o f riparian vegetation primarily affects the macroinvertebrate community through increased light and alterations to detrital inputs. Although the quantity of organic matter may be greater in undisturbed streams, an increased proportion of higher quality autochthonous resources allows an increase in energy availability in small streams after canopy removal (Bilby and Bisson 1992). This shift may be delayed while fast processing streamside debris breaks down in the initial years providing a pulse of organic matter o f lower nutritional quality. This boosts production overall, but the later onset o f increased autochthonous production, inferred from the scraper response in old clearcuts, indicates onset of increased higher quality resource production was delayed. Headwater stream ecosystems are a source for many downstream energy resources. Not only are the amounts o f detritus and invertebrate drift important (Wipfli and Gregovich 2002), the source o f these resources is of consequence because o f its influence on nutritional quality and food web stability. Forested headwater streams may have more food web stability than larger reaches due to the dominance of allochthonous inputs or donor control (Woodward and Hildrew 2002), and the high number o f weak trophic linkages (McCann et al. 1998). Therefore, although a shift from allochthonous to autochthonous resources induced by clearcut logging can result in increased production initially, trophic stability in small streams may be compromised. This examination of the effects of changes in abiotic and biotic variables from clearcutting adjacent to headwater streams in ESSF forests is one of the only studies to 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. use these two approaches simultaneously in the interior o f BC. The research provides baseline data that could be used for continued monitoring o f small stream for fisheries or water quality management. The data contribute to a large body of research on effects of forest harvest on small streams and expand the knowledge base. In particular, this work expands knowledge for a forest type that is commercially and ecologically important but has previously been little studied. Additionally, we have examined numerous biotic metrics to compare drift samples, and highlighted drawbacks and different sensitivities in some of these indicators through our data. This contributes to a body o f knowledge on use of biotic indicators for water quality monitoring. Each o f these contributions strengthens decision making by either providing scientific knowledge on which to base decisions, or by developing scientific tools for monitoring stream health. 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References Allan, J.D. 1995. Stream ecology. Structure and function of running waters. Chapman and Hall, New York. Allan, J.D., Wipfli, M.S., Caouette, J.P., Prussion, A., and Rodgers, J. 2003. Influence of streamside vegetation on inputs of terrestrial invertebrates to salmonid food webs. Canadian Journal o f Fisheries and Aquatic Sciences 60: 309-320. Barbour, M.T., Gerritsen, J., Snyder, B.D., and Stribling, J.B. 1999. Rapid bioassessment protocols for use in streams and wadeable rivers: Periphyton, benthic macroinvertebrates and fish, Second Edition. EPA 841-B-99-002. U.S. Environmental Protection Agency; Office of Water; Washington, D.C. Benke A.C., Huryn, A.D., Smock, L.A., and Wallace, J.B. 1999. Length-mass relationships for freshwater macroinvertebrates in North America with particular reference to the southeastern United States. Journal of the North American Benthological Society 18: 308- 343. Bennett, S. and Hewgill, K. 2002. A benthic invertebrate index o f biological integrity for streams in the Bulkley TSA. Prepared for Pacific Inland Resources, Ministry of Forests, and Ministry o f Water, Land and Air Protection, Smithers, BC. Bilby, R.E. and Bisson, P.A. 1992. Allochthonous versus autochthonous organic matter contributions to the trophic support o f fish populations in clear-cut and old-growth forested streams. Canadian Journal of Fisheries and Aquatic Sciences 49: 540-551. Bothwell, M.L., 2000. Effects of ultraviolet radiation on stream ecosystems and its links to forestry activities. Forest Renewal BC Research Program, Final Report. FRBC Ref# HQ96457-RE. Bothwell, M.L., Sherbot, D.M.J., Pollock, C.M. 1994. Ecosystem response to solar ultraviolet-B radiation: influence of trophic-level interactions. Science 265: 97-100. Bridcut, E. E. 2000. A study of terrestrial and aerial macroinvertebrates on river banks and their contribution to drifting fauna and salmonid diets in a Scottish catchment. Hydrobiologia 427: 83-100. 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Brittain, J.E. and Eikeland, T.J. 1988. Invertebrate Drift - A Review. Hydrobiologia 166: 77-93. Brisco, S., and Ziegler, S. 2004. Effects o f solar radiation on the utilization of dissolved organic matter (DOM) from two headwater streams. Aquatic Microbial Ecology 37: 197-208. Brooks, P.D., McKnight, D.M., and Bencala, K.E. 1999. The relationship between soil heterotrophic activity, soil dissolved organic carbon (DOC) leachate, and catchmentscale DOC export in headwater catchments. Water Resources Research 35: 18951902. Brooks, P.D., O ’Reilly, C.M., Diamond, S.A., Campbell, D.H., Knapp, R., Bradford, D., Com, P.S., Hossack, B, and Tonnessen, K. 2005. Spatial and temporal variability in the amount and source of dissolved organic carbon: implications for ultraviolet exposure in amphibian habitats. Ecosystems 8: 478-487. Brown, G.W., and Krygier, J.T. 1970. Effects o f Clear-cutting on stream temperature. Water Resources Research 6: 1133-1139. Burgherr, P. and Meyer, E.I. 1997. Regression analysis of linear body dimensions vs. dry mass in stream macroinvertebrates. Archiv fur Hydrobiologie 139: 101-112. Bumey, C.M. 1990. Seasonal and diel changes in dissolved and particulate organic material. In The biology of particles in aquatic systems. Edited by R.S. Wotton. CRC Press, Boston, pp. 83-116. Chamberlin, T.W., Harr, R.D., and Everest, F.H. 1991. Timber harvesting, silviculture, and watershed processes. American Fisheries Society Special Publication 19: 181206. Collier, K.J. 1995. Environmental factors affecting the taxonomic composition of aquatic m acroinvertebrate com m unities in low land waterways o f Northland, N e w Zealand. New Zealand Journal of Marine and Freshwater Research 29: 453-465. Cory, R.M, Green, S.A., and Pregitzer, K.S. 2004. Dissolved organic matter concentration and composition in the forests and streams of Olympic National Park, WA. Biogeochemistry 67: 269-288. 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Croft, C. 2003. Development and calibration o f a multimetric benthic invertebrate index of biological integrity: A biomonitoring study for assessing stream condition in the Morice and Lakes TSAs. Morice and Lakes Innovative Forest Practices Agreement No. 441.01, Telkwa, BC. Cuffney, T.F., and Wallace, J.B. 1988. Particulate organic matter export from three headwater streams: discrete versus continuous measurements. Canadian Journal of Fisheries and Aquatic Sciences 45: 2010-2016. Culp, J.M. and Davies, R.W. 1983. An assessment of the effects o f streambank clearcutting on macroinvertebrate communities in a managed watershed. Canadian Technical Report o f Fisheries and Aquatic Sciences No. 1208. Cummins, K.W. 1974. Structure and function of stream ecosystems. 1974. Bioscience 24:631-641. Cummins, K.W., Wilzbach, M.A., Gates, D.M., Perry, J.B., and Taliaferro, W.B. 1989. Shredders and riparian vegetation: leaf litter that falls into streams influences communities o f stream invertebrates. Bioscience 39: 24-30. Cummins, K.W., and Wilzbach, M.A. 2005. The inadequacy of the fish-bearing criterion for stream management. Aquatic Science 67: 486-491. DeLong, C. and Meidinger, D. 2003. Ecological variability of high elevation forests in central BC. The Forestry Chronicle 79: 259- 262. Eckhardt, B.W., and Moore, T.R. 1990. Controls on dissolved organic carbon concentrations in streams, Southern Quebec. Canadian Journal of Fisheries and Aquatic Sciences 47: 1537-1544. Edwards, E.D., and Huryn, A.D. 1995. Annual contribution of terrestrial invertebrates to a New Zealand trout stream. New Zealand Journal of Marine and Freshwater Research 29: 467-477. England, L.E., and Rosemond, A.D. 2004. Small reductions in forest cover weaken terrestrial-aquatic linkages in headwater streams. Freshwater Biology 49: 721-734. 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Feller, M. 1997. Influence of harvesting on the nutrient status of ESSFwc2 ecosystems. In Sicamous Creek Silvicultural Systems Project Workshop Proceedings, 24-25 April 1996, Kamloops, B.C. Edited by C. Hollstedt and A. Vyse. British Columbia Ministry of Forests, Victoria, BC. pp. 121-133. Fore, L.S., Karr, J.R., and Wisseman, R.W. 1996. Assessing invertebrate responses to human activities: evaluating alternative approaches. Journal of the North American Benthological Society 15: 212-231. Forest Renewal British Columbia (FRBC). 2000. Reconnaissance (1:20,000) fish and fish habitat inventory o f the Bone Creek Watershed. Prepared for Slocan Forest Products Ltd., Vavenby, BC. Forest Renewal British Columbia (FRBC). 2001. Reconnaissance (1:20,000) fish and fish habitat inventory o f the Winters Creek Watershed, Wolfcub Creek Watershed, Shatford Creek Watershed, selected sub-basins o f Penticton Creek Watershed, Campin Creek Watershed, and Two-John Creek Watershed. Prepared for Weyerhaeuser Company Ltd., Okanagan Falls, BC. Fuchs, S., Hinch, S., and Mellina, E. 2003. Effects o f streamside logging on stream macroinvertebrate communities and habitat in the sub-boreal forests o f British Columbia, Canada. Canadian Journal of Forest Research 33: 1408-1415. Giller, P.S. and Malmqvist, B. 1998. The biology o f streams and rivers. Oxford University Press, Oxford. Gore, J.A. 1996. Discharge measurements and streamflow analysis. In Methods in Stream Ecology. Edited by F.R. Hauer and G.A. Lamberti. Academic Press, Toronto, pp. 53-74. Gotelli, N.J., and Colwell, R.K. 2001. Quantifying biodiversity: procedures and pitfalls in the measurement and comparison o f species richness. Ecology Letters 4: 379-391. Greenhalgh, C.A., and Heise, B.A. 2004. Effects o f logging on export of organic matter from headwater streams. Unpublished report prepared for the British Columbia Forest Research Program. 30 pp. 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Gregory, S.V., Lamberti, G.A., Erman, D.C., Koski, K.V., Murphy, M.L. and Sedell, J.R. 1987. Influence o f forest practices on aquatic production. In Streamside management: forestry and fishery interactions, February 12-14 1986, Seattle Washington. Edited by E.O. Salo, T.W. Cundy. University of Washington, Institute of Forest Resources, Seattle, Washington, pp. 233-255. Gurtz, M.E., and Wallace, J.B. 1984. Substrate-mediated response o f stream invertebrates to disturbance. Ecology 65: 1556-1569. Hall, R.O., and Meyer, J.L. 1998. The trophic significance o f bacteria in a detritus-based stream food web. Ecology 79: 1995-2012. Hannam, K.D., and Prescott, C.E. 2003. Soluble organic nitrogen in forests and adjacent clearcuts in British Columbia, Canada. Canadian Journal of Forest Research 33: 1709-1718. Hawkins, C.P., Murphy, M.L. and Anderson, N.H. 1982. Effects of canopy, substrate composition, and gradient on the structure o f macroinvertebrate communities in cascade range streams o f Oregon. Ecology 63: 1840-1856. Heise, B.A. 2001. Sicamous Creek silvicultural systems project: effects o f timber harvesting and road construction on stream invertebrates in the Sicamous Creek watershed. Forest Renewal BC Research Program, Final Report. FRBC # TO96105-RE. Hernandez, O., Merritt, R.W. and Wipfli, M.S. 2005. Benthic invertebrate community structure is influenced by forest succession after clearcut logging in southeastern Alaska. Hydrobiologia 533: 45-59. Hetrick, N.J., Brusven, M.A., Meehan, W.R., and Bjomn, T.C. 1998. Effects of canopy removal on invertebrates and diet of juvenile coho salmon in a small stream in southeast Alaska. Transactions of the American Fisheries Society 127: 876-888. Hicks, B.J., Hall, J.D., Bisson, P.A. and Sedell, J.R. 1991. Responses o f salmonids to habitat changes. American Fisheries Society Special Publication 19: 483-518. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hilsenhoff, W.L. 1982. Using a biotic index to evaluate water quality in streams. Technical Bulletin No. 132. Wisconsin Department of Natural Resources, Madison, W I. Hood, E. McKnight, D.M., and Williams M.W. 2003. Sources and chemical quality of dissolved organic carbon (DOC) across an alpine/subalpine ecotone, Green Lakes Valley, Colorado Front Range, United States. Water Resources Research 39: 1188. Hood, E., Williams, M.W., and McKnight, D.M. 2005. Sources of dissolved organic matter (DOM) in a Rocky Mountain stream using chemical fractionation and stable isotopes. Biogeochemistry 74: 231-255. Hood, E., Gooseff, M.N. and Johnson, S.L. 2006. Changes in the character o f stream water dissolved organic carbon during flushing in three smalll watersheds, Oregon. Journal of Geophysical Research 111, GO 1007, doi: 10.1029/2005JG000082. Hosmer, D., and Lemeshow, S. 2000. Applied logistic regression. Second edition. John Wiley and Sons Inc. New York. Hurlbert, S.H. 1984. Pseudoreplication and the design o f ecological field experiments. Ecological Monographs 54: 187 - 211. Hynes, H.B.N. 1970. The ecology of running waters. The Blackburn Press, New Jersey. Jackson, C.R., Sturm, C.A., and Ward, J.M. 2001. Timber harvest impacts on small headwater stream channels in the coast ranges o f Washington. Journal of the American Water Resources Association 37: 1533-1549. Karr, J.R. and Chu, E.W. 1999. Restoring Life in Running Waters: Better Biological Monitoring. Island Press, Washington, D.C. Kawaguchi, Y. and Nakano, S. 2001. Contribution o f terrestrial invertebrates to the annual resource budget for salmonids in forest and grassland reaches of a headwater stream. Freshwater Biology 46: 303-316. Kelly, D.J., and Clare, J.J. 2001. Attenuation o f solar ultraviolet radiation by dissolved organic matter alters benthic colonization patterns in streams. Journal o f the North American Benthological Society 20: 96-108. 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kelly, D.J., and Bothwell, M.L. 2002. Avoidance of solar ultraviolet radiation by juvenile coho salmon (Onchorhynchus kisutch). Canadian Journal o f Fisheries and Aquatic Sciences 59: 474-482. Kelly, D.J., Bothwell, M.L., and Schindler, D.W. 2003. Effects o f solar ultraviolet radiation on stream benthic communities: an intersite comparison. Ecology 84: 27242740. Kiffney, P.M., Richardson, J.S., and Feller, M.C. 2000. Fluvial and epilithic organic matter dynamics in headwater streams of southwestern British Columbia, Canada. Archiv fur Hydrobiologie 149: 109-129. Kiffney, P.M, Richardson, J.S., and Bull, J.P. 2003. Responses o f periphyton and insect consumers to experimental manipulation o f riparian buffer width along headwater streams in forested landscapes. Journal o f Applied Ecology 40: 1060-1076. Krebs, C.J. 1999. Ecological Methodology. Harper Collins Publishers, New York. Kreutzweiser, D.P. and Capell, S.S. 2003. Benthic microbial utilization of differential dissolved organic matter sources in a forest headwater stream. Canadian Journal of Forest Research 33:1444-1451. Kreutzweiser, D.P., Capell, S.S., and Beall, F.D. 2004. Effects o f selective forest harvesting on organic matter inputs and accumulation in headwater streams. Northern Journal o f American Foresters 21: 19-30. Kohler, S.L. 1984. Identification of stream drift mechanisms: An experimental and observational approach. Ecology 66: 1383-1392. Lloyd, D., Angrove, K., Hope, G., and C. Thompson. 1990. A guide to site identification and interpretation for the Kamloops Forest Region. Research Branch, Ministry of Forests, Victoria, BC. Lemmon, P.E. 1956. A spherical densitometer for estimating forest overstorey density. Forest Science 2: 314-320. Lock, M.A. 1990. The dynamics o f dissolved and particulate organic material over the substratum o f water bodies. In The biology o f particles in aquatic systems. Edited by R.S. Wotton. CRC Press, Boston, pp. 117-144. 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Macdonald, J.S., Beaudry, P.G., Maclsaac, E.A., and Herunter, H.E. 2003a. The effects o f forest harvesting and best management practices on streamflow and suspended sediment concentrations during snowmelt in headwater streams in sub-boreal forests o f British Columbia, Canada. Canadian Journal o f Forest Research 33: 1397-1407. Macdonald, J.S., Maclsaac, E.A., and Herunter, H.E. 2003b. The effect o f variable retention riparian buffer zones on water temperatures in small headwater streams in sub-boreal forest ecosystems of British Columbia. Canadian Journal o f Forest Research 33: 1371-1382. McCann, K., Hastings, A., and Huxel, G.R. 1998. Weak trophic interactions and the balance of nature. Nature 395: 794-798. Merritt, R.W. and Cummins, K.W. {Editors). 1996. An introduction to the aquatic insects of North America. Third edition. Kendall/Hunt Publishing Co., Dubuque, Iowa. Moore, R.D., and Wondzell, S.M. 2005. Physical hydrology and the effects o f forest harvesting in the Pacific Northwest: a review. Journal of the American Water Resources Association 41: 763-784. Moore, T.R. 1989. Dynamics of dissolved organic carbon in forested and disturbed catchments, Westland, New Zealand. 1. Maimai. Water Resources Research 25: 1321-1330. Moore, T.R., and Jackson, R.J. 1989. Dynamics of dissolved organic carbon in forested and disturbed catchments, Westland, New Zealand. 2. Larry River. Water Resources Research 25: 1331-1339. Murphy, M.L. and Hall, J.D. 1981. Varied effects of clear-cut logging on predators and their habitat in small streams o f the Cascade Mountains, Oregon. Canadian Journal o f Fisheries and Aquatic Sciences 38: 137-145. Murphy, M.L., and Meehan, W.R. 1991. Stream ecosystems. American Fisheries Society Special Publication 19: 17-46. 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Nakano, S., Kawaguchi, Y., Taniguchi, Y., Miyasaka, H., Shibata, Y., Urabe, H., and Kuhara, N. 1999. Selective foraging on terrestrial invertebrates by rainbow trout in a forested headwater stream in northern Japan. Ecological Research 14: 351-360. Newbold, J.D., Erman, D.C. and Roby, K.B. 1980. Effects o f logging on macroinvertebrates in streams with and without buffer strips. Canadian Journal of Fisheries and Aquatic Sciences 37: 1076-1085. Noel, D.S., Martin, C.W., and Federer, C.A. 1986. Effects of forest clearcutting in New England on stream macroinvertebrates and periphyton. Environmental Management 10: 661-670. Parish, R., Coupe, R. and Lloyd, D. {Editors). 1996. Plants of southern interior British Columbia. B.C. Ministry of Forests and Lone Pine Publishing, Vancouver, B.C. Peters, G.T., Webster, J.R., and Benfield, E.F. 1987. Microbial activity associated with seston in headwater streams: effects o f nitrogen, phosphorus and temperature. Freshwater Biology 18: 405-413. Province of B.C. 1995. Riparian management area guidebook. Ministry of Forests and Ministry of the Environment. Province o f British Columbia, Victoria, B.C. Quinn, J.M., Boothroyd, I.K.G., and Smith, B.J. 2004. Riparian buffers mitigate effects o f pine plantation logging on New Zealand streams 2. Invertebrate communities. Forest Ecology and Management 191: 129-146. Resh, V.H. and Jackson, J.K. 1993. Rapid assessment approaches to biomonitoring using benthic macroinvertebrates. In Freshwater Biomonitoring and Benthic Macroinvertebrates. Edited by D.M. Rosenberg, and V.H. Resh. Chapman and Hall, New York. pp. 195-233. Rogers, L.E., Buschbom, R.L., and Watson, C.R. 1977. Length-weight relationships of shrub-steppe invertebrates. Annals o f the Entomological Society of America 70: 5153. Rosenberg, D.M., and Resh, V.H. {Editors). 1993. Freshwater Biomonitoring and Benthic Macroinvertebrates. Chapman and Hall, New York. 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SalmonWeb. 2001. Community based monitoring for biological integrity o f streams. http://www.cbr.washinton.edu/salmonweb Sample, B.E., Cooper, R.J., Greer, R.D., and Whitmore, R.C. 1993. Estimation of insect biomass by length and width. American Midland Naturalist 129: 234-240. Smock, L.A. 1980. Relationships between body size and biomass o f aquatic insects. Freshwater Biology 10: 375-383. Smock, L.A. 1996. Macroinvertebrate movements: drift, colonization, and emergence. In Methods in Stream Ecology. Edited by F.R. Hauer and G.A. Lamberti. Academic Press, Toronto, pp. 371-390. Siler, E.R., Wallace, J.B., and Eggert, S.L. 2001. Long-term effects o f resource limitation on stream invertebrate drift. Canadian Journal of Fisheries and Aquatic Sciences 58: 1624-1637. Silsbee, D.G., and Larson, G.L. 1983. A comparison of streams in logged and unlogged areas o f Great Smoky Mountains National Park. Hydrobiologia 102: 99-111. Stelzer, R.S., Heffeman, J., and Likens, G.E. 2003. The influence o f dissolved nutrients and particulate organic matter quality on microbial respiration and biomass in a forest stream. Freshwater Biology 48: 1925-1937. Stewart, K.W. and Stark, B.L. 2002. Nymphs of North American Stonefly Genera (Plecoptera). Second edition. The Caddis Press. Columbus, Ohio. Stone, M.K. and Wallace, J.B. 1998. Long-term recovery of a mountain stream from clear-cut logging: the effects of forest succession on benthic invertebrate community structure. Freshwater Biology 39: 151-169. Suberkropp, K.F. 1998. Microorganisms and organic matter decomposition. In River Ecology and Management. Edited by R.J. Naiman and R.E. Bilby. Springer-Verlag, New York, NY. pp. 120-143. Tabachnick, B.G., and Fidell, L.S. 1989. Using multivariate statistics. Second edition. HarperCollins, New York. 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tate, C.M., and Meyer, J.L. 1983. The influence of hydrologic conditions and successional state on dissolved organic carbon export from forested watersheds. Ecology 64: 25-32. Thomas, J.D. 1997. The role of dissolved organic matter, particularly free amino acids and humic substances, in freshwater. Freshwater Biology 38:1-36. Toews, D.A.A., and Moore, M.K. 1982. The effects of streamside logging on large organic debris in Carnation Creek. Land Management Report 11. Ministry of Forests, Victoria, BC. Tranvik, L.J. and Bertilsson, S. 2001. Contrasting effects o f solar UV radiation on dissolved organic sources for bacterial growth. Ecology Letters 4: 458-463. Troendle, C.A., and King, R.M. 1985. The effect of timber harvest on the Fool Creek Watershed, 30 years later. Water Resources Research 21: 1915-1922. Underwood, A.J. 1994. On beyond BACI: sampling designs that might reliably detect environmental disturbances. Ecological Applications 4: 3-15. Vannote, R.L., Minshall, G.W., Cummins, K.W., Sedell, J.R., and Cushing, C.E. 1980. The river continuum concept. Canadian Journal o f Fisheries and Aquatic Sciences 37:130-137. Vannote, R.L., and Sweeney, B.W. 1980. Geographic analysis o f thermal equilibria: a conceptual model for evaluating the effect of natural and modified thermal regimes on aquatic insect communities. The American Naturalist 115: 667-695. Wallace, J.B., and Gurtz, M.E. 1986. Response of Baetis mayflies (Ephemeroptera) to catchment logging. The American Midland Naturalist 115: 25-41. Wallace, J.B., Eggert, S.L., Meyer, J.L., and Webster, J.R. 1997. Multiple trophic levels of a forest stream linked to terrestrial litter inputs. Science 277: 102-105. Wallace, J.B. and Grubaugh, J.W. 1996. Transport and storage o f FPOM. In Methods in Stream Ecology. Edited by F.R. Hauer and G.A. Lamberti. Academic Press, Toronto, pp. 191-216. Waters, T.F. 1965. Interpretation o f invertebrate drift in streams. Ecology 46: 327-334. 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Webster, J.R., Gurtz, M.E., Hains, J.J., Meyer, J.L., Swank, W.T., Waide, J.B., and Wallace, J.B. 1983. Stability of stream ecosystems. In Stream Ecology. Edited by J.R. Barnes and G.W. Minshall. Plenum Press, NewYork. pp. 355-395. Webster, J.R., Golladay, S.W., Benfield, E.F., D'Angelo, D.J., and Peters, G.T. 1990. Effects of forest disturbance on particulate organic matter budgets o f small streams. Journal o f the North American Benthological Society 9: 120-140. Weishaar, J.L., Aiken, G.R., Bergamaschi B.A., Fram, M.S., Fugii, R., and Mopper, K. 2003. Evaluation of specific ultraviolet absorbance as an indicator o f the chemical composition and reactivity of dissolved organic carbon. Environmental Science and Technology 37: 4702-4708. Wiggins, G.B. 1996. Larvae of the North American Caddisfly Genera (Trichoptera). Second edition. University o f Toronto Press, Toronto. Williamson, C.E., Stemberger, R.S., Morris, D.P, Frost, T.M, and Paulsen, S.G. 1996. Ultraviolet radiation in North American lakes: attenuation estimates from DOC measurements and implications for plankton communities. Limnology and Oceanography 41:1024-1034. Wilzbach, M.A., Cummins, K.W., and Hall, J.D. 1986. Influence o f habitat manipulations on interactions between cutthroat trout and invertebrate drift. Ecology 67:898-911. Winkler, R.D., Spittlehouse, D.L., and Golding, D.L. 2005. Measured differences in snow accumulation and melt along clearcut, juvenile, and mature forests in southern British Columbia. Hydrological Processess 19: 51-62. Winterboum, M.J., and Townsend, C.R. 1991. Streams and rivers: one-way flow systems. In Fundamentals of aquatic ecology. Edited by R.S.K. Barnes and K.H. Mann. Blackwell, Oxford, pp. 230-242. Wipfli, M.S., and Gregovich, D.P. 2002. Export of invertebrate and detritus from fishless headwater stream in southeastern Alaska: implications for downstream salmonid production. Freshwater Biology 47: 957-969. 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wipfli, M.S., and Musslewhite, J. 2004. Density o f red alder (Alnus rubra) in headwaters influences invertebrate and detritus subsidies to downstream fish habitats in Alaska. Hydrobiologia 520: 153-163. Woodward G., and Hildrew, A.G. 2002. Food web structure in riverine landscapes. Freshwater Biology 47: 777-798. Wotton, R.S. 1990. Particulate and dissolved organic material as food. In The biology of particles in aquatic systems. Edited by R.S. Wotton. CRC Press, Boston, pp. 213261. Young, K.A. 2000. Riparian zone management in the Pacific Northwest: Who's cutting what. Environmental Management 26: 131-144. Zar, J.H. 1984. Biostatistical analysis. Prentice Hall, Toronto. Ziegler, S.E., and Brisco, S.L. 2004. Relationships between the isotopic composition of dissolved organic carbon and its bioavailability in contrasting Ozark streams. Hydrobiologia 513: 153-169. 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. R eproduced with perm issio n Appendix of the copyright o w n e r . Table A l. List of taxa and their designations as aquatic (A), terrestrial (T), clinger (CLG=Y); in functional feeding groups predator (PR), collector-gatherer (CG), collector-filterer (CF), scraper (SC), shredder (SH); their Hilsenhoff biotic index value (HBI); and their status as tolerant (TOL=T), and long-lived (Long-lived=Y). Biomass formulas estimate dry mass (DM) in mg using body length (L) measurements Fu rth er reprod uctio n prohibited without p e r m i s s i o n . Phylum Class Order Family G enus Arthopoda A rthopoda Arthopoda Arthopoda Arthopoda A rthopoda Arthopoda A rthopoda Arthopoda Arthopoda Arthopoda A rthopoda Arthopoda Arthopoda Arthopoda Arthopoda Arthopoda Arthopoda Arthopoda Arthopoda Arthopoda Arthopoda A rthopoda Arthopoda A rthopoda A rthopoda Arthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda Arthopoda A rthopoda Entoqnatha Arachnida Arachnida Arachnida Arachnida Arachnida A rachnida Arachnida insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Collembola A raneae Acarina Acarina Acarina Acarina Acarina A ran eae C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera C oleoptera Diptera Diptera Diptera Diptera Diptera Diptera Diptera Diptera Diptera Diptera Diptera Diptera Diptera Diptera n.s. Dolomedes Pisauridae "eqq-shape" n.s. n.s. “e lo n g a te ” “brow n/hard” n.s. n.s. "round” Styqothrombidiidae “elonqate/soft” n.s. A m phizoidae, ad C arabidae C hrysom elidae, ad C hrysom elidae Curculionidae, ad Dytiscidae, ad Dytiscidae Elmidae, ad H isteridae, ad H ydraenidae Hydrophilidae, ad Hydrophilidae Staphylinidae, ad Staphylinidae n.s., ad n.s. n.s., ad Athericidae C eratopoqonidae C hironom idae Culicidae Dixidae Em pididae n.s. P sychodidae n.s., pu Sciom yzidae Simuliidae Stratiom yidae Syrphidae Aqu. LongBiom ass formula CLG PR CG CF s c SH HBI TOL Terr. lived T A A A T A A T A T T T T A A A T T A A T T T T T A A A A A A A A A A A A 100 Y Y Y Y Y Y Y Y Y 107 n.d. n.d. 5 5 5 5 5 n.d. 1 n.d. 100 n.d. 100 n.d. 100 5 100 5 100 5 4 100 n.d. n.d. 100 50 50 5 5 100 100 n.d. 100 n.d. n.d. n.d. n.d. 100 2 100 6 6 50 50 8 50 50 100 2 100 6 n.d. 10 100 50 50 6 n.d. 100 100 6 100 8 10 100 100 100 100 100 100 100 100 100 100 n.d. n.d. n.d. n.d. n.d. n.d. T T n.d. n.d. n.d. n.d. n.d. n.d. n.d. T n.d. Y n.d. Y n.d. n.d. n.d. Y Y Y n.d. n.d. Y Y n.d. n.d. n.d. n.d. n.d. T n.d. n.d. n.d. n.d. T T DM =exp(-3.628)*power(L,2.494) DM =exp(-3.106)*(power(L,2.929)) DM =exp(3.682)*(power(L,2.761 ))/1000 DM =exp(3.682)*(power(L,2.761))/1000 D M =exp(3.682)*(power(L,2.761 ))/1000 DM =exp(3.682)*(power(L,2.761 ))/1000 DM=exp{3.682)*(power(L,2.761 ))/1000 D M =exp(-3.106)*(power(L,2.929)) DM=exp(-3.460)*(power(L,2.790) D M =0.0077*power(L,2.910) D M =exp(-2.427)*power(L,2.171) D M =.0392*(power(L,3.111)) D M =exp(-3.460)*(power(L,2.790) DM=.0618*power(L,2.502) DM=.0077*(power(L,2.910) DM =exp(-3.460)*(power(L,2.790) DM =exp(-3.460)*(power(L,2.790) D M =0.0077*power(L,2.910) DM =exp(-3.460)*(power(L,2.790) DM =.0077*(power(L,2.910) DM =exp(-3.460)*(power(L,2.790) DM =0.0077*power(L,2.910) DM =exp(-3.460)*(power(L,2.790) D M =0.0077*power(L,2.910) DM =exp(-3.293)*power(L,2.366) DM=.0040*power(L,2.586) DM=.0025*power(L,2.469) DM=.0018*power(L,2.617) DM=.0025*power(L,2.692) DM=.0025*power(L,2.692) DM=.0054*power(L,2.546) DM=.0025*power(L,2.692) DM=.0025*power(L,2.692) DM=.0025*power(L,2.692) DM=.0025*power(L,2.692) DM =.002*power(L,3.011) DM=.0025*power(L,2.692) DM=.0025*power(L,2.692) Ref. Note A B B B B B B B B B A D B C D B B D B D B B B D B D D D D D D D D D D D D D 4 4 4 4 4 1 5 R eproduced with perm issio n Family G enus of the c opyright o w n e r . F urther rep ro du ction prohibited without p e r m i s s i o n . Phylum C lass Order A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda A rthopoda Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta tn secta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Insecta Diptera T h aum aleidae Diptera Tipulidae E p h em eropter n.s., ad Ameletus E phem eropter A m eletidae E phem eropter B aetidae Baetis E phem eropter B aetidae n.s. E phem eropter E phem erellidae Attenella E phem eropter E phem erellidae Drunella E phem eropter E phem erellidae n.s. E phem eropter H eptaqeniidae Cinyqmula E phem eropter H eptaqeniidae Epeorus (Iron) E phem eropter H eptaqeniidae Epeorus (Ironopsis) E phem eropter H eptaqeniidae n.s. Rhithrogena E phem eropter H eptageniidae E phem eropter n s.s n.s. H em iptera H em iptera A phidae H em iptera H eteroptera (suborder) H ym enoptera n.s., ad L epidoptera n.s., ad L epidoptera n.s. A nisoptera (suborder) O d o n ata Z ygoptera (suborder) O d o n ata o ther n.s., ad Pleco p tera C hloroperlidae, ad P leco p tera C hloroperlidae Haploperla P leco p tera P leco p tera C hloroperlidae n.s. Suwallia Pleco p tera Chloroperlidae Pleco p tera C hloroperlidae Sweltsa n.s. Pleco p tera Despaxia P leco p tera Leuctridae n.s., ad P leco p tera N em ouridae n.s. P leco p tera N em ouridae Podmosta P leco p tera N em ouridae P leco p tera N em ouridae Zapada Peltoperlidae n.s. P leco p tera Soliperia P leco p tera Peltoperlidae Yoraperla P leco p tera Peltoperlidae n.s., s h re d d e r P leco p tera P leco p tera Perlodidae n.s. Setvena P leco p tera Perlodidae Taeniopterygidae n.s. P leco p tera LongAqu. B iom ass formula CLG PR CG CF s c SH HBI TOL lived Terr. 100 Y A A T A A A A A A A A A A A A A T T T T T A A T T A A A A A A T A A A A A A A A A A 50 100 100 100 100 100 100 Y Y Y Y Y Y Y Y 100 100 100 100 100 100 Y Y Y Y Y Y Y Y Y Y 108 100 100 100 100 100 100 11 3 n.d. 0 5 4 2 0 1 100 4 100 0 100 0 100 4 100 0 11 8 n.d. n.d. n.d. 100 5 100 5 11 8 n.d. n.d. 0 1 0 1 11 100 0 n.d. 100 2 100 2 100 2 100 1 100 1 100 1 100 11 2 2 2 I 50 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Y n.d. n.d. I n.d. I I I I n.d. n.d. n.d. Y DM=.0025*power(L,2.692) DM=.0029*power(L,2.681) D M =exp(-3.628)*power(L,2.494) DM =0.0077*power(L,2.588) DM =0.0033*power(L,3.196) DM =0.0053*power(L,2.875) D M =0.0103*power(L,2.676) D M =0.0019*power(L,3.46) DM =0.0103*power(L,2.676) DM =0.0108*power(L,2.754) DM=0.0056*power(L,2.926) DM =0.0056*power(L,2.926) D M =0.0108*power(L,2.754) DM=0.0108*power(L,2.754) DM=0.0071 *power(L,2.832) D M =exp(-3.308)*power(L,2.696) D M =exp(-3.308)*power(L,2.696) DM=exp(-2-998)*power(L,2.270) D M =exp(-3.871 )*power(L,2.407) D M =pow er(exp,-5.036)*pow er(L,3.122) D M =pow er(exp,-5.909)*p°w er(L,2.959) D M =0.0078*power(L,2.792) DM =0.0078*power(L,2.792) D M =pow er(exp,-4.357)*pow er(L,2.539) DM g>ow er(exp,-4.357)*pow er(L,2.539) DM=.0065*power(L,2.724) DM=.0065*power(L,2.724) DM=.0065*power(L,2.724) D M =.0065*power(L,2.724) DM =0.0094*power(L,2.754) D M =.0028*power(L,2.719) D M =pow er(exp,-4.357)*pow er(L,2.539) D M=.0056*power(L,2.762) D M=.0056*power(L,2.762) DM=.0056‘ power(L,2.762) DM =0.0170*power(L,2.737) DM=0.0170*power(L,2.737) DM=0.0170*power(L,2.737) DM=0.0094*power(L,2.754) DM=0.0196*power(L,2.742) DM=0.0196*power(L,2.742) D M =0.0094*power(L,2.754) Ref. Note D D A D C D D D D D D D D D D B B B B A A D D A A D D D D D D A D D D D D D D D D D 4 2 2 2 2 3 3 3 3 R eproduced with perm issio n Phylum C lass of the copyright o w n e r . Fu rth er reprod uctio n prohibited A rthopoda Insecta Insecta A rthopoda Insecta A rthopoda A rthopoda Insecta A rthopoda Insecta Insecta A rthopoda A rthopoda Insecta A rthopoda Insecta A rthopoda Insecta A rthopoda Insecta A rthopoda Insecta A rthopoda Insecta A rthopoda Insecta Insecta A rthopoda A rthopoda Insecta Insecta A rthopoda A rthopoda Insecta A rthopoda Insecta A rthopoda Insecta A rthopoda In secta A rthopoda Insecta A rthopoda Insecta A rthopoda Insecta Insecta A rthopoda Insecta A rthopoda A rthopoda Insecta A rthopoda Insecta M ollusca G astropoda N em atoda N em atom orpha A nnelida Clitellata Platyhelm inthes Turbellaria O rder Family T h y san o p tera n.s., ad T h y san o p tera n.s. T richoptera A pataniidae T richoptera A pataniidae T richoptera B rachycentridae T richoptera G lo sso so m a tid a e T richoptera G lo sso so m a tid a e T richoptera H ydropsychidae T richoptera H ydropsychidae T richoptera Lim nephilidae T richoptera Lim nephilidae T richoptera Lim nephilidae T richoptera Lim nephilidae T richoptera Lim nephilidae Lim nephilidae T richoptera T richoptera Lim nephilidae T richoptera Lim nephilidae Lim nephilidae T richoptera T richoptera Philopotam idae T richoptera n.s., pu T richoptera R hyacophilidae T richoptera R hyacophilidae T richoptera R hyacophilidae T richoptera U enoidae T richoptera U enoidae n.s. T richoptera n .s. n.s., terrestrial G enus Allomyia Moselyana Micrasema Glossosoma n.s. n.s. Parapsyche Chyranda Cryptochia Desmona Eocosmoecus Homophylax n.s. Onocosmoecus Pyschogtypha Spagnophylax Wormaldia Himalopysche n.s. Rhyacophila n.s. Neothremma Aqu. LongCLG PR CG CF S c lsH HBI TOL B iom ass formula Terr. lived T T A A A A A A A A A A A A A A A A A A A A A A A A T A A A A A Y 50 100 50 Y Y Y Y Y 50 100 100 100 100 50 Y Y 100 50 100 100 100 100 50 100 50 Y Y Y Y Y Y 50 100 100 100 100 50 50 50 50 100 100 100 n.d. n.d. 1 1 1 1 0 4 4 1 0 1 0 0 4 1 0 4 3 n.d. 0 0 0 0 0 n.d. n.d. 7 5 11 8 4 n.d. n.d. n.d. n.d. I I I I I I I n.d. I I n.d. n.d. Y Y Y Y n.d. n.d. n.d. Ref. Note without p e r m i s s i o n . D M =exp(-3.628)*power(L,2.494) D M =exp(-3.628)*power(L,2.494) DM =0.0056*power(L,2.839) DM =0.0056*power(L,2.839) DM =0.0083*pow er(L,2.818) DM =0.0082*power(L,2.958) DM =0.0082*power(L,2.958) DM =0.0046*power(L,2.926) DM=0.0046*power(L,2.926) DM=0.004*power(L,2.933) DM=0.004*power(L,2.933) DM =0.004*power(L,2.933) DM=0.004*power(L,2.933) DM=0.004*power(L,2.933) DM=0.004*power(L,2.933) DM=0.004*power(L,2.933) DM=0.004*power(L,2.933) DM=0.004*power(L,2.933) D M =0.005*pow er(L,2.511) DM =0.0056*power(L,2.839) DM =0.0099*power(L,2.480) DM =0.0099*power(L,2.480) DM =0.0099*power(L,2.480) DM=0.0056*power(L,2.839) DM =0.0056*power(L,2.839) DM =0.0056*power(L,2 .839) D M =exp(-3.628)*power(L,2.494) afdm =.0203*pow er(L,2.521) DM =0.2679 D M =0.2679 DM=0.025 DM =0.0082*pow er(L,2.168) A A D D D D D D D D D D D D D D D D D D D D D D D D A D E E E D DM= 0.03300 DM= 0.03300 D M =0.02644 E E E Subphylum: Crustacea A rthropoda A rthropoda A rthropoda Branchiopoda Maxillopoda C o p e p o d a (S u b cla ss) O stracoda 1 A A A 100 100 100 8 8 8 R eferen ces A. S am p le e t al. 1993, B. R o g ers e t al. 1977; C. Sm ock 1980; D. B enke e t al. 1999; E. m e a n DM of a t le a s t 10 individuals ta k e n from drift s a m p le s by C.A. M ackay in 2004. N otes 1. U sed desig n atio n s for S taphylinidae adult. 2. G enerally clingers, Merritt & C um m ins (1996). 3. Spraw lers/clingers, Merritt a n d C um m ins (1996). 4. T h e s e ta x a a re p arasites an d a re included a s p redators. 5. Diptera p u p a e d esignations a re b a se d on C hironom idae, a s majority o f p u p a e w ere s u sp e c te d bu t unconfirm ed C hironom idae. 6. T h e s e ta x a a re univoltine/semivoltine (Merritt a n d C um m ins 1996) an d a re th e re fo re not co n sid ered long-lived. 109 6 6 6 6