E F F E C T OF F O R E S T HARVESTING ON B I O L O G I C A L PROCESSES IN H E A D W A T E R STREAMS WITHIN T H E F L U M E C R E E K E X P E R I M E N T A L W A T E R S H E D , BRITISH C O L U M B I A by JENNIFER ANN H I E B L E R - C H A R I A R S E B.Sc , Willamette University, Salem, Oregon, U.S.A. 1999 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES Resource Management and Environmental Studies We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A April 2003 © Jennifer A . Hiebler-Chariarse, 2003 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for refernce and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Resource Management and Environmental Studies The University of British Columbia Vancouver, Canada o ABSTRACT Effects of forest harvesting on periphyton biomass (measured as Ash Free Dry Mass (AFDM) and Chlorophyll a), benthic organic matter (measured as A F D M ) and invertebrate density and diversity were studied in coastal headwater streams in southwestern British Columbia. These aquatic ecosystem variables were also evaluated along the longitudinal gradient of these streams to examine longitudinal patterns. In addition, photosynthetically active solar radiation (PAR), stream water temperatures and chemical concentrations of streamwater were measured and evaluated in order to determine their role in regulating the aforementioned variables. Six headwater streams (1 s t and 2 n d order), three undisturbed and three harvested (3 years post-harvest), were sampled over a six-week period during the early summer (May 31 - July 7, 2002). Each stream had nine study reaches (covering a total stream length of 500-700 meters), with harvested streams having three reaches above the harvested area, three reaches within the harvested area and three below - providing for comparisons between forested and harvested stream reaches within streams. Periphyton biomass sampling occurred once a week over the six-week study, while benthic macro invertebrates were sampled over four days during week 3 of the study. Physical and chemical data were also collected once a week for the six-week study period. No significant longitudinal trends were found within the undisturbed streams for any of the studied variables. However, along the harvested streams, Chlorophyll a and stream temperature increased along the downstream longitudinal gradient as the stream flowed through the harvested area and subsequently decreased as the streams flowed back into forested reaches. In addition, incident solar radiation decreased (however not statistically significant) as the streams flowed further into the forested area below the harvested boundary. Incident solar radiation and stream water temperatures were greater in the harvested reaches than in forested reaches of the harvested streams. Periphyton biomass levels were greater within harvested reaches than in forested reaches of the headwater streams. Incident solar radiation was considered to be a primary factor controlling periphyton biomass in these headwater streams, as periphyton biomass increased with increased incident solar radiation. Increased invertebrate density was positively related to periphyton biomass, suggesting bottom-up controls in the aquatic food web. While richness of aquatic invertebrate orders was greater in forested reaches below the harvested area, overall invertebrate diversity was not altered by previous harvesting activities. No differences were seen in benthic organic matter biomass nor were there significant relationships between nutrient concentrations with the aforementioned research variables during the study period. i i TABLE OF CONTENTS Abstract . ii List of Tables v List of Figures vi Acknowledgements vii CHAPTER 1 - INTRODUCTION 1 1.1 General Overview 1 1.2 Stream Structure and Processes 2 1.3 The Longitudinal Gradient 3 1.4 Impact of Forest Harvesting on Stream Processes 4 1.5 Objectives 6 CHAPTER 2 - RESEARCH DESIGN 7 2.1 Study Area Description 7 2.2 Experimental Design 9 CHAPTER 3 - PHYSICAL/CHEMICAL PROPERTIES 12 3.1 Introduction 12 3.2 Methods 20 3.3 Results 23 3.4 Discussion 39 3.5 Conclusion 45 CHAPTER 4 - PERIPHYTON 48 3.1 Introduction 48 3.2 Methods 52 3.3 Results 55 3.4 Discussion 62 3.5 Conclusion 66 CHAPTER 5 - INVERTEBRATES 68 3.1 Introduction 68 3.2 Methods 71 3.3 Results 75 3.4 Discussion 90 3.5 Conclusion 97 CHAPTER 6 - CONCLUSIONS 98 Cited References 103 i i i Appendix 1 - ANOVA's (undisturbed streams - dissolved ion concentrations)... 110 Appendix 2 - ANOVA's (harvested streams - dissolved ion concentrations) 112 Appendix 3 - Chemical concentrations (figures for Al Na, Si0 2 and S04) 114 Appendix 4 - ANOVA's (streamwater temperature) 116 Appendix 5 - ANOVA's (periphyton measures) 117 Appendix 6 - ANOVA's (undisturbed stream - invertebrate density) 118 Appendix 7 - ANOVA's (undisturbed stream - invertebrate diversity/O.M) 119 Appendix 8 - ANOVA's (harvested streams - invertebrate density) 120 Appendix 9 - ANOVA's (harvested streams - invertebrate richness) 121 Appendix 10 - ANOVA's (harvested streams - invertebrate diversity) 122 Appendix 11 - ANOVA (harvested streams - organic matter) 123 iv LIST OF T A B L E S Page# Table 2.1 Characteristics of headwater study streams. 11 Table 3.1 A N O V A table (split-plot design) for undisturbed streams. 22 Table 3.2 A N O V A table (split-plot nested design) for harvested streams. 23 Table 3.3 Average stream water chemistry concentrations for undisturbed streams. 25 Table 3.4 Average stream water chemistry concentrations for harvested streams. 26 Table 4.1 Average A F D M and Chlorophyll a values in the harvested streams. 57 Table 4.2 Regression data for periphyton biomass. 62 Table 4.3 Dissolved nutrient concentrations (N and P) in harvested streams. 62 Table 5.1 Invertebrate densities within the undisturbed stream. 76 Table 5.2 Invertebrate diversity within the undisturbed stream. 77 Table 5.3 Regression data for total invertebrate density. 89 Table 5.4 Regression data for Trichoptera density. 90 Table 5.5 Regression data for Chironomidae density. 90 v LIST OF FIGURES Page# Figure 2.1 Map of Study Area. 8 Figure 2.2 Layout of Study Streams. 10 Figure 3.1 Discharge levels for harvested streams F4 and F5 during study period. 24 Figure 3.2 Average concentrations of A l for undisturbed streams over the six-week study period. 27 Figure 3.3 Average concentrations of Ca, Mg, Na, N 0 3 and SiC>2 for undisturbed streams over six-week study period. , 28 Figure 3.4 Average concentrations of A l for harvested streams over six-week study period. 29 Figure 3.5 Average concentrations of Ca, Mg, Na, N 0 3 and S i 0 2 for harvested streams over six-week study period 30 Figure 3.6 Average concentrations of A l for treatments within the harvested streams. 30 Figure 3.7 Average concentrations of K for treatments within the harvested streams. 31 Figure 3.8 Percentage PAR for treatments within the harvested streams. 32 Figure 3.9 Data logger temperature data on harvested stream F4. 33 Figure 3.10 Data logger temperature data on harvested stream F5. 34 Figure 3.11 Data logger temperature data on undisturbed streams F6 and F7. 35 Figure 3.12 Average temperatures for undisturbed streams over the six-week study period. 36 Figure 3.13 Average temperatures for harvested streams over the six-week study period. 37 Figure 3.14 Average temperatures for treatments within harvested streams. 37 Figure 3.15 Average temperatures for reaches on harvested streams. 38 Figure 4.1 Average A F D M (periphyton) for undisturbed streams. 55 Figure 4.2 Average Chlorophyll a for undisturbed streams. 56 Figure 4.3 Average Chlorophyll a for undisturbed streams over the 6-week study period. 56 Figure 4.4 Average A F D M (periphyton) for treatments on harvested streams. 58 Figure 4.5 Average A F D M for harvested streams over the six-week study period. 58 Figure 4.6 Average A F D M (periphyton) for reaches on harvested streams. 59 Figure 4.7 Average Chlorophyll a for treatments on harvested streams. 60 Figure 4.8 Average Chlorophyll a for harvested streams over the six-week study period. 60 Figure 4.9 Average Chlorophyll a for reaches on harvested streams. 61 vi Figure 5.1 A F D M (organic matter) for regions within the undisturbed stream F7. 77 Figure 5.2 Average invertebrate densities for treatment areas on harvested streams and the undisturbed stream F7. 78 Figure 5.3 Average Plecoptera densities for treatments within the harvested streams. 79 Figure 5.4 Average Diptera densities for treatments within the harvested streams. 80 Figure 5.5 Average Diptera densities for treatments and reaches within the harvested streams. 80 Figure 5.6 Order richness for treatment areas within the harvested streams. 81 Figure 5.7 Ephemeroptera genera richness for treatments and reaches within the harvested streams. 82 Figure 5.8 Plecoptera genera richness for harvested streams. 82 Figure 5.9 Diptera genera richness for harvested streams. 83 Figure 5.10 Order level diversity for treatments within harvested streams. 84 Figure 5.11 Order level diversity for treatments and reaches within the harvested streams. 84 Figure 5.12 Ephemeroptera diversity for treatments and reaches within the harvested streams. 85 Figure 5.13 Plecoptera diversity for treatments and reaches within the harvested streams. 86 Figure 5.14 Diptera diversity for treatments and reaches within the harvested streams. 87 Figure 5.15 Benthic organic matter for treatments and reaches within the harvested streams. 88 vii ACKNOWLEDGEMENTS I would first like to thank my adviser Dr. Michael Feller along with my committee members, Drs. John Richardson, Rob Hudson and Les Lavkulich for their support, advice and assistance with all phases of this research. In addition I would like to thank Dr. Tony Kozak for his helpful assistance with the statistical analysis. I truly appreciated the hard work and assistance of Christine Bonish, Erin Pierce, Stefanie Pollock and Norah White in the field and laboratory. Their friendship and support during the field season was very important to the success of the project. For my mother and father, I want to express my great appreciation and thanks for their unending support, encouragement and love over the years as well as sacrifices they have both made to provide me with the opportunities I have had in life for pursuing my educational goals. I would not be where I am today were it not for their love and guidance. Lastly, I would like to thank my husband for his support, love and patience throughout this entire experience. And most importantly, for helping me remain focused and confident -realizing that we can accomplish anything with a little hard work and dedication. Thank you. viii CHAPTER 1. INTRODUCTION 1.1 GENERAL OVERVIEW Small headwater streams may constitute up to 85% of total stream length in a watershed, yet their riparian areas are often managed without consideration for the influence these small streams have on the overall stream ecosystem (Peterson et al. 2001). In order to fully understand how a stream ecosystem functions, we must first understand the connectivity between the watershed basin and the stream ecosystem (Giller and Malmqvist 1998). Any disturbance within the basin of a small headwater stream may ultimately affect downstream water quality and aquatic ecosystems. Anthropogenic disturbances have been shown to alter abiotic and biotic stream processes (Allan 1995, Montgomery and Buffington 1997). For example, forestry practices occurring on riparian hillsides have been linked to changes in channel morphology, water chemistry (i.e. N and P), water temperature, primary production and assemblage of aquatic fauna (Hansmann and Phinney 1973, Al lan 1995, Minshall et al. 1997, Carignan and Steedman 2000). First and second order stream systems (headwater streams) are vital components of the entire watershed ecosystem. While not always fish-bearing, they provide an important source of nutrients, primary production, food supply, sediment and organic debris for lower stream reaches (Burt 1996). However, in terms of management, these stream systems have not been provided the same level of protection that downstream fish-bearing stream reaches have been provided (Young 2000). If we are to manage our watersheds for both sustainable forestry and fisheries we must appreciate the important connection between small headwater streams and fish-bearing lower stream reaches. In addition, we must better understand the effect of forest harvesting activities on small headwater streams. 1 1.2 S T R E A M S T R U C T U R E A N D PROCESSES Stream structure and processes are a direct result of the interactions between the biological community, energy resources and physical habitat (Murphy and Meehan 1991). Stream processes are linked to physical conditions within the riparian zone (the area of linkage between terrestrial and aquatic systems). The riparian zone can be described as a three dimensional space consisting of the stream and stream-bed, upwards to the forest canopy and outwards to the edges of the floodplain/bankfull mark (Gregory et al. 1991). Interactions between the land and water in this zone may lead to changes in microclimate, nutrients, inputs of organic matter and composition of aquatic flora and fauna. Physical stream habitat is driven by underlying geology and soil type which controls stream structure and chemistry. Surrounding vegetation type has a strong influence on how streams transform available energy sources into food resources for upper trophic levels. For example, in heavily shaded streams, organic material from vegetation falling into the stream (allochthonous inputs) is colonized by microbes such as fungi and bacteria and is then consumed by invertebrate shredders (Murphy and Meehan 1991). In contrast, streams with little to no riparian canopy are dependent upon energy from the sun for the growth of periphyton and aquatic plants (autochthonous matter) which then provide a nutrient source for other organisms (Horne and Goldman 1994). Transformation of energy and nutrients differ on a spatial scale between headwater streams and downstream systems. For example, headwater streams are typically small and heavily shaded by riparian vegetation, transferring energy and nutrients through allochthonous inputs (organic material) (Gomi et al. 2002), while larger downstream systems with open riparian canopies rely upon particulate organic matter flushed from upstream tributaries as well as growth of algae and other aquatic macrophytes (i.e. moss) through photosynthesis (Murphy and Meehan 1991). Headwater systems have been shown to be a vital part of a watershed by supplying nutrients, sediments and particulate organic matter for downstream systems (Gomi et al. 2002). While headwaters play a significant role in the functioning of watersheds, their role is often underestimated and land management activities do not take into consideration 2 how headwater processes differ from downstream systems. Understanding how disturbances (e.g. forest harvesting, road building) alter headwater stream processes by altering organic matter inputs, chemistry, primary production and trophic food webs may help address how to better protect and manage these ecosystems. 1.3 T H E L O N G I T U D I N A L G R A D I E N T Forestry practices have been the primary land-use affecting forested headwater streams within the coastal mountains of western North America. In order to manage on the watershed scale, we must have better knowledge of the effects of removing the forest canopy across headwater streams as well as an understanding of any cumulative downstream effects. Better information of how forest harvesting may alter water quality and aquatic organisms along the longitudinal gradient may provide useful information for management decisions. Changes along a longitudinal gradient may be viewed on different spatial scales with large scale studies looking at changes throughout the entire watershed from headwaters to mouth, while smaller scales would consist of changes along reaches of streams from tens of meters to hundreds of meters in length. Understanding the connectivity of headwater streams to downstream systems along with recognition of stream processes functioning as a continuum was important to furthering the knowledge of the important role of headwaters in a watershed (Vannote et al. 1980). The river continuum concept describes how stream processes change along a longitudinal gradient based on changes in stream size, influence of riparian vegetation inputs and channel structure (Vannote et al. 1980). On a smaller scale, that of the present study, cycling of nutrients within stream reaches may influence biological processes along the longitudinal gradient of the stream reach depending on length of spirals (Newbold et al. 1982, Davis and Minshall 1999). Nutrient cycling or "spiraling" refers to the reuse and retention of nutrient ions as they cycle along a downstream gradient (Newbold et al. 1981). The length of a spiral is the distance an ion travels from dissolved to particulate and back to dissolved form (Essington and Carpenter 3 2000). Spiral length illustrates the distance a dissolved nutrient flows downstream until sequestered. The shorter the spiral length, the more limiting the nutrient may be within the system (Newbold et al. 1982). Nutrients are vital for many biotic processes, with nutrients such as nitrogen and phosphorous often limiting primary production in stream systems. Because nutrient levels affect biological processes, effects from nutrient spiraling may be seen as differences in biological processes along a downstream gradient. Logging may ultimately affect spiral lengths by altering flow paths and inputs of nutrients into stream systems. 1.4 E F F E C T OF F O R E S T H A R V E S T I N G O N S T R E A M PROCESSES Historical forest harvest practices had few regulations for the protection of water quality and stream habitat. Current forest harvest practices have attempted to protect stream structure and habitat through the use of buffers, restricted use of equipment near stream channels and through the practice of leaving woody debris in and around the stream channel as a nutrient source. While better forest harvesting practices are being utilized, changes to stream structure and water quality continue to occur, caused primarily by changes to the hydrologic regime (Harr 1986). Research has produced mixed results regarding the effect of forest removal on hydrologic regimes. This is most likely due to uncontrollable factors such as underlying geology (soils), vegetation differences and climatic influences. However, in western North America studies have shown that forest harvesting often leads to increased frequency of large peak flows (Hicks et al. 1991, Jones and Grant 1996, Toth 1998, Hudson 2001b). One reason for increased peak flows is the increased snow accumulation within harvested areas, followed by subsequent rain-on-snow events and increased rates of melting caused by direct solar radiation (Harr 1986, Jones and Grant 1996, Hudson 2001b). Within a forested watershed, the forest canopy wi l l intercept snowfall, leading to less accumulation on the forest floor and slower melt rates (Harr 1986). Removal of the forest canopy and vegetation causes decreased evapotranspiration and increased soil moisture, leading to overall increased water 4 yields (Jones and Grant 1996). Other possible explanations for higher peak flows, include the impact of overland flow re-routing caused by logging roads (Jones and Grant 1996). Larger peak flows, with increased frequency, are capable of significantly altering stream structure, habitat and biological communities (Toth 1998). While structural changes are often the root cause of subsequent effects to the aquatic system, biological processes are also directly affected by riparian forest harvesting. For example, research in a variety of geographical locations has shown that periphyton levels in harvested stream reaches are often higher than levels seen in forested streams (Hansmann and Phinney 1973, Noel et al. 1986, Feminella et al. 1989, Quinn et al. 1997, Kiffney and Bul l 2000). Increased periphyton biomass is primarily a result of increased light levels following removal of the riparian forest canopy. Forest harvesting has also been linked to changes in algal species, often moving towards an algal community dominated by filamentous algae which are not as easily digested by certain invertebrate consumers (Hansmann and Phinney 1973, Lowe et al. 1986). A shift in the algal community, which provides the main food source for grazing herbivores, has the potential to alter herbivorous invertebrates as well as other trophic levels within the stream food web (Anderson 1992, Stone and Wallace 1998). In addition to increased levels of primary production, there is often an effect on aquatic organisms following forest harvesting. Some studies have reported overall increases in total invertebrate densities following forest harvesting in riparian zones (Murphy and Hal l 1980, Hawkins et al. 1982). However, in contrast, a study in British Columbia reported decreases in invertebrate densities after logging due to habitat degradation from increased levels of sedimentation in study streams (Shortreed and Stockner 1983). Changes in riparian canopy density and vegetation type caused by logging often leads to shifts in invertebrate community composition (Anderson 1992, Stone and Wallace 1998). These varied results illustrate the importance of understanding why individual watersheds respond differently to forest harvesting. 5 1.5 OBJECTIVES While previous studies have evaluated the effect of riparian forest removal on variables such as water quality, fish species, periphyton biomass and macro invertebrates, more studies are needed to understand how stream processes differ geographically and how forest harvesting may affect the aquatic system along the longitudinal gradient of a stream as it flows through a harvested area and back into a forested area. The present study evaluates the effects of forest harvesting on several variables (periphyton biomass, benthic macro invertebrates and benthic organic matter). In addition, the present study evaluates the longitudinal gradient of the study streams in order to test for any variability between the harvested area and downstream recovery as a function of longitudinal distance along the stream channel. Specifically, I want to examine the following questions; 1. How do periphyton biomass, benthic macro invertebrate densities and diversity and benthic organic matter change along a longitudinal gradient within headwater streams? 2. Does forest harvesting across a section of stream affect incident solar radiation, stream temperature, periphyton biomass, benthic macro invertebrates and benthic organic matter within the harvested section and downstream of the harvested area? 3. What are the relationships between periphyton biomass, benthic macro invertebrates, benthic organic matter, water chemistry/nutrient concentrations, water temperature and incident solar radiation within these headwater streams? 6 C H A P T E R 2. R E S E A R C H DESIGN 2.1 STUDY A R E A DESCRIPTION The study site was located within the Roberts Creek Study Forest (RCSF) on the Sechelt Peninsula, approximately 40 km northwest of Vancouver, British Columbia (Figure 2.1). The study forest falls within the Pacific Ranges Drier Maritime variant of the Coastal Western Hemlock Zone ( C W H d m l ) . Climate is relatively mild, consisting of dry summers and wet winters with only small snow packs accumulating at various times during the winter. Average annual precipitation (primarily as rainfall) ranges from 1210mm to 1530mm annually with most precipitation falling between November-March (values taken from an Environment Canada weather station between 1984-1992 in Gibsons, B .C . ) . Average annual temperature ranges from 9.9°C to 11.3°C (values taken from Environment Canada weather station between 1989-1992 in Gibsons, B .C) . The topography within the drainage basin consists of relatively uniform stream gradients ranging from 13% to 17% with streambeds consisting of coarse gravel and larger boulders in most stream reaches with soil in the upper-most reaches. Underlying soils consist of glacial til l and colluvial soils (Dystric Brunisol) without much soil moisture storage due to rapid drainage (Valentine et al. 1978). The forest appears to have regenerated from a fire that burned through the watersheds approximately 140 years ago. Evidence of the fire is based upon tree coring of the oldest standing trees along with remnants of burnt snags and fallen logs throughout the watersheds. The forest composition is dominated by Douglas-fir (Pseudotsuga menziesii) with western hemlock (Tsuga heterophylla) and western red cedar (Thuja plicata) found among the forest canopy and understory. Some harvesting occurred in the late 1870's through the 1950's and consisted primarily of removal of dead and downed western red cedar (Hudson 2001b). Primitive roads and trails that were utilized for previous harvesting activities are still seen around the watersheds. Each of the harvested streams has at least two recently built road crossings with culverts for the harvesting that occurred between 1998 and 1999. 7 J A V J • Ar U \ Strait of ^ Georgia Q & n u > i i < m | Flume Creek Experimental Watershed Legend v Weir A Meteorological Site Creek Catchment Boundary Road Tenure Boundary Forested Area Harvested Area Scale (metres) 200 400 GOO fiOO Figures 2.1 Map of Study Area 8 2.2 E X P E R I M E N T A L D E S I G N Six headwater streams (1 s t and 2na order) were selected for sampling in the present study. Three of the streams (harvested streams) had been logged in 1998 and 1999. Three of the streams selected as forested streams (undisturbed streams) had not been actively harvested within the past 50 to 100 years. Each stream had nine study reaches for an average stream length of 700 meters from the uppermost reach (reach 1) to the downstream reach (reach 9) (Figure 2.2). On harvested streams, reaches 1-3 are referred to as above-harvested (forested), reaches 4-6 are harvested and reaches 7-8 are below-harvested (forested reaches below the harvested area). A l l six study streams were located on the relatively uniform Mount Elphinstone slope at elevations between 400 to 750 meters, with a southerly aspect and are first and second order creeks draining narrow elongated catchments (Hudson 2001b). These streams are classified as S6 under the Forest Practices Code of B C Act and Regulations. They are given this designation, because they are small, non-fish bearing streams. Because these streams are small and are not used directly by fish-species or for drinking water, there was no requirement for streamside riparian buffers under the Forest Practices Code. The harvested streams were labeled F4, F5 and F8 and the undisturbed streams were F6, F7 and F9. Logging occurred between fall of 1998 and fall of 1999, carried out as an experiment in the Roberts Creek Study Forest (RCSF) of the Sunshine Coast Forest District to determine the effect of logging on the water quality and discharge levels of streams F4, F5 and F6 (control) (Hudson 2001b). The logging types tested included variable retention and strip shelterwood. The variable retention harvesting involved both grouped (many trees in a clustered reserve) and dispersed retention (individual trees left as habitat refuge) and left approximately 18% of the canopy within the harvested boundary. The strip shelterwood harvesting left approximately 49% of the canopy within the harvested area in leave strips that crossed the main stream channel of watershed F5. However, for the present study a first-order tributary of the F5 watershed which did not receive the leave strip treatment was chosen for sampling in order to have replicates of three headwater streams that had been logged across the stream channel. 9 HARVESTED STREAMS Stream size was variable, but each of the streams averaged between 1 to 3 meters in width. Further characteristics of the study streams are given in Table 2.1. Table 2.1 Characteristics of the study streams: Values are approximated. Discharge levels are averaged for the study period of May 1 to July 31, 2002 (reflecting low flow conditions). Harvested Streams: Study Stream Avg. Stream Slope Watershed Area Mean Discharge Levels Harvested Length % of watershed area that was harvested F4 17% 39 ha .0043 m3/s 750 meters 38.5% F5 17% 61 ha .0069 m3/s 550 meters 17.3% F8 16% 45 ha .0048 m3/s 250 meters 5.8% Undisturbed Streams: Study Avg. Stream Watershed Mean Discharge Stream Slope Area Levels F6 13% 16 ha .0024 m3/s F7 13% 65 ha .0098 m3/s F9 15% 39 ha .0041 m3/s 11 CHAPTER 3. WATERSHED PHYSICAL AND CHEMICAL PROPERTIES 3.1 INTRODUCTION Nutrients/Chemistry: -Nitrogen and Phosphorus-Understanding the chemical properties and cycles of a stream is important for understanding the biological processes within the stream (Home and Goldman 1994). In particular, nitrogen and phosphorus are considered the two most important nutrients for aquatic plants and animals. Nitrogen is most commonly found in the forms of nitrate ( N O 3 ) and ammonium ( N H / ) and phosphorus is most common in the form of phosphate (PO4 3 ) within aquatic systems. Both of these nutrients are considered to be growth-limiting, with increases or decreases in these two nutrients leading to changes in the growth rates of aquatic flora and fauna (Lowe et al. 1986, Rosemond 1994, H i l l et al. 1995). The concentration of nutrients within streams is directly related to adjacent riparian land use such as forest harvesting (Home and Goldman 1994). While nitrogen and phosphorus cycles are similar in different parts of the world, the major difference between geographical locations is the amount of rainfall. Nitrogen is derived primarily from rainfall and therefore areas of higher rainfall may have relatively higher levels of nitrogen within stream systems (Krebs 1994). -Other Nutrients-Other chemicals important within stream systems include Si02, Ca, M g , Na, K , CI and Fe (Home and Goldman 1994). For example, S1O2 is important for the cell walls of diatoms (algae), while C a is important for metabolic processes and provides the material necessary for skeletal structures of many organisms. Magnesium is not considered a limiting factor as it is typically present in sufficient quantities; however, it is a crucial ion for energy-transfers in living cells. Sodium and K are also not considered limiting; however, too much Na can negatively impact species that are unable to tolerate high Na levels. Chloride ions are 12 required by photosynthesizing cells, and are typically measured in order to determine previous levels of free chlorine (a toxic element - which is converted to chloride in a few minutes in water) within aquatic systems. Of the "trace elements" that are required in moderate levels for organisms, Fe is the most important for organisms and is often considered to be in demand within aquatic ecosystems (Home and Goldman 1994). -Nutrient Cycling-Nutrients may cycle through forests at different rates, yet the overall cycle is similar between forests and generally consists of vegetation taking up nutrients during growth and accumulating those nutrients within their leaves and wood. Nutrients are subsequently returned to the soil layers through the dispersal of vegetation falling and decomposing onto the forest floor. Nutrients within the soil layers are either taken up again by growing vegetation, filtered down through the soil layers to groundwater or directly flushed into stream systems during rain events (Krebs 1994). -Seasonal Cycles-Chemical cycling and concentrations within forested streams are controlled by a variety of factors including underlying geology, climate (i.e. precipitation), bacterial activity, plant uptake and forest canopy (Home and Goldman 1994). Seasonal cycles are dominated by increased concentration of nutrients in late summer and fall followed by a decrease in concentration in winter and spring (Feller and Kimmins 1979; Mulholland and Hil l 1997). This is primarily a result of an accumulation of nutrients on land during the dryer periods when warmer temperatures increase biological and geological breakdown followed by subsequent flushing of sediments into stream channels during rain events (Home and Goldman 1994). Similar seasonal cycles were seen in the Flume Creek study streams prior to harvesting of the study sites (Hudson 2001a). 13 -Longitudinal Trends/Nutrient Spiraling-Nutrient cycling within stream systems is similar to cycles found within terrestrial systems. Nutrients that have flushed into the stream system from rainfall, groundwater and vegetation falling into the stream eventually become a part of the stream nutrient cycle. Once the nutrients have entered the stream system, they begin to cycle within the stream. Nutrient cycling combined with downstream transport of dissolved ions is referred to as longitudinal "spiraling" of nutrients from dissolved to particulate and back to dissolved form as the ion flows downstream (Newbold et al. 1981, Newbold et al. 1982). The length of a spiral is the downstream distance a dissolved ion travels starting from a dissolved ion to particulate form and back to a dissolved ion (Essington and Carpenter 2000). For example, nitrogen may follow a cycle that consists of a nitrate or ammonium ion entering the stream through rainfall where it is taken up by algae for growth, the algae are then consumed by invertebrates and the nitrogen is then excreted as ammonium which is recycled back to algae or flushed downstream (Home and Goldman 1994). When dissolved nitrogen is in excess of demands within the system, it is passed downstream until taken up again by algae growing downstream. Nutrient spiraling is an important ecological process in streams, because recycled nutrients may often be the main source of nutrients for primary producers in dryer regions and seasons (Tripathi and Singh 1994). In addition, nutrient spiraling lengths illustrate the distance a dissolved ion travels downstream until sequestered. The shorter this distance, the more limiting the nutrient may be within the stream system (Davis and Minshall 1999). When nutrients are growth-limiting, the impact may be felt within the trophic levels of a stream. Typical spiraling lengths range from as little as 22-97 meters found in streams in Tennessee (Mulholland et al. 1985) to distances as long as 188-660 meters in small streams in Oregon (Munn and Meyer 1990). Different spiraling lengths may be a result of a variety of factors including nutrient concentrations within the stream, and the ability of particulates to sequester nutrients and discharge levels, with higher levels of water flow increasing the transport of dissolved nutrients downstream (Essington and Carpenter 2000). 14 Other longitudinal changes have been found in headwater streams at Hubbard Brook in a study that compared chemical processes and concentrations at different elevations along headwater streams (Johnson et al. 2000). Biogeochemical processes were found to be relatively heterogeneous between watersheds and elevations (longitudinal gradients between 550-750 meters), with minor differences between study sites caused by differences in base mineral soils, organic matter and discharge levels. Johnson et al. (2000) suggested that the minor differences seen in dissolved ion concentrations were a result of differences in stream substrate type and levels of benthic organic matter in the streambeds at different elevations. These differences may impact the cycling and flushing of nutrients to downstream reaches. For example, headwater streams typically have increased levels of uptake and transformation of inorganic nitrogen compared to lower stream reaches, showing the important role that headwaters may play in the cycling and transport of nutrients (Peterson et al. 2001). -Effects of Harvesting-Water quality is one of the primary concerns of local communities that depend on watersheds for their water supply. Forest harvesting occurring in these watersheds could ultimately affect water quality in various ways. The effects of forest harvesting on water quality often includes increased water temperatures, increased fine sediments and increased nutrient levels (Murphy and Hal l 1981; Johnson and Jones 2000). The disruption of biogeochemical cycles, and thus water quality, is caused by changes to nutrient sources, increased soil temperature and increased sediment input (Carignan and Steedman 2000). Any alteration of the physical/chemical properties of a stream may ultimately impact the biotic processes (i.e. primary production) within a stream system (Horne and Goldman 1994). Forest harvesting often reduces uptake of nutrients by vegetation, increases soil decomposition rates and allows precipitation to reach mineral horizons thereby increasing availability of nutrients within soils (Brown and Binkley 1994). While forest harvesting may increase concentrations of a variety of chemicals, NO3 and PO4 are most important for the aquatic ecosystem (Horne and Goldman 1994). While significant changes in PO4 are 15 uncommon following forest harvesting, N O 3 levels were seen to increase in nearly 70% of studies in a review of forest harvesting effects on dissolved ion concentrations (Brown and Binkley 1994). Other studies have also shown significant increases in stream nutrient levels following forest harvesting (Murphy and Hal l 1981, Feller and Kimmins 1984). Nitrate in particular is considered a sensitive indicator of stream disturbance and forest harvesting often results in increased N O 3 concentrations (Home and Goldman 1994). However, increases in nutrient concentrations during summer may be offset by the nutrient demands of aquatic vegetation in harvested areas (Likens et al. 1970). Extensive studies carried out on nutrient cycling and budgets at the Hubbard Brook Experimental Forest in New Hampshire, U . S . A . , evaluated the impact of removing vegetation through logging and herbicide application on nutrient levels and nutrient budgets (Likens et al. 1970, Bormann et al. 1974). Cutting of trees and depressing regrowth of vegetation led to increases in debris within the stream system (primarily from debris left on the ground following treatment) along with increased discharge levels (Likens et al. 1970). A l l major ions increased following forest harvesting, with nitrate increasing 40 to 60-fold and exceeding drinking water recommendations within the first two years. Other increases of over 100% in Ca, M g , K and Na were also seen within the first two years after treatment (Likens et al. 1970). With nutrients considered a growth-limiting factor for many aquatic organisms, increases and/or decreases in nutrient concentrations in stream water may affect growth of aquatic flora and fauna. In logged areas, where solar radiation is no longer limiting to periphyton growth, nutrients have been shown to be growth-limiting (Hansmann and Phinney 1973, Kiffney and B u l l 2000). Therefore, increases to nutrient levels caused by logging in these streams may lead to further increases in periphyton growth. Increased primary production may ultimately lead to impacts on other trophic levels within the food-web. 16 Solar Radiation: -Longitudinal Trends-Solar radiation is subsequently referred to as "light" in the text. Changes in light levels along the longitudinal gradient are directly related to canopy coverage. For example, light levels may naturally increase along the longitudinal gradient of a stream as the stream becomes larger and canopy opens up allowing more light to reach the stream (Vannote et al. 1980). -Effects of Harvesting-Light is recognized as a primary controlling factor within stream ecosystems because of its affect on stream temperatures which then may alter growth and metabolism rates in many aquatic species (Vannote et al. 1980). In headwater streams with dense riparian canopies, the ecosystem functions with lower light levels reaching the stream. Disturbances that lead to forest canopy removal (i.e. harvesting) increase light levels which increases stream temperatures which ultimately may alter stream biology (Noel et al. 1986). Light levels were 13 to 30 times higher in clear-cut stream reaches compared to controls in small headwater streams in British Columbia (Kiffney and B u l l 2000). Light is often found to be the primary limiting factor for primary production in stream ecosystems (Murphy and Hal l 1980, H i l l and Knight 1988). Increased light levels following forest canopy removal increase stream temperatures and primary production (Noel et al. 1986, Quinn et al. 1997, Kiffney and B u l l 2000). Temperature: Temperature is a vital component in the metabolic processes of aquatic flora and fauna. Stream temperature controls the rate of biotic and abiotic processes which influence overall ecosystem dynamics (Johnson and Jones 2000). For example, increased water temperature enhances primary production and causes accelerated decay rates of organic matter, which in turn may lead to increases in aquatic invertebrates (Rempel and Carter 1986). Increased 17 water temperatures may also impact invertebrate and fish emergence and growth. Higher stream temperatures have been associated with earlier emergence from eggs and faster growth in many invertebrate and fish species (Holtby 1988). Stream temperature patterns consist of both seasonal and diurnal (temperature variation occurring within one day) variation, running on a relatively predictable temperature regime (Vannote and Sweeney 1980). -Longitudinal Trends-Temperature change along a stream's longitudinal gradient typically consists of lower temperatures in the headwaters and increasingly higher temperatures in lower stream reaches. Longitudinal changes occur in all climactic environments including the arctic where stream reaches in the lowland of the Brooks River in Alaska can reach 15°C in summer (Home and Goldman 1994). In addition, a harvested headwater stream (800 meter stream length) in the northeastern region of the U.S . resulted in cumulative increases in water temperature along the longitudinal gradient from the headwaters to the mouth of the stream (Lynch et al. 1984). -Effects of Harvesting-Forest harvesting often causes increases in peak stream temperatures and larger diurnal fluctuations (Hansmann and Phinney 1973, Shortreed and Stockner 1983, Brown 1985, Noel et al. 1986, Holtby 1988, Hetrick et al. 1998b, Johnson and Jones 2000). Removal of riparian canopy may raise peak stream temperatures by 5°C or more (Brown and Binkley 1994). Temperature is controlled by a variety of factors including amount of solar radiation reaching the stream, soil temperatures, discharge levels, elevation, topography and weather patterns (Ward and Stanford 1982). However, the primary factor controlling stream temperatures is the amount of solar radiation reaching the stream and surrounding soils (Brown 1985, Brosofske et al. 1997). In recent years studies have illustrated a strong relationship between soil temperatures and water temperatures (Brosofske et al. 1997). Headwater streams are often controlled by 18 influxes of groundwater and therefore any increase in soil temperatures could lead to increases in stream temperatures (Vannote and Sweeney 1980). The strong correlation between increased soil temperatures and increased stream water temperature (even with harvested areas 50 meters or more from the stream) may put into question the effectiveness of small width riparian buffers for the protection of stream temperatures (Brosofske et al. 1997). However, riparian buffers have been shown to minimize increases in peak temperatures to less than 2°C in most cases (Brown and Binkley 1994). Forest harvesting often affects flow-paths and discharge levels, leading to increased peak flows and lower discharge levels in dryer seasons. Water temperatures are typically higher in lower flow conditions, showing that discharge levels may also be a controlling factor for stream temperature increases (Hetrick et al. 1998b). The influence of canopy on headwater stream temperature is also an important variable to consider for management of stream ecosystems. In small streams, the canopy density can be greater than 90% and therefore these systems may be more responsive to the presence or absence of vegetation (Brown and Binkley 1994). Objectives: This chapter focuses on the current physical/chemical characteristics of the study streams. The physical/chemical data were obtained for the streams to see if they could help to explain any observed changes in biological variables (i.e. periphyton and invertebrates). However, due to the lack of pre-harvest data, this data was not analyzed to determine if harvesting had any effect on the physical/chemical properties of the study streams. Chemistry studies should consist of pre-harvest data over a sufficient period of time in order to produce definitive results regarding harvesting effects. However, trends in the data may illustrate summertime conditions and differences between streams and treatment areas during summer conditions. Any relationships between the physical/chemical characteristics and biological factors (i.e. periphyton and invertebrates) within the study streams wi l l be discussed in subsequent chapters. 19 3.2 METHODS Study Area: r The study site was located within the Roberts Creek Study Forest (RCSF) on the Sechelt Peninsula, approximately 40 km northwest of Vancouver, British Columbia (Figure 2.1). Six streams were selected for study, three of which had been recently harvested (F4, F5 and F8) and three of which remained undisturbed (F6, F7 and F9). A l l six study streams were located on the relatively uniform Mount Elphinstone slope with a southerly aspect and consist of first and second order creeks draining narrow elongated catchments (Hudson 2001b). A l l study reaches were located between elevations of approximately 400 to 750 meters above sea level and drained slopes of approximately 13-18% (Figure 2.2). Stream size was variable; however, each of the streams averaged between 1-3 meters in width. Logging occurred between fall of 1998 and fall of 1999. The study area is described in more detail in Chapter two. Field Sampling and Laboratory Analysis: -Water Chemistry-At the bottom end of each study reach, a lOOmL grab sample of stream water was collected once a week during the six-week sample period (n=6 samples per reach). Samples were frozen upon return to the laboratory and processed at the Glyn Road laboratory operated by the B . C . Ministry of Forests Research Branch. Cations ( A l , Cu , Fe, Na, M n , Zn) were measured on an A R L 3560 ICP instrument using off-peak background correction and subsequent subtraction of a M i l l i - Q (18 megohm) water blank to compensate for any plasma on-peak interference. The ICP instrument was calibrated using NIST traceable ICP calibration standards. Major cations (Ca, Na, NH4, M g , K) and anions (CI, PO4, N O 3 , SO4) were analyzed by ion chromatography using a Waters "Action Analyzer" H P L C system. The cations were separated on a Waters IC-Pak Cation M / D column and detected by conductivity. Anions were separated on an Anion H R column and detected by conductivity, except for N O 3 which was detected by a U V visible wavelength detector for better detection limits. 20 Calibration was done using NIST traceable ion chromatography standards (Hudson 2001a). During week 5 where too many samples were missing (due to dried streams) all samples were discarded. Only weeks one and two had full sample sets. In weeks 3, 4 and 6 the only reaches and streams chosen for analysis were those where the reach maintained stable water flows and data was complete for statistical analysis. -Solar Radiation-Photosynthetically active radiation (PAR) readings were taken on a uniformly cloudy day (with a little precipitation) on March 19, 2003 using a Sunfleck Ceptometer (Decagon, Pullman, WA). Three readings, one in the upper, middle and lower sections of each study reach were taken between 10:00 and 14:00 in order to capture average light levels during midday. Six readings (three readings taken in the beginning of sampling and end of sampling time) were taken within the harvested area and averaged to determine above canopy incident PAR from forest canopy. Absolute values were not calculated since readings were taken in early spring and would not reflect annual, seasonal or maximum summer levels. The PAR reading for each reach was expressed as the percentage of the total incident PAR by dividing the above canopy PAR by the average PAR measured along a reach. -Temperature-Temperature measurements were made in each stream and reach once a week during the study period (with the exception of the first week). These measurements were taken with thermometers placed into the stream reach near the location of periphyton sampling sites. Measurements were made between 08:30 and 12:00 in the harvested streams and between 10:00 and 14:00 in the undisturbed streams, with measurement times consistent throughout the weeks. In addition, data loggers (Unidata "Macro" 7000) were utilized to measure stream temperature every 15 minutes from the beginning of June until the end of August. This was done near reach 9 (lowest reach) in each of the undisturbed streams (F6 and F7) and in the bottom reach of the harvested area and below-harvested (approximately 50 meters into the forested area) in two of the harvested streams (F4 and F5). 21 Statistical Analyses: While the main objective was to utilize physical and chemical data for comparisons with biological factors, analysis of variance was used to test for differences longitudinally and between the harvested and forested areas to see if results could be used to explain results in the biological factors. The physical/chemical data was not analyzed to determine effects of the previous forest harvesting. The statistical analysis for undisturbed streams was analyzed as a General Linear Model ( G L M ) analysis of variance ( A N O V A ) split-plot design; with the split-plot design utilized because of the added effect of time (weeks) on the design ( S Y S T A T 1998) (Table 3.1). Table 3.1 A N O V A table for undisturbed streams showing split-plot design. ANOVA: Split-Plot (Undisturbed Streams) Design Source of Variation Degrees of Freedom Whole Plot Reaches g - l = 8 Error 1 n ( g - 1 ) = 24 Split Plot Week t - 1 =5 Reaches x Week ( g - 1 ) (t - 1 ) = 40 Error 2 g ( t - l ) ( n - l ) = 90 g =9 reaches t = 6 weeks n = 3 replicates/streams To analyze for differences in nutrient concentrations and stream water temperatures between harvested and forested sites, a General Linear Model ( G L M ) analysis of variance ( A N O V A ) was applied for a nested split-plot design using the statistical program Systat 9.0 ( S Y S T A T 1998) (Table 3.2). The General Linear Model ( G L M ) was used due to missing values in the data set. The experimental design was nested because reaches were treated as being nested within the treatments (above-harvested, harvested and below-harvested areas) and streams. This accounts for the fact that reach 1 on harvested stream F4 is not the same as reach 1 on harvested stream F5 and therefore reach's 1 cannot be treated as replicates during analysis. The split-plot design was utilized because of the added effect of time (weeks) on the design. 22 Table 3.2 ANOVA table for harvested streams showing split-plot nested design. ANOVA: Split-Plot (Nested) (Harvested Streams) Design Source of Variation Degrees of Freedom Whole Plot Treatment k - 1 =2 Reaches (Treatment) k ( g - l ) = 6 Error 1 k g ( n - l ) = 1 8 Split Plot Week t - 1 = 5 Treatment x Week ( k - 1 ) ( t - 1 ) = 10 Week x Reaches (Treatment) t g ( k - l ) = 30 Error 2 k g ( t - l ) ( n - l ) = 90 k — 3 treatment t — 6 weeks g = 3 reaches n = 3 replicates/streams When the above statistical analyses showed significant differences for effects, Tukey's multiple comparison test was used to determine specific differences. To account for instances of non-normality within data and heterogeneous variances, data were Logio (X,- + 1) transformed. Data in tables in this chapter are presented as untransformed means ± 1 standard error. 3.3 RESULTS Stream Discharge: Discharge data was available from harvested streams F 4 and F 5 (data used with permission of Rob Hudson, B.C. Ministry of Forests). Discharge levels revealed that there were fluctuations in discharge with rain events in early spring and then declining levels throughout the study period (Figure 3.1) 23 0.030 5/06 5/20 6/03 6/17 7/01 7/15 Date Figure 3.1 Discharge levels for harvested streams F4 and F5 during May 1 - July 18, 2002. Water Chemistry: Average concentrations for dissolved ions within each of the undisturbed streams are listed in Table 3.3, while average concentrations within each section of the harvested streams are listed in Table 3.4. Those dissolved ion concentrations showing no significant differences among treatments and streams are not reported or discussed within this chapter, however average concentrations are listed in the tables below for all dissolved ions sampled. 24 Table 3.3 Average stream water chemistry concentrations (standard deviation) (mg/L) for each undisturbed stream during the study period of May 31 -July 7, 2002. (ND refers to concentrations not detectable in laboratory tests). Dissolved Undisturbed Undisturbed Undisturbed Nutrients Stream Stream Stream F6 (n=27) F7 (n=41) F9 (n=29) Al .01 (.01) .03 (.02) .03 (.01) NH 4 .02 (.03) .02 (.02) .02 (.04) Ca .99 (.24) 1.30 (.06) 1.04 (.16) Cu .02 (.02) .02 (.02) .02 (.03) CI .88 (.18) .81 (.09) .76 (.08) Fe .01 (.01) .02 (.01) .02 (.03) K .06 (.07) .10 (.04) .05 (.03) Mg .24 (.05) .27 (.03) .18 (.01) Mn ND ND ND N0 3 .02 (.02) .12 (.09) .05 (.08) P0 4 ND ND ND Si0 2 2.15 (.95) 2.24(1.01) 1.26 (.57) Na 1.28 (.27) 1.25 (.09) 1.08 (.08) S0 4 1.70 (.32) 1.37 (.23) 1.09 (.20) Zn ND ND ND 25 a o u ** C3 ( « c is e S 5. s ® "3 •sL-' ro •2 ? >• o ' — ' 05 /—* .2 i i 2 .S ^ *j 1- © e g -w u ^ s-C se © C u a ^. u _ u u C -w ••• «5 03 w 5 ^ 3 u « a u « & w j». 4i 13 5 * § 2 » S © 8 5 •b -S 2 « 2 c 2 .5 S 3 '5 8 CO s 08 H • c CD •g 'o c Above-Harvested Harvested Below-Harvested Figure 3.8 Photosynthetically active radiation expressed as a percentage of the total incident PAR. Error bars are + 1 standard error. Bars with different letters are significantly different at P < 0.05 (Tukey's Test). Temperature: -Data Loggers-During the course of the study (June 6 - July 18) temperatures within harvested reaches averaged 13°C with undisturbed reaches averaging 9.9°C. However, harvested streams exhibited greater diurnal fluctuations with average daily temperature differences of 5°C (range of 8°C to 19°C) (Figures 3.9 and 3.10) compared to forested reaches where the average temperature fluctuation was 2° (range of 8°C to 11°C) (Figure 3.11). 32 33 Lf) 0 0 O o CD CO Q h CD I s -CD O CO O O l-H SO a 3 I in fa C3 J3 Si s a u a> a 2! o 22 3 O s & o u 3 34 -Undisturbed Streams-Temperature changed significantly with time in the undisturbed streams, with temperature significantly lower in week 2 than in all other weeks (Appendix 4) (Figure 3.12). -Harvested Streams-Temperatures within the harvested streams exhibited significant treatment by week interactions and differed significantly with time and between treatments (Appendix 4). Harvested reaches exhibited higher temperatures compared to the above and below-harvested reaches during the first 5 weeks, but similar temperatures in week 6 to those found in the below-harvested reaches. Temperatures were significantly lower in week 2 than in weeks 3, 4 and 5 and significantly higher in week 3 than in weeks 4, 5 and 6 (Figure 3.13). In general, temperatures in above-harvested reaches were significantly lower than those in harvested and below-harvested reaches (Figure 3.14). Temperatures increased throughout the harvested reaches and subsequently decreased as the streams flowed into the forested reaches (Figure 3.15). Tempera tu re 13 -i — — | 12 -SQ 10 -9 -7 2 3 4 5 6 Weeks Figure 3.12 Average temperatures for undisturbed streams from week 2 to week 6 of the study period. Data points with different letters are significantly different at P < 0.05 (Tukey's Test). Error bars are ± 1 standard error. 36 Temperature 13 -i 7 H 6 H 1 1 1 1 1 1 2 3 4 5 6 Weeks Figure 3.13 Average temperatures for harvested streams from week 2 to week 6 of the study period. Data points with different letters are significantly different at P < 0.05 (Tukey's Test). Error bars are ± 1 standard error. Tempera ture 4 2 0 a T --Above-Harvested Harvested Below-Harvested Figure 3.14 Average temperatures for treatments within the harvested streams (averaged from data during six-week study period). Bars with different letters are significantly different at P < 0.05 (Tukey's Test). Error bars are + 1 standard error. 37 Temperature 13 " I 12 H 11 H 10 9 H I I ^ • T 4 5 6 Harves ted 1 2 3 A b o v e - H a r v e s t e d T— T " T 7 8 9 B e l o w - H a r v e s t e d R e a c h e s Figure 3.15 Average temperatures for each reach in harvested streams for each reach. Reaches 1-3 are above-harvest reaches, 4-6 harvested reaches and 7-9 below-harvested reaches. Data are averaged for all weeks and error bars are + 1 standard error. 38 3.4 DISCUSSION Stream Discharge: Discharge levels in harvested streams F4 and F5, while not statistically analyzed, exhibited decreasing trends during the study period. Discharge levels decreased from an average of 0.025 m 3/second to 0.002 nrVsecond over the study period of M a y 31 - July 7, 2002. Stream flow was sporadic along the longitudinal gradients of the study streams, with the streams at times flowing underground and then resurfacing downstream (personal observation). Chemicals/Nutrients: -Undisturbed Streams-Dissolved ions within the undisturbed streams differed significantly only with time. Aluminum concentrations decreased over the six-week study period. Decreasing discharge during the study period may be the primary reason for decreases in A l . Aluminum concentrations have been found to be positively related to discharge (Likens et al. 1970). In contrast, Ca, M g , Na, NO3 and S i O i concentrations increased during the study period. Increases in these dissolved ions may also be linked to discharge levels as concentrations of these ions are typically inversely related to discharge (Feller and Kimmins 1979). Lower water flows and reliance upon groundwater flows during base flow conditions may often lead to increases in concentrations within a stream system (Likens et al. 1970, Feller and Kimmins 1979, Mulholland and H i l l 1997). No significant longitudinal differences were found within the undisturbed streams. This revealed that there was no obvious major change in nutrient cycling processes or in abiotic and biotic processes which may alter chemical concentrations along a longitudinal gradient. 39 -Harvested Streams-General: While the overall water quality of forested streams tends to be very high, forest harvesting often alters water quality through increased sedimentation and increased concentrations of dissolved nutrients (Binkley et al. 1999). Forest harvesting often leads to increases in certain ions dissolved in stream water (Likens et al. 1970, Bormann et al. 1974, Silsbee and Larson 1983). High concentrations of N O 3 along with N H 4 and toxic levels of metals such as Hg are seen as harmful to both aquatic species and drinking water (Binkley et al. 1999). For example, NO3-N concentration guidelines for drinking water both in Canada and in the U.S . have 10 mg/liter as the upper limit, while average levels of NO3-N found in forested watersheds are typically around 0.2 mg/liter (Binkley et al. 1999). In the present study NO3 levels were similar to those reported by Binkley et al. (1999) with concentrations ranging from .01 to .73 mg/liter. Neither NO3 nor any other chemical had concentrations exceeding recommended drinking water levels (Guidelines 1987). While safe levels were met for drinking water quality, it is still important to evaluate dissolved ions and their interactions with biological processes. Although water chemistry data in the present study was utilized primarily for explaining any trends found in biological characteristics (periphyton/invertebrates) within the study streams, some observations can be made from the statistical analysis and trends within the chemical data. Significant differences were found in dissolved ion concentrations ( A l and K) between treatment areas as well as between streams, suggesting that underlying geological differences and groundwater flow-paths may be contributing to these differences. While differences were seen among some dissolved ions, nitrate and phosphate (important nutrients for primary production) did not exhibit significant differences between treatment areas or watersheds over the six-week study period. 40 Treatment by Week Interactions: Significant interactions between treatments and weeks illustrated different trends in ion concentrations between treatments during weeks within the study period. Differences between treatment areas may be a function of declining discharge during the study period, with above-harvested reaches drying out by the end of the sample period (Figure 3.14). For example, all treatments had decreasing A l concentrations during the first three weeks as in the undisturbed streams. However, after week 3 A l concentrations increased in the above-harvested reaches but remained stable in the harvested and below-harvested reaches. Differences in substrate composition and discharge levels along the longitudinal gradient of the stream may explain these differences, with substrate in the above-harvested reaches consisting primarily of mineral soil and gravels and in the harvested and below-harvested reaches consisting primarily of gravel and boulders. Finer mineral substrates may lead to increased leaching of A l directly from the soils into the stream water (Likens et al. 1970). Sodium and SiC>2 concentrations generally increased over the study period. Increases over the study period were most likely a result of decreases in stream discharge concentrating the remaining dissolved ions within the stream. However, differences between treatments were most likely a result of mineralogical differences between reaches, as Na and Si02 concentrations are derived primarily from mineral weathering (Feller and Kimmins 1979) Sulfate concentrations within harvested reaches were higher than in above-harvested reaches in week 1 but decreased during week 2 to values lower than in the above-harvested reaches. However, the absolute increase over the study period was only 0.2 mg/L with concentrations in week 6 decreasing to concentrations 0.6 mg/L less than in previous weeks. While the above trends in dissolved ions help to illustrate underlying differences between reaches, these particular dissolved ions are typically not limiting in aquatic systems and therefore should not impact productivity of aquatic organisms. 41 Treatments: Aluminum and K concentrations differed significantly between treatments. Aluminum exhibited higher concentrations in the above-harvested reaches compared to the harvested and below-harvested reaches. Aluminum is primarily derived from the breaking down of minerals from bedrock and ti l l (Likens et al. 1970). Differences in concentrations seen in the present study may be a result of different substrate material as previously stated for other dissolved ions. Potassium concentrations were significantly lower in above-harvested reaches compared to the harvested and below-harvested reaches. Differences found within the present study may be a result of overall greater quantities of organic matter and quicker rates of breakdown of organic matter occurring in the harvested reaches, as K concentrations have been linked to leaching of potassium from organic matter (Likens et al. 1970). Higher concentrations in below-harvested reaches may be a result of downstream transport of dissolved K ions. Weeks: Aluminum, Ca, K , M g , Na, NO3 and SiC>2 concentrations differed significantly with time. Aluminum concentrations decreased within the harvested streams as they did in the undisturbed streams, with concentrations in week 1 significantly lower than in weeks 3 and 4. Concentrations of the other ions generally increased from weeks 1 and 2 to subsequent weeks. These results are most likely related to decreasing discharge levels within the study streams as discussed above. Potassium concentrations are not as clearly related to discharge levels (Feller and Kimmins 1979) so increases over the study period may be due to differences in rates of mineral weathering and ground water inputs during the study. Solar Radiation: -Longitudinal Trends-Solar radiation did not exhibit any longitudinal trends along the undisturbed streams in the present study. This is primarily a result of no differences in canopy coverage along the 42 undisturbed streams. While the river continuum concept states that light levels change along a longitudinal gradient, the concept involves changes in stream order and size, as the stream size increases and riparian canopy naturally opens up light levels increase (Vannote et al. 1980). The present study streams did not change in order or size along the longitudinal gradient studied. -Effects of Harvesting-Solar radiation was greater in harvested areas than in forested reaches, reflecting the effects of forest canopy removal on light levels. Other studies have shown similar increased light levels reaching the streambed when the forest canopy is removed (Noel et al. 1986, H i l l and Knight 1988). Photosynthetically active radiation (PAR) (covering the spectrum of light from 400 to 700 nm) is used for measuring light levels because it records the under-water spectrum of light used by algae for photosynthesis (Horne and Goldman 1994). Therefore, increases in P A R lead to increased light for primary production. Stream algae have been shown to be growth limited by light availability, with increases occurring following forest canopy removal (Hi l l and Knight 1988, Kiffney and Bul l 2000, Kiffney et al. 2003). Temperature: -Longitudinal Trends-Downstream temperature increases along the longitudinal gradient may also be important to the biological processes of stream systems. No significant differences were seen within the undisturbed streams, suggesting that the length of stream studied (without change in order or size of stream) did not cause any differences in stream temperature longitudinally. While longitudinal results were not significant on the harvested streams, stream temperatures did reveal decreasing trends as the streams flowed into the below-harvested reaches. This is referred to as downstream recovery with a study in California streams reporting immediate declines of 1-2°C within the first 130 meters and full recovery to forested stream temperatures 1.6km below the harvested area (McGurk 1988). Some studies have reported stream temperature increases to be a function of the length of stream reach logged (Holtby 43 1988). With longer harvested stream reaches leading to greater effects on downstream temperatures and recovery times. -Effects of Harvesting-While average stream temperatures in forested and harvested streams were not significantly different, increases in peak temperatures and increased heat loss during night were seen in harvested reaches compared to forested reaches. Harvested areas had increased diurnal fluctuations, with peak stream temperatures reaching 7°C higher than peak temperatures in forested streams during the summer months. Increases of 7°C for peak temperatures are similar to those found in other studies within British Columbian small streams following forest harvesting (Feller 1981, Hartman and Scrivener 1990). Removal of forest canopy, leading to increased solar radiation, may increase peak temperatures by 5°C or more (Brown and Binkley 1994). In addition, stream temperatures were significantly lower in above-harvested reaches than in harvested and below-harvested reaches. Harvested reaches exhibited higher daytime stream temperatures than the two forested sections, although the differences were not statistically significant. Statistically significant results may have been found i f temperature readings had been made later in the day when differences are greater. However, due to sampling schedules, temperature data were collected between the hours of 8:30 and 11:00 for each of the study weeks. Temperature trends within the present study are consistent with those reported elsewhere with forest harvesting leading to increased stream temperatures and larger daily fluctuations (Feller 1981, Brown 1985, Holtby 1988, Hetrick et al. 1998b, Johnson and Jones 2000). Temperature increases are primarily a result of increased solar radiation reaching the stream. However, stream temperatures may also be controlled by a variety of other factors, including discharge levels, soil temperatures, ground-water influx, elevation, topography and climate (Ward and Stanford 1982). For example, there was no significant temperature difference between the control streams F6 and F7. Stream F6 has a denser riparian canopy and is a smaller stream with less discharge than F7 that has a less dense canopy. The greater 44 discharge levels of F7 may allow it to dissipate the increased solar radiation it receives. Other characteristics of these watersheds (underlying geology, elevation, topography and climate) were all similar. While increased stream temperatures have been shown to be detrimental to biological processes within most stream ecosystems, logging to the stream banks of forested streams in southeast Alaska, where temperatures are considered cooler than optimum for salmon fry, improved the overall condition of the salmon fry (Thedinga et al. 1989). However, in northern temperate streams, research has shown that increased water temperatures are typically detrimental to fish species where levels above 25°C can be lethal to cold-water fish species (Bjornn and Reiser 1991). While these study streams do not contain fish, temperature increases seen in these streams may still impact other aquatic organisms and biological processes. Possible interactions between temperature levels with periphyton and invertebrates are discussed in the following chapters (Chapter 4 and Chapter 5). Thermal recovery refers to the amount of time a stream takes to return to pre-harvest temperature levels. In this study (3 years post-harvest), vegetation within the harvested areas has not grown back to a level which would provide protection from solar radiation. Thermal recovery is primarily a function of vegetation growth, and temperatures have been found to decrease to pre-harvest levels in streams in southwest British Columbia 7 years following harvesting (Feller 1981). This suggests that temperatures wi l l decrease with regrowth of vegetation in the harvested areas. 3.5 C O N C L U S I O N S Chemistry/Nutrients: No significant differences in dissolved ion concentrations were seen longitudinally in the undisturbed streams. This is either a result of no major change in nutrient cycling processes along the longitudinal gradient studied or a relatively low statistical power to determine potentially small shifts along the longitudinal gradient. While drinking water quality may not 45 be a concern within the present study (3 years post-harvest), shifts in chemical concentrations may still affect biological processes. For example, changes in nutrients considered growth-limiting to aquatic plants (N and P) may lead to increases in primary production (Home and Goldman 1994). While differences between A l and K concentrations were noted between treatment areas, these ions may not be significantly impacting aquatic organisms and primary production as they are not typically limiting for primary production. In addition, while significant differences with time were found, absolute concentrations did not vary drastically over the study period or between streams. These relatively small differences should not be significantly impacting biological processes in the study streams. Solar Radiation: Availability of light is vital for stream ecosystem functions and controls a variety of processes such as stream temperature and primary production. The lack of differences in light levels along the longitudinal gradient of the undisturbed streams is most likely a result of no significant shifts in stream order or stream size, which would lead to a more open canopy and greater light levels reaching the streambed. Harvested streams exhibited greater light levels in the harvested study reaches than in undisturbed forested reaches as a direct result of removal of tree canopy. Light levels would be expected to return to normal with the regrowth of the vegetation in the harvested areas. Temperature: Temperature is also a vital component of the stream ecosystem and controls many of the metabolic processes of aquatic flora and fauna. There were no significant longitudinal changes seen in the undisturbed streams most likely as a result of no changes in canopy cover and stream order/size along the longitudinal gradient studied. Forest harvesting resulted in increased temperatures as a result of removal of tree canopy and increased solar radiation reaching the stream. Any alteration of the temperature regime may impact primary production, decay rates and growth of upper-trophic levels such as invertebrates and fish (Rempel and Carter 1986, Holtby 1988, Johnson and Jones 2000). Buffer-strips are often the 46 preferred solution for protecting stream temperature increases following harvesting of a watershed and have been shown to decrease the effect of canopy removal on stream temperatures (Brown 1985, Brosofske et al. 1997). However, in these headwater stream systems where zero-order streams and temporary channels spread out over the hillside, protecting each of these channels with buffer-strips wide enough to protect the stream system would pose difficulties for management and harvesting. These study streams did not require buffer strip protection because they are classified as S6 under the Forest Practices Code of B C Act and Regulations, since they are not utilized directly for drinking water and do not contain fish. While increased temperatures were found in the harvested areas, downstream recovery began within only 100 meters distance below the harvested area. These results suggest that harvested stream lengths of 250-750 meters were not sufficient to significantly increase downstream temperatures. 47 C H A P T E R 4. P E R I P H Y T O N 4.1 I N T R O D U C T I O N Periphyton, also referred to as biofilm or aufwuchs, is a complex assemblage of algae, bacteria, fungi, dissolved organic matter and midges which grows on the surface of rocks and leaves in the bottom of a stream. While the term periphyton refers only to the algal portion of biofilm, the words will be used interchangeably throughout this chapter. To measure solely for the algal content, one may measure the amount of Chlorophyll a (the most abundant pigment found in plants) within the biofilm. A measure of total biofilm is achieved through a laboratory process of filtering a sample of periphyton, with subsequent ashing to burn away all of the organic matter belonging to the biofilm and finally weighed again to determine ash free dry mass (AFDM). A F D M is a measurement of the total amount of biofilm material for a given area. Primary production (periphyton growth and biomass) in streams is controlled by a variety of factors such as overhead canopy density (light), nutrient levels and grazing pressures. Limiting Factors: -Solar Radiation-Riparian canopy removal has been shown to increase periphyton biomass (ash free dry mass) and algal biomass (Chlorophyll a) in northern temperate streams, primarily due to increases in the total amount of solar radiation reaching the streambed (Hansmann and Phinney 1973, Murphy and Hall 1980, Hetrick et al. 1998a, Kiffney and Bull 2000). Periphyton have been shown to be limited primarily by light levels, as studies from various parts of the world, including New Zealand (Quinn et al. 1997) and the eastern coast of the United States (Noel et al. 1986) have all reported similar relationships between periphyton and light. Forest harvesting, which removes tree canopy above streams, has led to increased light levels reaching streams, and increased periphyton growth (Hansmann and Phinney 1973, Noel et al. 1986, Kiffney and Bull 2000). Periphyton levels have been shown to return to pre-harvest 48 levels once vegetation has regrown and provided sufficient shade to the stream. In the absence of such shading, post-harvesting periphyton growth wi l l remain high independent of the number of years post-harvest (e.g. Hansmann and Phinney 1973 for <1 year post-harvest results; Noel et al. 1986 for 3 year post-harvest data and Lowe et al. 1986 for 6 year post-harvest data). While increased light has been shown to increase periphyton biomass, other abiotic factors may reduce periphyton growth and standing crop in harvested areas. These factors include cell damage to periphyton caused by excessive exposure to light (photo inhibition). Photo inhibition is a plausible explanation for reports of similar periphyton biomass levels in harvested and forested stream reaches (Hil l and Knight 1988, H i l l et al. 1995). Increased water flows and freshets have also been shown to impact periphyton levels by limiting the growing period and causing loss due to scouring (Shortreed and Stockner 1983, Hetrick et al. 1998a). Relatively small floods in Arizona streams eliminated certain algal species which were loosely attached to substrates (i.e. cyanobacteria and chlorophytes), with larger floods leading to the scouring of all types of algae (Grimm and Fisher 1989). Periphyton biomass has also been found to decrease with increasing stream discharge in British Columbia (Kiffney et al. 2000). -Nutrients-Periphyton may also be limited by nutrients, in particular N and P (Shortreed and Stockner 1983, Lowe et al. 1986, H i l l and Knight 1988, Rosemond 1994, H i l l et al. 1995, Rosemond et al. 2000, Kiffney and Richardson 2001). Nitrogen, more specifically in the forms of NO3 and NH4 for aquatic flora/fauna, provides the building blocks for proteins and is often considered to be limiting in stream systems. In addition, P is often considered to be the most common growth-limiting element in aquatic ecosystems (Home and Goldman 1994). Increases in nutrients, either manually or through harvesting activities, enhance periphyton growth in stream channels with open canopies where light conditions are no longer limiting to periphyton growth (Lowe et al. 1986, H i l l and Knight 1988, Kiffney and Richardson 2001). 49 -Grazing Herbivores-Flora and fauna interact in all ecosystems, with plants growing and reproducing while herbivores consume the plant material in order to grow and reproduce as well. Within aquatic systems herbivorous invertebrates consume periphyton as a primary food source. Grazers within the aquatic ecosystem have been shown to alter the structural and functional attributes of periphyton (Hauer and Lamberti 1996). Grazing pressure from invertebrates (i.e. caddisflies (Trichoptera) and mayflies {Ephemeroptera)) as well as other herbivores (i.e. snails) have been identified as a factor controlling periphyton growth and standing crop biomass (Feminella et al. 1989, H i l l and Harvey 1990, Rosemond 1994, H i l l et al. 1995, Rosemond et al. 2000). However, this grazer-periphyton interaction is often dependent upon other abiotic factors including levels of light, temperature, nutrients, stream discharge and time of year, and cannot easily be explained without controlled experiments in the field. In contrast to grazers depressing periphyton growth, periphyton levels may act to limit grazer levels. For example, in streams with increased levels of periphyton, often resulting from increased light and nutrient levels, increases in grazer densities have suggested resource limitation for stream herbivores (Fuller et al. 1986, Hart and Robinson 1990, Kiffney and Richardson 2001) A review of experiments involving periphyton/herbivore interactions showed that out of nearly 100 experiments, more than 70% showed significant effects of periphyton on grazers. The majority of these significant effects resulted when a manipulated decrease in periphyton levels led to decreases in numbers of grazers or when a manipulated increase in periphyton levels led to increased grazer densities (Feminella and Hawkins 1995). Periphyton may also decrease grazer movement as a result of grazer ability to seek out areas rich in food, leading to less emigration and more time spent during feeding in periphyton rich areas (Lamberti and Resh 1983, Kohler 1984, Richards and Minshall 1988, Feminella et al. 1989). 50 Food Webs: Understanding the factors that limit and control the amount of periphyton in the food web is important. However, it is also important to understand how the species and community structure of periphyton is altered by forest canopy removal. Some studies have suggested that increased periphyton biomass may lead to increased numbers of invertebrates and potentially more food supply for fish species (Perrin et al. 1987, Hetrick et al. 1998a). However, it is important to consider the effect of any shifts in the periphyton community assemblage or invertebrates on other stream trophic levels. For example, in areas of riparian canopy removal the assemblage of periphyton often shifts from a community of diatoms to a community of filamentous algae which are not as easily digested by certain invertebrate consumers (Hansmann and Phinney 1973, Lowe et al. 1986). In addition, while there may ultimately be greater periphyton biomass in harvested areas, periphyton in forested areas have been shown to be more efficient, containing more carbon per cell than periphyton in harvested areas (Hill et al. 1995). Changes in periphyton assemblage not only affect local herbivores; algal community shifts may ultimately affect downstream food webs. Periphyton are a significant part of the stream food web, providing the food base for invertebrates, which in turn are fed upon by larger prey species such as fish. The stream ecosystem may be influenced by bottom-up processes with a link between increased periphyton biomass and increases in grazer abundance (Richards and Minshall 1988, Hart and Robinson 1990, Kiffney and Richardson 2001). However, the linkages might not be that straightforward, as other factors may be influencing grazer densities. For example, Kiffney and Bull (2000) found that while periphyton levels increased in harvested headwater streams in British Columbia, there was no effect on invertebrate density, possibly suggesting no bottom-up influence in the stream ecosystems they studied. However, they also suggested that other factors, such as increased sediment levels in the harvested stream, could have led to the decreased invertebrate densities found during the study. In addition, top-down processes showing that herbivores regulate their food resources, suggest that stream ecosystems are not solely controlled by abiotic factors (Peterson et al. 1993, Feminella and Hawkins 1995). This illustrates how stream ecosystems, like terrestrial systems, have a 51 variety of abiotic and biotic factors interacting and influencing the structure and function of local food webs. Objectives: The overall objective of the present study is to determine 1) trends along a longitudinal gradient and 2) the effects of forest harvesting (3 years post-harvest) on periphyton biomass within stream reaches with no riparian canopy. 3) Factors potentially explaining any variation in periphyton biomass, such as incident P A R , NO3 and PO4, were also assessed. 4.2 METHODS Study Area: The study site was located within the Roberts Creek Study Forest (RCSF) on the Sechelt Peninsula, approximately 40 km northwest of Vancouver, British Columbia (Figure 2.1). Six streams were selected for study, three of which had been recently harvested (F4, F5 and F8) and three of which remained undisturbed (F6, F7 and F9). A l l six study streams were located on the relatively uniform Mount Elphinstone slope with a southerly aspect and consist of first and second order creeks draining narrow elongated catchments (Hudson 2001b). A l l study reaches were located between elevations of approximately 400 to 750 meters above sea level and drained slopes of approximately 13-18% (Figure 2.2). Stream size was variable; however, each of the streams averaged between 1-3 meters in width. Logging occurred between fall of 1998 and fall of 1999. The study area is described in more detail in Chapter two. Field Sampling: Periphyton sampling was carried out over a six week period from M a y 24 to July 5, 2002. The research design involved the placement of six unglazed ceramic tiles (each with a surface area of 15cm2) in riffle segments within each of the nine study reaches of the six 52 study streams at the beginning of the sampling period. Tiles were used with the assumption that unglazed ceramic tiles are representative of conditions on natural substrata, as discussed by Kiffney and B u l l (2000) and Rosemond et al. (2000). One tile, selected randomly from each of the riffle segments, was to have been removed at weekly intervals and processed giving a total of 253 samples. However, 71 samples could not be processed due to lack of sufficient streamflow in certain study reaches. Processing of tiles in the field involved placing the tile in a plastic bag, wrapping in aluminum foil to avoid exposure to sunlight, then freezing upon return to the laboratory until further analysis. Laboratory Analysis: In the laboratory, tile surfaces were scraped using a razor blade, scrubbed using a toothbrush and rinsed using distilled water to ensure removal of all periphyton. Each sample was divided in two for determination of ash-free dry mass ( A F D M ) and Chlorophyll a (Hauer and Lamberti 1996). The A F D M sample was filtered through ash-free filters (Millipore Prefilters A P I 5 47mm; Bedford, M A ) that were pre-heated in order to remove any impurities then weighed. After filtering, samples were then oven-dried (70°C) for 24 hours to a constant mass. Once samples were dried, they were cooled in a dessicator to room temperature and weighed (dry mass). The dried filters were then placed into a muffle furnace at 550°C for two hours in order to burn off all organic matter. The samples were removed from the furnace, allowed to cool in a dessicator and re-weighed (ashed mass). Subtracting the ashed mass (inorganic material left over after ashing) from the dry mass (dried sample) and initial filter mass provided the total A F D M of organic matter within the samples. Dividing the amount of organic matter (pg) by the surface area of the tile (15cm2) provided a measurement of A F D M as pg c m - 2 . For Chlorophyll a the sample was filtered through filters (Millipore Prefilters A P I 5 25mm; Bedford, M A ) . This was followed with extraction of the Chlorophyll a pigments using 10ml of a 90% acetone solution. The filters were soaked in the acetone solution for 24 hours in a darkened refrigerator. Approximately 30 minutes prior to testing, the containers were removed from the refrigerator, shaken and allowed to settle before pouring off the solution 53 into a cuvette. Chlorophyll a levels (ug cm" 2) in each cuvette were measured using a fluorometer (TD-700 Laboratory Fluorometer - Turner Designs; Sunnyvale, C A ) . Statistical Analyses: Objective 1: To test for any changes along the longitudinal gradient. The statistical analysis for longitudinal trends on the undisturbed streams was analyzed as a General Linear Model ( G L M ) analysis of variance ( A N O V A ) for a split-plot design using the statistical program Systat 9.0 ( S Y S T A T 1998) (Table 3.1). Objective 2: To test for the effects of forest harvesting (3 years post-harvest) on periphyton biomass within stream reaches with no riparian canopy. For the analysis of differences in nutrient concentrations and stream water temperatures between harvested and forested sites, I applied a G L M A N O V A for a randomized nested split-plot design using the statistical program Systat 9.0 ( S Y S T A T 1998). The General Linear Model ( G L M ) was used due to missing values in the data set. The experimental design was nested because reaches were treated as being nested within the treatments and streams. This accounts for the fact that reach 1 on harvested stream F4 is not the same as reach 1 on harvested stream F5, so the different reaches cannot be treated directly as replicates during analysis. The split-plot design was utilized because of the added effect of time (weeks) on the design (Table 3.2). When the above statistical analyses showed significant differences for effects, Tukey's multiple comparison test was used to determine specific differences. Objective 3: To determine the relationships between research variables. Regression analysis was conducted to determine any relationships between periphyton ( A F D M and Chlorophyll a) and the physical and chemical parameters measured such as incident P A R , NO3 and PO4. Untransformed data was first plotted to see relationships, then different regression models were tested to see which model had the highest coefficient of 54 determination (Rz) and the lowest standard error of the estimate, and was significant at P < 0.05. To account for instances of non-normality within the data sets and heterogeneity of variances, data were transformed when necessary using Logio (Xj + 1). Data are presented as untransformed means ± 1 standard error. 4.3 R E S U L T S Longitudinal Trends: No significant differences in A F D M were found along the longitudinal gradient in undisturbed streams (Appendix 5.1) (Figure 4.1). Chlorophyll a did not differ significantly between reaches along the longitudinal gradient in the undisturbed streams (Appendix 5.2) (Figure 4.2), yet did exhibit differences with time, with levels in weeks 1 and 2 being significantly lower than in all other weeks and levels in weeks 3 and 4 being significantly lower than in weeks 5 and 6. However, Chlorophyll a levels did not differ significantly between weeks 3 and 4 and between weeks 5 and 6 (Figure 4.3). A F D M 500 -i 400 300 E o O ) 200 100 5 6 Reaches Figure 4.1 Average A F D M values + 1 standard error for the undisturbed streams, (averaged from data during six-week study period). 55 Chlorophyll a 100 -i 80 E o CD 3. 60 40 -\ 20 I _I_ 4 5 6 Reaches Figure 4.2 Average Chlorophyll a values + 1 standard error for the undisturbed streams, (averaged from data during six-week study period). Ch lo rophy l l a 120 100 H E o O) 3. Figure4.3 Average Chlorophyll a concentrations ± 1 standard error for undisturbed streams over the six-week study period. Data points with different letters are significantly different at P < 0.05 (Tukey's Test). 56 Effects of Forest Harvesting: Overall results revealed that A F D M and Chlorophyll a levels over the six week study period (May-July 2002) differed significantly between harvested reaches and forested reaches. Average A F D M levels were 2.5 times greater in harvested reaches than in undisturbed reaches and average Chlorophyll a concentrations (reflecting the algal component) were 4.5 times greater in the harvested reaches (Table 4.1). Table 4.1 Average A F D M and Chlorophyll a values in the harvested streams. (AFDM and Chlorophyll a ± 1 Standard Error). Study Area n = streams x reaches x samples AFDM (us/cm2) Chlorophyll a (ue/cm2) Above Harvested 39 197.83 ±47.63 25.76 ±6.16 Harvested 50 485.23 ±65.57 129.61 ±24.26 Below Harvested 50 246.18 ±54.69 32.65 ±5.21 - A F D M -A F D M differed significantly between harvested reaches and above and below-harvested reaches in the harvested streams (Appendix 5.3) (Figure 4.4). In the harvested streams A F D M also exhibited significant differences with time, with levels in week 1 significantly lower than levels in weeks 3, 4, 5 and 6 (Figure 4.5). There were no significant differences between reaches within the harvested or forested sections (Figure 4.6). 57 A F D M 600 500 400 E o 300 2 200 100 Above-Harvested Harvested Below-Harvested Figure 4.4 Average A F D M values + 1 standard error for the harvested streams, (averaged from data during the six-week study period). Bars with different letters are significantly different at P < 0.05 (Tukey's Test). A F D M 600 i i i 1 1 r 1 2 3 4 5 6 Weeks Figure 4.5 Average A F D M values ± 1 standard error for the harvested streams over the six week study period. Data points with different letters are significantly different at P< 0.05 (Tukey's Test). 58 A F D M 800 600 o 400 3. 200 1 2 3 A b o v e - H a rves ted 4 5 6 Harves ted R e a c h e s B e l o w - H a r v e s t e d Figure 4.6 Average A F D M levels + 1 standard error for harvested streams. Reaches 1-3 are above-harvested reaches (forested), reaches 4-6 are harvested and reaches 7-9 are below-harvested reaches (forested). -Chlorophyll a-Chlorophyll a levels within harvested streams differed significantly between the harvested reaches and the above and be low-harvested reaches (Appendix 5.4) (Figure 4.7). Chlorophyll a levels exhibited significant increases with time, with weeks 1, 2 and 3 having lower levels compared to weeks 5 and 6 in the study. However, no significant differences were noted between weeks 4, 5 and 6 (Figure 4.8). Chlorophyll a amounts did not differ significantly between reaches within treatments. The large difference in Chlorophyll a levels between reach 4 and 6 was not significantly different probably due to low statistical power (Figure 4.9). 59 Ch lo rophy l l a 180 E o Dl =i. Above-Harvested Harvested Below-Harvested Figure 4.7 Average Chlorophyll a concentrations + 1 standard error for the harvested streams (averaged from data during the six-week study period). Bars with different letters are significantly different at P < 0.05 (Tukey's Test). Chlorophy l l a 200 -180 -160 -140 -120 -CM E 100 -o cn 80 -60 -40 -20 -0 -Figure 4.8 Average Chlorophyll a concentrations ± 1 standard error for harvested streams over the six-week study period. Data points with different letters are significantly different at P < 0.05 (Tukey's Test). 60 Ch lo rophy l l a 250 E o O) 200 -\ 150 100 50 1 2 3 Above-Harvested I 4 5 6 Harvested Reaches 7 8 9 Below-Harvested Figure 4.9 Average Chlorophyll a levels + 1 standard error for harvested streams. Reaches 1-3 are above-harvested reaches (forested), reaches 4-6 are harvested and reaches 7-9 are below-harvested reaches (forested). Relationships between Periphyton and Stream Characteristics: Regression analyses revealed that percent incident P A R explained 29% of the variation in A F D M and 38% of variation in Chlorophyll a (Table 4.2). Both of the regressions revealed significantly positive relationships. However, these relationships were not well defined with A F D M and Chlorophyll a both exhibiting relatively large standard errors of the estimate. The relatively large standard errors are a result of the high variability within the data set and a poor relationship between the variables. No significant relationships were found between dissolved nutrient concentrations and periphyton biomass. Nitrate and PO4 concentrations were not significantly different between sections in the harvested streams (Table 4.3 - refer to Chapter 3 for complete chemistry results). 61 Table 4.2 Regressions with periphyton measures (AFDM and Chlorophyll a) as a function of percent incident photosynthetically active radiation (PAR), (n = 54) Regression Equation Coefficient of Determination R 2 Standard Error of the Estimate SEE P Value AFDM = 192.80 + 2.73 (PAR) 0.29 158.7 <0.001 Chlorophyll a = 25.95 + 0.95 (PAR) 0.38 45.97 <0.001 Table 4.3 Dissolved nutrient concentrations (mg/L) ± 1 standard error in treatment areas of the harvested study streams. Treatment Area Nitrate Phosphorous Above-harvested 0.17+0.03 0.006 ± 0.002 Harvested 0.17+0.02 0.005 + 0.000 Below Harvested 0.24 ± 0.03 0.006 ± 0.001 4.4 DISCUSSION Forest canopy removal across riparian zones often leads to increased solar radiation with subsequent increases in periphyton (Hansmann and Phinney 1973, Noel et al. 1986, Kiffney and B u l l 2000). The results of the present study revealed that periphyton biomass levels, measured as A F D M and Chlorophyll a, were higher in harvested reaches than in undisturbed reaches within these headwater study streams. While periphyton mass was significantly different between harvested and forested reaches in the harvested streams, it did not vary significantly along a longitudinal gradient between reaches in the undisturbed streams. Although the longitudinal changes were not significant, average Chlorophyll a levels on 62 harvested streams decreased as the streams flowed through the forest downstream of the harvested area. This trend was not seen in A F D M as the data exhibited more variability between reaches. Regression analyses indicated both Chlorophyll a and A F D M were more closely related to P A R light readings than to the other stream characteristics assessed, suggesting that light may be a primary controlling factor for periphyton biomass in these headwater streams. However nutrients were not fully assessed and their significance as a controlling factor on periphyton biomass cannot be fully determined. Longitudinal Trends: Neither A F D M nor Chlorophyll a exhibited significant changes along the longitudinal gradient of the undisturbed streams (a distance of 600 meters of stream length from the uppermost reach (#1) to the furthest downstream reach (#9)). Variation between these reaches seemed to be a result of natural variability between specific habitats within the study streams. For example, reach 3 on the undisturbed stream F7 had an average Chlorophyll a level of 49.5 pg c m - 2 as a result of an open overhead canopy, while reach 7 had a Chlorophyll a level of 12.1 pg c n r 2 with a denser overhead canopy limiting growth of algae with lowered light levels. However, these differences did not follow any longitudinal trends and were not statistically significant. This is not surprising since periphyton have been shown to depend primarily on light with consistent longitudinal shifts seen only along greater distances as stream size increases substantially and canopy naturally opens (i.e. the shift from stream order 1 to stream order 5) (Vannote et al. 1980). For harvested streams, differences between periphyton levels in the three reaches within a section (i.e. within the harvested area, above-harvested and/or below-harvested area) were not statistically significant. However, Chlorophyll a within the forested reaches below the harvested area decreased from an average of 45 pg c m - 2 in the reach closest to the harvest boundary to an average of 21 pg c n r 2 at the reach located 150 meters into the forested section. Thus, once the stream had flowed back into a forested area, Chlorophyll a levels returned to those found in the upstream-forested reaches. This suggests an influence of decreasing light levels as the stream flows back into more densely forested sections and that 63 algal biomass returned to upstream conditions after only 50 meters downstream of the harvested boundary. This is most easily explained by the amount of light available at varying distances from the harvested boundary, with decreasing light available for algal growth as the stream flows further into the forested area and away from the harvested areas. Increased shading caused by increases in riparian canopy density results in lowered levels of periphyton when compared to areas with higher light levels (Hi l l and Knight 1988, Kiffney and Bu l l 2000) and Kiffney et al. (2003) found that A F D M and Chlorophyll a in small streams in southwest British Columbia decreased as width of riparian forested buffer strips increased. Effects of Harvesting: Both measures of periphyton - A F D M (biofilm) and Chlorophyll a (algal component of biofilm) - showed that periphyton biomass was greater within the harvested stream reaches than in upstream and downstream forested reaches in the study streams. However, in contrast, a study in Carnation Creek on Vancouver Island, British Columbia, showed that periphyton levels did not increase following forest harvesting. This finding was attributed to an absence of a significant increase in phosphorus levels following logging which may have limited growth of periphyton (Shortreed and Stockner 1983). However, the present study found no significant difference in phosphorus levels between harvested and undisturbed reaches (Chapter 3), yet also found significant differences in A F D M and Chlorophyll a between harvested and undisturbed areas. It is important to note that the results of the present study are for one season (summer) during low flow conditions in headwater stream systems three years post-harvest. Periphyton biomass was greater in harvested streams 2-3 years following forest harvesting in New England streams (Noel et al. 1986) while a study in the state of Washington showed greater levels of periphyton 7 years after harvesting (Bilby and Bisson 1992). Increased periphyton levels were even found in Oregon streams up to 17 years after logging, with elevated rates of growth declining within the first 10-20 years (Murphy and Hal l 1980). Such increased periphyton growth has been related primarily to increased light levels and water temperature 64 following tree removal. Water temperatures up to 36°C were found to enhance autotrophic production in experimental outdoor channels in Ontario (Rempel and Carter 1986); however, such high temperatures may ultimately damage other aquatic organisms in actual stream settings. As the riparian canopy recovers and autotrophic production slows, measures of periphyton biomass have been shown to return to levels found in forested streams (Murphy and Hal l 1980). While studies in various geographical locations have reported similar trends in periphyton growth and limitation, growth limiting factors may differ with season. For example, in a woodland stream in Tennessee over the course of two years nutrients were limiting to periphyton growth in summer and fall, while light was limiting in winter and spring (Rosemond et al. 2000). Seasonal variation in small streams in British Columbia showed Chlorophyll a accumulation rates reaching a low of .31 pg/cm/day in January and a high of 3.6 pg/cm/day during May (Kiffney et al. 2000). Seasonal differences were similar in Oregon streams with peak production occurring during spring/summer and explained by increased light levels and temperature combined with lower more stable discharge levels allowing periphyton to become easily established on stream substrates (Bilby and Bisson 1992). Due to such seasonal and annual differences, the results of the present study should not be extrapolated or compared to conditions found in other seasons or in study sites where vegetation has had sufficient time to recover and provide shade to the streams. Differences between seasons and riparian canopy types may ultimately lead to different interactions and processes between the abiotic and biotic conditions which would alter periphyton production. Relationships between Periphyton and Stream Characteristics: Regression analyses of A F D M and Chlorophyll a levels as a function of streamwater NO3 and PO4 concentrations revealed no significant relationships. These results are consistent with the statistical analyses showing no significant differences or trends among streams or treatments in NO3 and PO4 concentrations during the study period (Table 4.2). Harvesting induced nutrient changes have been shown to be greatest within the first 2 years and decreasing in subsequent years (Feller and Kimmins 1984). Data in the present study were 65 collected three years post-harvesting when harvesting-induced changes may not be seen. Research regarding effects of harvesting on nutrient levels should typically include pre-harvest data to allow definitive conclusions to be made. In addition, the present study occurred during a period with few rainfall events which would act to flush nutrients into the stream, helping to reveal any difference between treatments. While there was no relationship of A F D M and Chlorophyll a to stream nutrient concentrations, studies 1-2 years post-harvesting and studies which have experimentally increased nutrient levels (N and P) have shown significant relationships between nutrient concentrations and periphyton levels, with stronger effects occurring once light is no longer limiting within the stream ecosystem (Shortreed and Stockner 1983, Lowe et al. 1986, Perrin et al. 1987, H i l l and Knight 1988, Rosemond 1994, H i l l et al. 1995, Rosemond et al. 2000). Regression analyses with A F D M and Chlorophyll a as a function of percent incident P A R and other variables suggested that periphyton biomass may be primarily controlled by light within the study streams. While primary production was not directly measured, greater biomass in harvested reaches suggested increased primary production as a result of increased light availability. These results suggest that light may be a primary limiting factor; however, nutrients were not fully assessed and their significance as a controlling factor on periphyton growth and biomass cannot be determined. 4.5 C O N C L U S I O N S Longitudinal Trends: Longitudinal gradient changes along the undisturbed streams were not significant. No significant differences were found between the three reaches of the different sections (i.e. reaches within the harvested area and above and below areas); however, Chlorophyll a levels decreased with increasing distance downstream of the harvested area. N o trends were found for A F D M , however. This suggested that the algal portion of periphyton biomass is capable of returning to undisturbed forest levels below harvested sites within the first 100 meters in 66 these densely forested headwater streams. However, these results should not be extrapolated to streams where forest canopy may be less dense than those found in the study streams. Effects of Harvesting and Relationships between Periphyton and Stream Characteristics: Both measures of periphyton, A F D M (biofilm) and Chlorophyll a (algae), showed that periphyton levels within the harvested stream reaches were at least twice the levels in the forested reaches. Regression analyses suggested that greater periphyton biomass in harvested areas was a result of increased light levels within harvested reaches which is consistent with previous studies suggesting that in small headwater streams with a dense riparian canopy, light is the primary limiting factor for periphyton growth (Kiffney and B u l l 2000). 67 C H A P T E R 5. I N V E R T E B R A T E S 5.1 I N T R O D U C T I O N While all species within the aquatic ecosystem are important to stream processes, macro invertebrates have been extensively studied because of their important role in ecosystem productivity and their use as sensitive indicators of water quality (Roby and Azuma 1995, Minshall et al. 1997, Giller and Malmqvist 1998). Benthic macro invertebrates are typically chosen for research because they are abundant and have relatively long aquatic life stages, providing researchers with a good basis for studying changes and disturbances in aquatic ecosystems (Roby and Azuma 1995). For example, disturbances caused from the removal of forest canopy within riparian zones often cause significant changes in both abundance of invertebrates and diversity of species (Swank and Crossley 1988). Macro invertebrates have been studied extensively both because of their wide distribution and for their importance in the trophic food web as a primary food source for commercial and sport fisheries (Merritt and Cummins 1996). Aquatic insects and other invertebrates have adapted to life in streams in a variety of ways, with different groups flourishing in their respective habitats. Functional feeding groups have been utilized as a way of defining the different adaptations of aquatic insects in various stream habitats. Aquatic insects are placed into functional groups based upon similar feeding mechanisms and food resources (Hauer and Lamberti 1996). For example, shredders feed on woody debris and other allochthonous material, collectors filter fine particulate organic matter, scrapers graze on periphyton, piercers suck fluids from tissues and cells, predators attack prey and eat other insects while parasites live off of the tissue of hosts (Merritt and Cummins 1996). Therefore, shredders may be more abundant in areas with greater inputs of allochthonous material, while scrapers may be more abundant in areas with higher rates of periphyton production. Organic matter found within a stream system originates from the surrounding terrestrial environment and is transported into the stream either by overland flow, wind or from litterfall 68 (Hauer and Lamberti 1996). Material entering the stream from terrestrial sources is often referred to as allochthonous material or detritus. Breakdown of organic matter within a stream depends primarily on stream temperature, with increasing temperatures leading to increases in microbial activity causing increased decay rates, as well as consumption by invertebrates as a source of food (Rempel and Carter 1986). Allochthonous input from riparian vegetation is an important energy source for primary production and a food source for many invertebrates (Hansmann and Phinney 1973, Feminella et al. 1989, Wallace et al. 1997, Hetrick et al. 1998a). Amount and type of streamside vegetation directly affects how much organic matter enters the stream as well as the rates of processing (Gregory et al. 1991). Stream processes are directly coupled with the terrestrial landscape and because of this, any activities occurring in the riparian zone may influence organic matter budgets and stream structure (Hauer and Lamberti 1996). Effects of Harvesting: Forest harvesting leads to changes in abundance and community composition of invertebrates due to a variety of factors including increased light and nutrient levels causing increased rates of primary production (periphyton) (Wootton and Power 1993, Stone and Wallace 1998); increased water temperatures leading to increased breakdown of organic material and direct influences on growth and emergence of invertebrates (Vannote and Sweeney 1980, Noel et al. 1986); changes to sources of allochthonous debris which may alter food sources (Silsbee and Larson 1983, Stone and Wallace 1998); as well as changes in fine-sediment levels within streams (Shortreed and Stockner 1983, Zweig and Rabeni 2001). The above factors all may affect invertebrates either directly, by affecting growth and emergence, as well as indirectly through increased fine sediments affecting habitat. In addition, increases in primary production and changes to allochthonous inputs affect food quality and abundance. Forest harvesting may alter each of these factors to different extents. Harvesting effects on local communities of aquatic invertebrates wi l l depend on which factors are important within the ecosystem and which are significantly altered. 69 Removal of the riparian canopy above small forested streams often leads to increases in invertebrate densities as a result of increases in primary production (Murphy and Hal l 1980, Hawkins et al. 1982). Increases in lower trophic levels could ultimately lead to increased food supply for fish species and other predators (Hetrick et al. 1998a). However, increases in water temperature and fine sediment that often accompany riparian forest canopy removal may nullify any benefits from the increased food production (Murphy and Meehan 1991). In contrast, some studies have found decreases in invertebrate densities following forest harvesting as a result of increases in sedimentation affecting invertebrate habitat and food resources (Swank and Crossley 1988, Hartman and Scrivener 1990). While macro invertebrate density and biomass are often measured in field studies, community composition is also important, as increases or decreases in certain species may be indicative of changes in habitat structure and quality. Changes in periphyton biomass, such as increases in diatoms and filamentous algae caused by logging, often lead to changes in invertebrate community composition such as increases in invertebrate species which feed primarily on algae (i.e. scrapers & grazers) (Anderson 1992, Stone and Wallace 1998). Increases or decreases of allochthonous inputs have also affected invertebrate species specialized in eating detritus as their main food source. Studies in small streams in Carnation Creek, British Columbia, showed increased densities of Ephemeroptera in areas with higher levels of detritus, as well as increases in Chironomidae which are both families of invertebrates with species specializing in the consumption of allochthonous material (i.e. shredders) (Hartman and Scrivener 1990). In particular, one particular group of Ephemeroptera called Baetis (generalists that are able to inhabit many habitat types), often accounted for a major proportion of increases in harvested sites (Noel et al. 1986). Forest harvesting may often directly reduce the amount of allochthonous input (e.g. leaf litter and L W D ) entering the stream due to removal of the vegetation source. In small streams where the riparian zone and stream channel are tightly coupled, riparian vegetation inputs play a major role in the function and structure of the stream food web (Hauer and Lamberti 1996). Allochthonous inputs have often decreased within stream sections that have been harvested (Silsbee and Larson 1983, Bi lby and Bisson 1992) and have been accompanied by 70 increased export of dissolved solids (organic matter) from the local stream system (Likens et al. 1970). For example, Silsbee and Larson (1983) showed that even 45 years after initial logging, forested streams had nearly four times as much large woody debris (LWD) as previously harvested streams. There are some exceptions to these findings however. Carlson et al. 1990 found no difference in levels of L W D between previously harvested streams and forested streams in northeast Oregon 6-17 years post-harvest and Hetrick et al. 1998a found decreased allochthonous inputs with no significant change to benthic organic matter in a stream in southeast Alaska. However, since woody debris is continuously broken down and transported within a stream system, forest harvesting which removed sources of L W D w i l l eventually lead to reductions in the total amount of L W D within a stream (Fausch and Northcote 1992). Allochthonous inputs are vital to the stream system, providing structure and protection for aquatic fauna, as well as food for primary producers and some invertebrate species (Bilby and Bisson 1992, Wallace et al. 1997). Objectives: The general objective of the present study is to determine 1) the changes in benthic macro invertebrate density and diversity and benthic organic matter along the longitudinal gradient 2) the effects of forest harvesting (3 years post-harvest) on benthic macro invertebrate density and diversity and benthic organic matter within stream reaches with no riparian canopy, and 3) any relationships between benthic macro invertebrate density and factors potentially influencing densities, such as periphyton biomass ( A F D M and Chlorophyll a), benthic organic matter, N O 3 , PO4 and streamwater temperature. 5.2 METHODS Study Area: The study site was located within the Roberts Creek Study Forest (RCSF) on the Sechelt Peninsula, approximately 40 km northwest of Vancouver, British Columbia (Figure 2.1). Six 71 streams were selected for study, three of which had been recently harvested (F4, F5 and F8) and three of which remained undisturbed (F6, F7 and F9). A l l six study streams were located on the relatively uniform Mount Elphinstone slope with a southerly aspect and consist of first and second order creeks draining narrow elongated catchments (Hudson 2001b). A l l study reaches were located between elevations of approximately 400 to 750 meters above sea level and drained slopes of approximately 13-18% (Figure 2.2). Stream size was variable; however, each of the streams averaged between 1-3 meters in width. Logging occurred between fall of 1998 and fall of 1999. The study area is described in more detail in Chapter two. Field Sampling: -Benthic macro invertebrates-Benthic macro invertebrates were collected over a three-day period from June 12-14, 2002. Three replicate samples were collected from riffles in each of the nine sampling sites per stream (27 samples per stream) with the exception of dry reaches where no samples could be taken. Due to insufficient stream flow in many reaches, only the following streams and reaches were able to be sampled: stream F4 and F5 harvested and below-harvested reaches (reaches 4-9), stream F8 only harvested reaches (reaches 4-6) and the undisturbed reaches of stream F7 (reaches 1-9). A Surber sampler (mesh size 500 um) was utilized for the collection of macroinvertebrate samples. The area within the Surber Sampler grid (930 cm 2) was disturbed with a hand-held garden rake for approximately one minute with larger rocks rubbed by hand to ensure removal of any clinging invertebrates. Processing in the field involved filtering samples through a small sieve (mesh size 250 pm) in order to remove invertebrates rinsed into the end of the Surber sampler net. Any organisms still attached to the mesh of the Surber sampler were hand removed with forceps to ensure no individuals were lost. Invertebrates were then washed into a 500mL plastic container with stream water and preserved using a formalin solution (30%). 72 -Benthic Organic Matter-Benthic organic matter washed into the Surber sampler net during benthic macro invertebrate sampling was collected and transferred to the 500mL plastic container where it was preserved with the invertebrate sample. Laboratory Analysis: In the laboratory, invertebrates were separated from organic matter under a dissecting microscope and stored in ethanol (70%) for preservation. After sorting, invertebrates were identified primarily to the order and/or genus level (Merritt and Cummins 1996). Benthic organic matter separated from the invertebrates was placed in porcelain crucibles and dried at 70°C for 24 hours to a constant mass. After drying, samples were cooled to room temperature and weighed (dry mass). Crucibles were then placed into a muffle furnace at 550°C for two hours, cooled, then re-weighed (ashed mass). Subtracting the ashed mass from the dry mass provided the total ash free organic matter dry mass ( A F D M ) which was converted to a mass per m . Statistical Analyses: Analysis of Longitudinal Trends: -Invertebrate Density/Richness-For taxonomic richness, rarefaction was applied to the data in order to standardize samples of different sizes and account for differences in sampling intensity between samples and sites. Rarefaction is a statistical procedure that estimates the number of taxa expected in a random sample of individuals from a population (Krebs 1999). After rarefaction, taxonomic richness values were tested using analysis of variance. The undisturbed stream (only stream F7 had sufficient data due to lack of streamflow), with nine reaches along the longitudinal gradient, 73 was split up into three regions each containing three sampled reaches (upper, middle, lower regions) in order to test for any upstream differences in invertebrate density and richness along the longitudinal gradient. A one-way analysis of variance was utilized with regions as the main factor being tested. -Invertebrate Diversity-Analysis of community diversity involved the use of nonparametric measures of heterogeneity that take into account both species richness and abundance (evenness) for the calculation of index values (Magurran 1988). Simpson's reciprocal index was utilized for the data in the present study as it is more sensitive to abundances of commonly occurring taxa (Krebs 1999). Analysis of variance was used to identify significant differences between the three regions of the undisturbed stream (F7). -Benthic Organic Matter-Analysis of variance was used to identify significant differences between the three regions of the undisturbed stream (F7). Analysis of Harvesting Effects: -Invertebrate Density/Richness-Analysis of differences between harvested and forested reaches in invertebrate density and taxonomic richness (following rarefaction) was carried out with a General Linear Model ( G L M ) analysis of variance ( A N O V A ) using the statistical program Systat 9.0 ( S Y S T A T 1996). G L M was used due to missing values in the data set. Invertebrate densities at different taxonomic levels were tested for differences between treatments (harvested vs. below-harvested), reaches, and streams. When the A N O V A showed significant differences for effects, Tukey's multiple comparison test was used to determine specific differences. 74 -Invertebrate Diversity-The calculated Simpson's reciprocal index values were analyzed using G L M analysis of variance as described in the previous paragraph. -Benthic Organic Matter-Differences between harvested and forested reaches in organic matter (ash free dry mass) were determined using G L M analysis of variance as described above. Analysis of Relationships: Regression analysis was used to determine any relationships between benthic macro invertebrate densities and study response variables (periphyton biomass ( A F D M and Chlorophyll a), benthic organic matter, streamwater NO3 and PO4 concentrations and streamwater temperature) as described in Chapter 4 (page 54 and 55). Best-fit regression equations were selected on the basis of lowest standard error of the estimate and highest coefficient of determination (R ). To account for instances of non-normality within the data and heterogeneity of variances, data were (Logio) transformed. 5.3 RESULTS Longitudinal Trends: -Invertebrate Density-Invertebrate densities in the undisturbed stream F7 did not differ significantly between the three regions and average invertebrate densities were nearly equivalent to levels found in the below-harvested reaches on the harvested streams (Table 5.1) (Appendix 6.1). 75 Further analysis of invertebrate densities within the various order levels revealed that there were no significant differences in Ephemeroptera (mayflies) or Plecoptera density along the longitudinal gradient of the undisturbed stream F7 (Table 5.1) (Appendix 6.2 and 6.3). Diptera densities were significantly different between regions within the undisturbed stream F7 with middle reaches having lower densities than both the upper and lower reaches (Table 5.1) (Appendix 6.4). However, absolute differences were small and probably not biologically significant. Table 5.1 Invertebrate densities (number/m2 (standard error)) in the three regions of the undisturbed stream F7. Different letters indicate significant differences at P < 0.05 (Tukey's test). Reqion Total Invertebrates Ephemeroptera Plecoptera Diptera Upper 497 (112) 194 (35) 111 (14) 19 (6)a Middle 316 (40) 134 (22) 91 (21) 4 (2)b Lower 318 (72) 114 (26) 90 (32) 25 (6)a -Invertebrate Diversity-The undisturbed stream exhibited no significant differences in diversity at the order level along the longitudinal gradient (Table 5.2) (Appendix 7.1). Ephemeroptera diversity was significantly greater in upper reaches than in middle reaches, while lower reaches fell between the other two regions and did not differ significantly in diversity from either of them (Table 5.2) (Appendix 7.2). Plecoptera diversity differed significantly between regions with middle reaches having a lower diversity than lower reaches; however, differences were small and probably not biologically significant (Appendix 7.3) (Table 5.2). Diptera genera diversity did not differ significantly between regions within the undisturbed stream F7 (Appendix 7.4) (Table 5.2). 76 Table 5.2 Invertebrate diversity (Simpson's Reciprocal Index (standard error)) in the three regions of the undisturbed stream F7. Different letters indicate significant differences at P < 0.05 (Tukey's test). Reqion Order Level EohemeroDtera Plecoptera Diotera Upper 3.5 (0.3) 2.5 (0.1 )a 1.3 (0.1 ) a b 1.3(0.8) Middle 3.5 (0.1) 1.7(0.2)b 1.1 (0.0)b 0.6 (0.3) Lower 3.9 (0.5) 1.8 (0.3)ab 1.6 (0.2)a 1.0(0.8) -Benthic Organic Matter-Ash free dry mass ( A F D M ) did not differ significantly between regions along the longitudinal gradient on the undisturbed stream F7 (Appendix 7.5) (Figure 5.1). Upper Middle Lower Figure 5.1 Ash free dry mass (AFDM) + 1 standard error for the regions within the undisturbed stream F7. 77 Effects of Harvesting: -Invertebrate Density-Total invertebrate density did not differ significantly between harvested and forested areas; however, invertebrate density exhibited a significant interaction between streams and treatment (Appendix 8.1). This was a result of greater differences between harvested and below harvested reaches for stream F4 but not for stream F5 (Figure 5.2). 1400 1200 1000 800 E c 600 -I 400 200 1 • , . r — . , , , , , , , r F4-Harvest F4-Forest F5-Harvest F5-Forest F8-Harvest F7-Undisturbed Figure 5.2 Average invertebrate densities + 1 standard error for treatment areas on harvested streams and averaged for all regions on the undisturbed stream (averages taken from three replicate samples per reach). Bars with different letters are significantly different at P<0.05 (Tukey's Test). Further analysis of invertebrate densities within the various order levels revealed that there were no significant differences in Ephemeroptera density within the harvested streams (Appendix 8.2). Plecoptera densities exhibited significantly greater densities in harvested reaches than in below-harvested reaches (Appendix 8.3) (Figure 5.3). Greater Diptera densities were found in harvested than in below-harvested areas (Figure 5.4). However, significant interactions for Diptera densities were found between streams and treatments as 78 well as between streams and reaches within streams (Appendix 8.4) (Figure 5.5). Densities in the uppermost harvested reach were significantly higher in stream F4 than in streams F5 and F8, while densities in the lowest harvested reach were significantly higher in stream F8 than in all the other harvested reaches within streams F4 and F5 (Figure 5.5). There was no significant difference between the three reaches within a treatment (harvested or below-harvested area) for streams F4 and F5, but there was for stream F8 which had lower Diptera densities in the uppermost harvested reach and higher densities compared to streams F4 and F5 in the lowermost harvested reach (primarily consisting of Chironomidae) (Figure 5.5). 250 200 E 150 CD E 1 100 H 50 Harvest Be low-Harvest Figure 5.3 Average Plecoptera densities + 1 standard error for the harvested streams (averages taken from three replicate samples per reach). Bars with different letters are significantly different at P<0.05 (Tukey's Test). 79 350 Harvest Below-Harvest Figure 5.4 Average Diptera densities + 1 standard error for the harvested streams (averages taken from three replicate samples per reach). Bars with different letters are significantly different at P<0.05 (Tukey's Test). 6 0 0 4 0 0 2 0 0 4 5 6 H a r v e s t e d R e a c h e s B e lo w - H a r v e s te d R e a c h e s Figure 5.5 Average Diptera densities + 1 standard error for the harvested streams with reaches 4-6 harvested and 7-9 below-harvested (averages taken from three replicate samples per reach). Bars with different letters are significantly different at P < 0.05 (Tukey's Test). 80 -Invertebrate Richness-Rarefaction on order level invertebrate data showed significant differences between the harvested and below-harvested reaches for order level richness. Greater richness (# of orders/treatment) was found in the below-harvested than in the harvested reaches (Appendix 9.1) (Figure 5.6). Similar analyses were carried out for Ephemeroptera, Plecoptera and Diptera richness at the family level. Ephemeroptera richness exhibited a significant stream by treatment interaction, with overall greater richness in the harvested reaches of stream F5 than in the harvested reaches of stream F4. However, which streams by treatments were significantly different could not be determined (Appendix 9.2) (Figure 5.7). Neither Plecoptera nor Diptera richness (family level) differed significantly between treatments in the harvested streams (Appendices 9.3 and 9.4 respectively) (Figure 5.8 and 5.9 respectively). 7 CD CD 2 -> CO 1 -o -I 1 1 1 Harvest Be low-Harves t Figure 5.6 Average number of orders + 1 standard error for the treatment areas within the harvested streams (averages taken from three replicate samples per reach). Bars with different letters are significantly different at P<0.05 (Tukey's Test). 81 Figure 5.7 4 5 6 H a r v e s t e d R e a c h e s 7 8 9 B e l o w - H a r v e s t e d R e a c h e s Average number of Ephemeroptera Families + 1 standard error for reaches within each of the harvested streams with reaches 4-6 harvested and reaches 7-9 below-harvested (averages taken from three replicate samples per reach). 4 C/5 a) 3 E 03 CD _Q E 3 C OJ CT> CD k_ CD > CD 2 1 1 X Harvest Below-Harvest Figure 5.8 Average number of Plecoptera families + 1 standard error for the treatment areas within the harvested streams (averages taken from three replicate samples per reach). 82 CO 1 CO CD E =3 C CD CD ca CD > ca Harvest Below-Harvest Figure 5.9 Average number of Diptera families + 1 standard error for the treatment areas within each of the harvested streams (averages taken from three replicate samples per reach). -Invertebrate Diversity-Invertebrate diversity at the order level, differed significantly between streams, but not between the harvested and below-harvested reaches (Figure 5.10). However, there were significant differences between reaches within treatments, as well as a significant treatment by stream interaction (Appendix 10.1). The within treatment diversity was significantly lower in reach 6 than in reach 4 for stream F8 and significantly lower in reach 9 than in reaches 7 and 8 within the forested area of stream F4 (Figure 5.11). The significant stream by treatment interaction revealed that stream F5 had lower diversity in harvested reaches than in forested reaches, while stream F4 exhibited greater variability and no noticeable difference between the harvested and forested reaches; however, it was not able to be determined i f these trends were significant using Tukey's post-hoc test (Figure 5.11). 83 X CD CO O O i— Q. o CO cc £= O a E to T a • T b -Stream F4 Stream F5 Stream F8 Figure 5.10 Diversity of orders (Simpson's Reciprocal values + 1 standard error) in harvested streams (averages taken from 6 reaches in streams F4 and F5, with 3 reaches from stream F8). Bars with different letters are significantly different at P<0.05 (Tukey's Test). m cc | C/3 Figure 5.11 4 5 6 H a r v e s t e d R e a c h e s 7 8 9 B e l o w - H a r v e s t e d R e a c h e s Diversity of orders (Simpson's Reciprocal values + 1 standard error) for reaches within harvested streams of the harvested treatment area. Bars with different letters are significantly different at P<0.05 (Tukey's Test). 84 Ephemeroptera (Genus Level): Diversity of Ephemeroptera genera differed significantly between reaches within treatments and also exhibited significant stream by treatment interactions (Appendix 10.2). The uppermost reach of streams F4 and F5 within the harvested area had significantly lower diversity than the other reaches within the harvested treatments. The significant interaction between stream and treatments resulted from stream F5 having significantly greater diversity in the below-harvested reaches than in the harvested reaches, while F4 did not differ significantly (Figure 5.12). es ted R e a c h e s B e l o w - H a r v e s t e d R e a c h Figure 5.12 Diversity of Ephemeroptera genera (Simpson's Reciprocal values + 1 standard error) in harvested streams (averages taken from three replicates per reach). Bars with different letters are significantly different at P<0.05 (Tukey's Test). 85 Plecoptera (Genus Level): No significant differences were found in harvested streams for diversity of Plecoptera genera (Appendix 10.3) (Figure 5.13). X CD T3 _ C 75 o o CD CC w "iz o c/> a . S t r e a m F 4 S t r e a m F 5 S t r e a m F 8 4 5 6 H a r v e s t e d R e a c h e s B e l o w - H a r v e s t e d R e a c h e s Figure 5.13 Diversity of Plecoptera genera (Simpson's Reciprocal values + 1 standard error) in harvested streams (averages taken from three replicates per reach). 86 Diptera (Genus Level): Diptera genera diversity did not differ significantly between treatments, streams or reaches within the harvested streams (Appendix 10.4) (Figure 5.14). 3.0 4 5 6 H a rve sted R e a c h e s 7 8 9 Below-Harves t ed R e a c h e s Figure 5.14 Diversity of Diptera genera (Simpson's Reciprocal values + 1 standard error) in harvested streams (averages taken from three replicates per area). 87 Benthic Organic Matter: Benthic organic matter ( A F D M ) varied considerably between reaches and streams, but not significantly between the harvested and below-harvested reaches or between streams for the harvested streams (Appendix 11) (Figure 5.15). 4 5 6 H arves ted R e a c h e s 7 8 9 B e l o w - H a r v e s t e d R e a c h e s Figure 5.15 Ash free dry mass + 1 standard error in the harvested streams (averages taken trom three replicate samples per reach). 88 Relationships between Invertebrates and Stream Characteristics: Regression analysis with invertebrate density as a function of A F D M (periphyton), Chlorophyll a levels, benthic organic matter, streamwater nutrient concentrations and stream temperature indicated that total invertebrate density was significantly related only to A F D M , Chlorophyll a and stream temperature (Table 5.3). Each variable had a positive relationship with total invertebrate density, with A F D M explaining 45% of the variation in total invertebrate density while Chlorophyll a and streamwater temperature explained 55% and 25%, respectively. Similar regression analyses with Ephemeroptera and Plecoptera densities did not reveal significant relationships with periphyton measures, while Trichoptera was found to be significantly related to A F D M and Chironomidae densities were found to be significantly related to Chlorophyll a (with A F D M explaining 30% of the variation in Trichoptera densities and Chlorophyll a explaining 40% of variation in Chironomidae densities)(Tables 5.4 and 5.5). In addition, Trichoptera density was positively related to benthic organic matter (with organic matter explaining 41% of the variation in Trichoptera density) (Table 5.4). A l l best-fit regressions were linear equations. While the following equations best explain the variation in invertebrate densities, it is important to note that these equations should not be extrapolated beyond the range of data sampled. Table 5.3 Regressions with total invertebrate density (TID) as a function of periphyton biomass (AFDM and Chlorophyll a) and streamwater temperature, (n = 24). Regression Eauation Coefficient of Determination R 2 Standard Error of the Estimate SEE P Value TID = 314.40 + 0.76 (AFDM) 0.45 327.60 <0.001 TID = 331.62 + 7.01 (Chlorophyll a) 0.55 295.87 <0.001 TID = -1415.06 + 162.72 (Temperature) 0.25 380.11 0.012 89 Table 5.4 Regressions with Trichoptera density (TD) as a function of periphyton biomass (AFDM) and benthic organic matter, (n = 24). Coefficient of Standard Error of the Regression Equation Determination Estimate P Value E i SEE TD = 15.04 + 0.02 (AFDM) 0.30 12.40 0.006 TD - 13.40 + 9.07 (benthic O.M.) 0.41 11.41 0.001 Table 5.5 Regressions with Chironomidae density (CHIR) as a function of Chlorophyll a levels, (n = 24). Coefficient of Standard Error of the Regression Equation Determination Estimate P Value R 2 SEE CHIR = 7.84 + 2.61 (Chlorophyll a) 0.40 148.53 0.001 5.4 DISCUSSION Longitudinal Trends in the Undisturbed Stream (F7): Invertebrate densities only exhibited significant differences for Diptera with lower densities in the middle reaches than in the upper and lower reaches. With average differences of 12 invertebrates per square meter between the regions, these minor differences were most likely a result of particular habitat differences between reaches along the undisturbed stream. Taxonomic richness did not differ significantly longitudinally. Invertebrate diversity differed significantly between stream regions only for Ephemeroptera and Plecoptera with the upper reaches exhibiting greater diversity of Ephemeroptera than the middle reaches while 90 Plecoptera exhibited lower diversity in the middle reaches than in the lower reaches. The longitudinal differences in Ephemeroptera and Plecoptera diversity are most likely also a result of natural variation between habitat types. While there were some significant differences between certain reaches of the stream, no longitudinal trends were seen along the stream gradient. Benthic organic matter did not vary significantly along the longitudinal gradient of the undisturbed stream. Physical and biological properties of streams typically change along the longitudinal gradient; with shredders dominating upper reaches of streams where allochthonous inputs are high, scrapers dominating intermediate reaches where light levels are higher and collectors dominating downstream reaches as they filter small particles transported from upstream (Hawkins and Sedell 1980). Stream processes function along a longitudinal gradient from headwaters to mouth, with changes in biological processes related to changes in structure and size of streams as explained by the river continuum concept (Vannote et al. 1980). While understanding longitudinal changes that may occur along a linear gradient helps us to understand basic shifts in ecological processes, perhaps a more appropriate theory would include the concept that watersheds are a network of tributaries and main channels (branch shaped) with headwaters exhibiting processes related to their strong coupling with the hillslope (Gomi et al. 2002). While longitudinal changes are seen within an entire watershed, sampling within the study streams occurred over 600 meters stream length in headwater streams at elevations between 500 and 750 meters without substantial changes in stream channel structure, order or riparian vegetation types. Thus, few differences in the stream's physical environment were observed along the longitudinal gradient studied. Therefore, the longitudinal changes described by the river continuum concept were not relevant to the relatively short stream lengths of the present study. In addition, it is important to note that if any longitudinal changes were present in the study streams, the loss of data due to dry summertime conditions caused reduced statistical power and differences would not easily be detected by statistical tests. 91 Effects of Harvesting: -Density-Total invertebrate densities were only statistically significantly greater in harvested reaches than in below-harvested reaches for stream F4. However, both harvested streams had greater densities of total invertebrates in harvested areas with 1480 invertebrates / m 2 and 578 / m 2 in 2 2 harvested reaches versus 300 / iri and 358 / m in the below-harvested areas of streams F4 and F5, respectively. Invertebrate density differences in the study streams may best be explained by variations in habitat quality and greater periphyton biomass caused by forest canopy removal. For example, Chlorophyll a levels (algal biomass) were greater in harvested reaches in stream F4 than in stream F5 harvested reaches during invertebrate sampling, perhaps explaining the differences in total invertebrate densities and more specifically the scrapers which utilize periphyton as a primary food source (e.g. Plecoptera) seen between the harvested and below-harvested reaches of F4 (see Chapter 4 for complete Chlorophyll a results). Forest canopy removal can lead to increased algal production and biomass, providing a greater food base for stream reaches. If forest harvesting produces more complex habitats and woody debris then greater invertebrate densities may result (Hax and Golladay 1993, Downes et al. 2000), while decreased invertebrate densities have been found in streams with increased sedimentation following forest harvesting (Shortreed and Stockner 1983, Kiffney and B u l l 2000). Forest canopy removal leading to increased primary production may lead to greater food production which may increase invertebrate densities (Murphy and Hal l 1980, Noel et al. 1986, Stone and Wallace 1998). In addition to increased primary production, differences in habitat type may be a primary reason for differences in invertebrate abundance, with stream F4 having more stable substrates and channel structure than in stream F5 which had more fine sediments as substrate. However, while visual observations suggested this, substrate types and channel characteristics were not quantified. At the order level, Plecoptera densities were significantly greater in harvested reaches (avg. 2 2 173 stoneflies / m ) than in below-harvested reaches (avg. 58 stoneflies / m ). Stoneflies in 92 general are associated with depositional and erosional habitats and contain shredders, collector-gatherers, predators and some scrapers (Merritt and Cummins 1996). Logging activities in the study streams resulted in the deposition of woody debris into the stream channels. Increases in slash and woody debris within the harvested reaches should benefit shredders and collector-gatherers by providing habitat and food. Allochthonous inputs provide a food source and nutrients for invertebrates. Experimental exclusion of litter has led to decreases in invertebrate densities in small streams (Wallace et al. 1997). Overall differences in total invertebrate densities in the present study were primarily a result of increased numbers of Chironomidae (Diptera) (comprising nearly 40% of the invertebrates in the harvested reaches). This was also found in harvested sites in a New England stream (2-3 years post-logging), where increased invertebrate densities were caused by increases in Diptera and Ephemeroptera (Noel et al. 1986). In the present study, increased densities of Chironomidae may have provided more food sources for predators within the families of Plecoptera, leading to greater abundance of Plecoptera within harvested areas. In harvested small streams in Oregon, increases in overall invertebrate densities resulted from increases in predators, shredders and collector-gatherers as a result of differences in food availability (Hawkins et al. 1982). -Richness-Taxonomic richness was analyzed at the order and family level in the present study. Richness at the order level was greater in the downstream forested reaches than in the harvested reaches. However, this difference averaged only 1 order so biologically it may not be significant. Ephemeroptera exhibited highly variable results for genera level richness between reaches and streams with stream F5 decreasing in richness through the harvested area while the other streams maintained stable levels. Overall higher taxonomic richness in the below-harvested reaches could be a result of a variety of factors such as increased habitat complexity in forested reaches (Downes et al. 2000), negative effects of sedimentation on certain invertebrate groups in harvested reaches (Kiffney and B u l l 2000) or natural drift of invertebrates downstream into the below-harvested reaches. 93 Effects of forest harvesting on aquatic invertebrate richness/number of taxa have been found to be variable depending upon underlying differences in stream structure. For example, increased taxonomic richness was found in a) clear-cut stream reaches in Oregon with greatest increases in small high-gradient streams (Murphy and Hal l 1980), and b) in streams in Great Smoky Mountains Park, Tennessee and North Carolina (Silsbee and Larson 1983). Numbers of taxa were similar between harvested and forested streams 2-3 years post-logging in New England (Noel et al. 1986), with similar results seen 6-17 years post-logging in streams in Oregon (Carlson et al. 1990). However, often studies looking at differences between number of taxa compare harvested streams to streams flowing through relatively young forests. For example, Anderson (1992) working in Washington and Oregon found that a stream flowing through old growth forest (450 year-old conifers) had greater species richness than a stream flowing through a clear-cut, which, in turn, had greater species richness than a stream flowing through a 40-year-old primarily deciduous forest. -Invertebrate Diversity-Greater invertebrate diversity at the order level in streams F4 and F5 than in F8 were small and most likely a result of the lack of Oligochaeta within invertebrate samples in stream F8 (i.e. less taxonomic richness). In addition, streams F4 and F5 consisted of orders with similar numbers of organisms within each of the orders as opposed to stream F8 which consisted primarily of Diptera with smaller numbers of organisms within the other orders (i.e. less evenness). Further analysis at the family level revealed that Ephemeroptera had lower diversity in the uppermost harvested reach of streams F4 and F5 than in the other harvested reaches as a result of lower densities of Baetidae, Ephemeridae and Heptageneidae which were found in greater numbers in the other harvested reaches. Stream F5 had significantly greater Ephemeroptera diversity in the below-harvested reaches than stream F4. 94 These results suggest that forest harvesting in the study area has affected community composition in the study streams 2-3 years post harvesting, similar to results found for Oregon streams 6-17 years post-logging (Carlson et al. 1990). However, variability found between reaches and streams within treatment areas are most likely a result of differences in habitat conditions. -Benthic Organic Matter-Benthic organic matter did not differ significantly between harvested and below-harvested reaches in the study streams, with substantial variability both within treatments and between study streams. Visually, these differences appeared to result from the amount of woody debris found within the specific study reaches, but this could not be verified as levels of woody debris were not measured. While depletion of organic matter and woody debris may not occur immediately following forest harvesting in these study streams, a study on a stream 45 years post-logging found four times as much woody debris in an undisturbed stream than in logged streams (Silsbee and Larson 1983). With the importance of allochthonous inputs to stream invertebrate habitat and food supply, continued riparian vegetation inputs are vital for the riparian food web (Wallace et al. 1997). Factors explaining invertebrate densities: The increase in total invertebrate density with A F D M , Chlorophyll a and streamwater temperature found in the present study is consistent with the results of other studies which indicate that increases in certain invertebrates may be a result of increases in periphyton which is their food supply (Fuller et al. 1986, Hart and Robinson 1990, Quinn et al. 1997, Kiffney et al. 2003). Increased streamwater temperatures are a result of increased light levels following forest harvesting, which act together to aid in the increased growth of periphyton and aquatic insects. While total invertebrate densities increased with periphyton biomass, it is important to note that measures of periphyton biomass are not measures of periphyton productivity and therefore a positive relationship between food availability (measured as biomass) and invertebrate density does not necessarily suggest that there is increased productivity within these study streams. 95 At the order level, Trichoptera numbers increased with A F D M and benthic organic matter and Chironomidae numbers increased with periphyton biomass ( A F D M and Chlorophyll a) and streamwater temperature. The Trichoptera family consists primarily of shredders which benefit from greater amounts of food - organic matter and detritus (Hawkins and Sedell 1980). Increases in Trichoptera and Chironomidae with measures of periphyton biomass suggest increased food supply for invertebrate consumers caused by increases in light levels and streamwater temperatures after harvesting (Murphy and Hal l 1980, Hawkins et al. 1982, Kiffney et al. 2003). In addition, Chironomids may also increase as a result of greater levels of detritus material (i.e. leaf litter) within streams providing additional sources of food for invertebrate collectors (Hutchins and Wallace 2002). Higher levels of algal biomass and increased water temperatures within the harvested reaches may have led to increased densities of invertebrates caused primarily by an increased food supply. Increases in Trichoptera density (consisting primarily of shredders) were considered to be a result of increases in benthic organic matter and algal biomass ( A F D M ) . Chironomidae densities have been shown elsewhere to be related to Chlorophyll a levels (algal biomass) with most chironomids feeding through collector-gatherer processes and filtering of particles within the water (Merritt and Cummins 1996). It is important to note that increases in streamwater temperatures following forest harvesting may increase the breakdown of detritus making food available more quickly to invertebrates (Noel et al. 1986), as well as directly affect invertebrate densities through increased growth rates and fecundity (Vannote and Sweeney 1980). While temperature increases may be beneficial to some invertebrates, there is an optimal temperature range where higher or lower temperatures may negatively affect food supply and invertebrate growth (Vannote and Sweeney 1980). Overall results in the present study are similar to those found in other studies finding bottom-up processes dominate food webs in streams, with increases in periphyton production leading to increases in grazing herbivores (Fuller et al. 1986, Carlson et al. 1990, Hart and Robinson 1990, Kiffney and Richardson 2001) However, other researchers have found top-down processes dominate, where high levels of invertebrates act as a control and grazing leads to 96 lower levels of periphyton biomass (Feminella et al. 1989, H i l l and Harvey 1990, Taylor et al. 2002). Regardless whether an aquatic system is functioning top-down or bottom-up, the interaction between these trophic levels illustrates the important trophic function of periphyton within the stream food-web and aquatic ecosystem. 5.5 C O N C L U S I O N S There were few statistically significant differences in invertebrate density, richness and diversity and benthic organic matter along a longitudinal gradient in the undisturbed stream F7. The few statistically significant but small differences between regions of a stream were considered to be caused by differences in local habitat and substrate differences from reach to reach in the undisturbed stream without following any longitudinal trends. Forest harvesting three years prior to sampling in the study streams resulted in significantly greater total invertebrate densities for stream F4, and greater densities in stream F5 although not significantly greater. In addition, harvested reaches exhibited greater densities of Plecoptera than in below-harvested reaches, most likely due to higher levels of periphyton -a main food source for Plecoptera. Taxonomic richness (order level) was less in harvested reaches than in below-harvested reaches, but overall diversity (order level) did not differ significantly between harvested and forested reaches. Significant differences in diversity between reaches within the harvested area and between the study streams were most likely due to natural variability and differences in habitat (i.e. channel structure, substrate). Benthic organic matter did not differ significantly between harvested and forested reaches. Total invertebrate densities and those of Trichoptera and Chironomidae in particular, increased with periphyton biomass, as a result of additional food resources. In the streams of the present study, bottom-up control of invertebrates, with their numbers dependent on periphyton food supply, appears to be more important than top-down control of invertebrates on periphyton. 97 CHAPTER 6. CONCLUSIONS AND RECOMMENDATIONS Conclusions: Longitudinal Trends: -Objective #1-Longitudinal trends for periphyton biomass, benthic macro invertebrate densities and diversity and benthic organic matter. The present study evaluated the longitudinal trends of stream characteristics (nutrient concentrations, incident solar radiation, stream temperatures, periphyton biomass, benthic organic matter and benthic macro invertebrate density/diversity) during summer in coastal headwater streams of British Columbia. No consistent statistically significant longitudinal trends were seen for any of the study variables along the 600 meter length of stream sampled in the undisturbed streams. Studies evaluating longitudinal trends over longer stream lengths (i.e. from headwaters to mouth), with significant changes in stream order and vegetation types, have found significant differences in stream biological processes following the predictions of the River Continuum Concept (Hawkins and Sedell 1980). However, study streams did not exhibit changes in stream order or size along the shorter stream lengths evaluated in the present study. Effects of Forest Harvesting: -Objective #2-Effects of forest harvesting across a stream on incident solar radiation, streamwater temperature, periphyton biomass, benthic macro invertebrates and benthic organic matter within the harvested area and downstream of the harvested area. Forest harvesting consisting of removal of riparian canopy across the study streams, resulted in greater incident solar radiation and streamwater temperatures in harvested reaches than in 98 forested reaches. Increased levels of periphyton were found in harvested reaches than in forested reaches. Both Chlorophyll a and streamwater temperature increased as streams flowed through the harvested areas then subsequently decreased as the streams flowed further downstream through forest again. This suggests that the harvesting-induced changes may primarily be a function of light availability. Invertebrate densities were greater in harvested reaches (although only statistically significantly greater for stream F4), with statistically significantly greater densities for Plecoptera and Diptera (Chironomidae). Invertebrate taxonomic richness was greater in the forested reaches (order level). However, invertebrate diversity was found to be similar between harvested and forested reaches. Benthic organic matter was unaffected by forest harvesting. Relationships between Research Variables: -Objective #3-Relationships between periphyton biomass, benthic macro invertebrates, benthic organic matter, streamwater chemistry/nutrient concentrations, streamwater temperature and incident solar radiation (light) within the studied streams. Periphyton biomass was positively related to incident solar radiation and streamwater temperatures during the study period. Increased food supply subsequently increased invertebrate densities. Shifts in periphyton may have also been the cause of changes in community composition. Most notably, Plecoptera and Diptera (Chironomidae) densities were greater in harvested reaches than in forested reaches. The results presented above are similar to results found in other harvested small streams in the northern temperature regions (Hansmann and Phinney 1973, Murphy and Hal l 1980, Silsbee and Larson 1983, Noel et al. 1986, Stone and Wallace 1998, Kiffney and B u l l 2000). However, while the previously stated results were similar to those found in most other studies, overall greater taxonomic richness (order level) in the forested (below-harvested) reaches than in harvested reaches differs from findings of other studies which have reported increased taxonomic richness in harvested areas (Murphy and Hal l 1980, Silsbee and Larson 99 1983). These results, together with visual observations of stream channels, suggest that forest harvesting has led to increased invertebrate densities, primarily in the Plecoptera and Diptera invertebrate groups, as a result of increased periphyton biomass (food resources) and increased organic debris (food and habitat resources) from logging activities near the stream channels. Logging debris left behind after logging, which subsequently falls into stream channels, may provide food and habitat for invertebrate species (Noel et al. 1986). While forest harvesting increased light, streamwater temperature and primary and secondary production, downstream effects were minimal as the preceding variables returned to undisturbed stream conditions within the first 100 meters of stream length downstream into the forest from the harvested area boundary. This is suggested to be primarily a result of reduction in light availability, with light considered to be a primary controlling factor of the biological responses seen in these headwater streams. However, while nutrient levels were not significantly related to any of the biological factors within the present study, their possible effect on primary production should not be ruled out. Recommendations: Recommendations for Further Research: While the present study provided a snapshot of the biological processes within the harvested and undisturbed study streams, further research should focus on how individual biological processes control the results found in the present study. In addition, further research within these streams should focus on controlling certain variables such as light (e.g. shade manipulation), allochthonous inputs (e.g. leaf-litter exclusion) and streamwater nutrient concentration (e.g. nutrient additions) to develop a greater understanding of how these streams, function ecologically. In particular, I would suggest that a study looking at the differences between organic debris and sediment levels in the harvested reaches would potentially help explain differences seen in invertebrate densities and composition between the different study streams. In terms of gaining an overall better understanding of harvesting effects on headwater stream processes, more work should be done to understand the longer term effects of harvesting, rates of recovery and downstream cumulative effects. In addition, 100 studies should attempt to be integrated in order to avoid situations where research findings are not able to be extrapolated to other stream systems (Feminella and Hawkins 1995). It would be desirable to conduct a similar study to the present one during a period in which stream flow remains sufficient to allow for all necessary sampling to be completed. In the present study, stream reaches continued to dry up as the study progressed, which resulted in considerable missing data, making statistical analysis less powerful. In addition, differences between streams in structural and physical characteristics should be understood and determined as thoroughly as possible. For example, while streams within the same watershed may be relatively similar, it is important to quantify basic differences in stream characteristics (i.e. discharge, organic debris, substrate type) in order to help explain why apparently similar nearby streams may differ biologically. Recommendations for Management: Considerable energy and work has been put into developing an understanding of how to best manage our forests for both sustainable forestry and fisheries production. Relatively small scale forest harvesting which cuts across riparian zones of small streams (such as the harvesting which occurred in the study streams), can significantly alter the stream system from one that depends upon allochthonous inputs to one that obtains energy from the sun through photosynthesis (autochthonous inputs). While this did cause changes within the biological processes of the harvested reaches, these changes did not have any apparent significant effect directly downstream of harvested areas, which could ultimately affect downstream fisheries. Proper watershed management should include the protection of headwater streams as important sources of sediment, nutrients and particulate organic matter for downstream ecosystems (Gomi et al. 2002). For proper management decisions, we must have a more complete understanding of recovery rates for various land uses and human caused disturbances. For example, in small forested streams, recovery rates are directly coupled with re-growth of vegetation which wi l l ultimately return the system to forested conditions (Murphy and Hal l 1980). Rates of recovery following logging have been found to be dependent on underlying substrate 101 conditions, with streams having bedrock streambeds recovering more quickly than streams containing depositional areas (Stone and Wallace 1998). Therefore forest management plans should take into consideration the effects of harvesting on different stream structures and vegetation types. The preservation of vegetation leave strips (riparian buffers) around streams has been utilized as a means of protecting stream ecosystems from the effects of forest harvesting activities (Young 2000). While use of buffers may protect against land uses that can degrade aquatic , systems (Richardson and Jackson 2000), buffering small branching headwater systems with sufficiently sized buffers of 30 meters, may not even protect these small densely forested streams from the effects of increased light levels (Kiffney et al. 2003). While buffering may provide for some protection from the effects of harvesting, results of the present study suggest that harvesting across the riparian zones in these small headwater streams did not lead to dramatic changes in the biological system that would result in degradation of downstream ecosystems. However, as stated previously, this study was carried out during summertime conditions, 3 years after harvesting and therefore further studies would be needed to determine whether or not the streams would function differently at different times of the year and/or in different years post-harvest. If we are to properly manage aquatic systems for protection of aquatic ecosystem processes, it seems important that we protect these streams against harvesting activities that would result in long lengths of stream being harvested to the streambank. While the harvested lengths of 300-750 meters in the present study, did not result in significant downstream effects, harvesting of entire lengths of headwater streams without the preservation of upstream and/or downstream forested areas could potentially result in more dramatic changes to in-stream processes and longer recovery times. Aquatic species are well adapted to living in harsh environments with natural disturbances and as long as there are areas of refugia preserved in nearby stream reaches, disturbed reaches wi l l recover if given adequate time. However, until we have a complete understanding of how harvesting activities in headwater streams affect downstream reaches both in the short and long-term, protection of basic stream structure and ecosystem processes should be implemented. 102 R E F E R E N C E S : Allan, J. D. 1995. Stream Ecology: Structure and function of running waters. Chapman & Hall, London, United Kingdom. Anderson, N . H. 1992. Influence of disturbance on insect communities in Pacific Northwest streams. Hydrobiologia 248:79-92. Bilby, R. E., and P. A . Bisson. 1992. Allochthonous versus autochthonous organic matter contributions to the trophic support of fish populations in clear-cut and old-growth forested streams. Canadian Journal of Fisheries and Aquatic Sciences 49:540-551. Binkley, D., H. Burnham, and H. L. Allen. 1999. Water quality impacts of forest fertilization with nitrogen and phosphorus. Forest Ecology and Management 121:191-213. Bjornn, T. C , and D. W. Reiser. 1991. Influences of forest and rangeland management on salmonid fishes and their habitats. American Fisheries Society Special Publication 19:83-138. Bormann, F. H., G. E. Likens, T. G. Siccama, R. S. Pierce, and J. S. Eaton. 1974. The export of nutrients and recovery of stable conditions following deforestation at Hubbard Brook. Ecological Monographs 44:255-277. Brosofske, K. D., J. Chen, R. J. Naiman, and J. F. Franklin. 1997. Harvesting effects on microclimatic gradients from small streams to uplands in western Washington. Ecological Applications 7:1188-1200. Brown, G. W. 1985. Forestry and Water Quality. Oregon State University, Corvallis. Brown, T. C , and D. Binkley. 1994. Effect of management on water quality in North American forests. U.S. Forest Service. General Technical Report RM-248:l-27. Burt, T. P. 1996. Chapter 2: The Hydrology of Headwater Catchments, in G. Petts and P. Calow editors. River Flows and Channel Forms. Blackwell Science, Oxford. Carignan, R., and R. J. Steedman. 2000. Impacts of major watershed perturbations on aquatic ecosystems. Canadian Journal of Fisheries and Aquatic Sciences 57:1-4. Carlson, J. Y. , C. W. Andrus, and H. A. Froehlich. 1990. Woody debris, channel features, and macroinvertebrates of streams with logged and undisturbed riparian timber in northeastern Oregon, U.S.A. Canadian Journal of Fisheries and Aquatic Sciences 47:1103-1111. Davis, J. C , and G. W. Minshall. 1999. Nitrogen and phosphorus uptake in two Idaho (USA) headwater wilderness streams. Oecologia 119:247-255. Downes, B. J., P. S. Lake, E. S. G. Schreiber, and A. Glaister. 2000. Habitat structure, resources and diversity: the separate effects of surface roughness and macroalgae on stream invertebrates. Oecologia 123:569-581. Essington, T. E., and S. R. Carpenter. 2000. Nutrient cycling in lakes and steams: Insights from a comparative analysis. Ecosystems 3:131-143. 103 Fausch, K. D., and T. G. Northcote. 1992. Large woody debris and salmonid habitat in a small coastal British Columbia stream. Canadian Journal of Fisheries and Aquatic Sciences 49:682-693. Feller, M . C. 1981. Effects of clearcutting and slashburning on stream temperatures in southwestern British Columbia. Water Resources Bulletin 17:863-868. Feller, M . C , and J. P. Kimmins. 1979. Chemical characteristics of small streams near Haney in southwestern British Columbia. Water Resources Research 15:247-258. Feller, M . C , and J. P. Kimmins. 1984. Effects of clearcutting and slash burning on streamwater chemistry and watershed nutrient budgets in southwestern British Columbia. Water Resources Research 20:29-40. Feminella, J. W., and C. P. Hawkins. 1995. Interactions between stream herbivores and periphyton: a quantitative analysis of past experiments. Journal of the North American Benthological Society 14:465-509. Feminella, J. W., M . E. Power, and V . H . Resh. 1989. Periphyton responses to invertebrate grazing and riparian canopy in three northern California coastal streams. Pages 445-457 in Stream periphyton, grazing invertebrates and canopy. University of California at Berkeley. Fuller, R. L. , J. L . Roelofs, and T. J. Fry. 1986. The importance of algae to stream invertebrates. Journal of the North American Benthological Society 5:290-296. Giller, P. S., and B. Malmqvist. 1998. The Biology of Streams and Rivers. Oxford University Press, New York. Gomi, T., R. C. Sidle, and J. S. Richardson. 2002. Understanding processes and downstream linkages of headwater streams. Bioscience 52:905-916. Gregory, S. V. , F. J. Swanson, W. A. McKee, and K. W. Cummins. 1991. An ecosystem perspective of riparian zones: Focus on links between land and water. Bioscience 41:540-551. Grimm, N . B. , and S. G. Fisher. 1989. Stability of periphyton and macroinvertebrates to disturbance by flash floods in a desert stream. Journal of the North American Benthological Society 8:293-307. Guidelines, T. F. o. W. Q. 1987. Canadian Water Quality Guidelines. Water Quality Branch, Inland Waters Directorate, Ottawa, Ontario. Hansmann, E. W., and H. K. Phinney. 1973. Effects of logging on periphyton in coastal streams of Oregon. Ecology 54:194-199. Harr, D. R. 1986. Effects of clearcutting on rain-on-snow runoff in western Oregon: A new look at old studies. Water Resources Research 22:1095-1100. Hart, D. D., and C. T. Robinson. 1990. Resource limitation in a stream community: phosphorous enrichment on periphyton and grazers. Ecology 71:1494-1502. 104 Hartman, G. F., and J. C. Scrivener. 1990. Impacts of forestry practices on a coastal stream ecosystem, Carnation Creek, British Columbia. Canadian Bulletin of Fisheries and Aquatic Sciences 223:p.l48. Hauer, F. R., and G. A. Lamberti, editors. 1996. Methods in Stream Ecology. Academic Press, San Diego, California. Hawkins, C. P., M . L . Murphy, and N . H. Anderson. 1982. Effects of canopy, substrate composition, and gradient on the structure of macroinvertebrate communities in Cascade Range streams of Oregon. Ecology 63:1840-1856. Hawkins, C. P., and J. R. Sedell. 1980. Longitudinal and seasonal changes in functional organization of macroinvertebrate communities in four Oregon streams. Ecology 62:387-397. Hax, C. L. , and S. W. Golladay. 1993. Macroinvertebrate colonization and biofilm development on leaves and wood in a boreal river. Freshwater Biology 29:79-87. Hetrick, N . J., M . A . Brusven, T. C. Bjornn, R. M . Keith, and W. R. Meehan. 1998a. Effects of canopy removal on invertebrates and diet of juvenile coho salmon in a small stream in southeast Alaska. American Fisheries Society 127:876-888. Hetrick, N . J., M . A . Brusven, W. R. Meehan, and T. C. Bjornn. 1998b. Changes in solar input, water temperature, periphyton accumulation, and allochthonous input and storage after canopy removal along two small salmon streams in southeast Alaska. Transactions of the American Fisheries Society 127:859-875. Hicks, B. J., R. L. Beschta, and R. D. Harr. 1991. Long-term changes in streamflow following logging in western Oregon and associated fisheries implications. Water Resources Bulletin 27:217-226. Hil l , W. R., and B. C. Harvey. 1990. Periphyton responses to higher tropic levels and light in a shaded stream. Canadian Journal of Fisheries and Aquatic Sciences 47:2307-2314. Hil l , W. R., and A. W. Knight. 1988. Nutrient and light limitation of algae in two northern California streams. Journal of Phycology 24:125-132. Hill , W. R., M . G. Ryon, and E. M . Schilling. 1995. Light limitation in a stream ecosystem: responses by primary producers and consumers. Ecology 76:1297-1309. Holtby, L . B. 1988. Effects of logging on stream temperatures in Carnation Creek, British Columbia, and associated impacts on the coho salmon (Oncorhynchus kisutch). Canadian Journal of Fisheries and Aquatic Sciences 45:502-515. Home, A . J., and C. R. Goldman. 1994. Limnology, Second edition. McGraw-Hill, Inc., United States. Hudson, R. 2001a. Roberts Creek Study Forest: Pre-harvest chemical characteristics of three S6 creeks, Flume Creek Experimental Watershed. Forest Research Technical Report TR-013, B.C. Ministry of Forests. British Columbia, Nanaimo, B.C. 105 Hudson, R. 2001b. Roberts Creek Study Forest: preliminary effects of partial harvesting on peak streamflow in two S6 Creeks. Forest Research Extension Note EN-007, B.C. Ministry of Forests. British Columbia, Nanaimo, B.C. Hutchins, J. J., and J. B. Wallace. 2002. Ecosystem linkages between southern Appalachian headwater streams and their banks: Leaf litter breakdown and invertebrate assemblages. Ecosystems 5:80-91. Johnson, C. E., C. T. Driscoll, T. G. Siccama, and G. E. Likens. 2000. Element fluxes and landscape position in a northern hardwood forest watershed ecosystem. Ecosystems 3:159-184. Johnson, S. L., and J. A. Jones. 2000. Stream temperature responses to forest harvest and debris flows in western Cascades, Oregon. Canadian Journal of Fisheries and Aquatic Sciences 57:30-39. Jones, J. A. , and G. E. Grant. 1996. Peak flow responses to clear-cutting and roads in small and large basins, western Cascades, Oregon. Water Resources Research 32:959-974. Kiffney, P. M . , and J. P. Bull. 2000. Factors controlling periphyton accrual during summer in headwater streams of southwestern British Columbia, Canada. Journal of Freshwater Ecology 15:339-351. Kiffney, P. M . , and J. S. Richardson. 2001. Interactions among nutrients, periphyton, and invertebrates and vertebrate (Ascaphus truei) grazers in experimental channels. Copeia 2:422-429. Kiffney, P. M . , J. S. Richardson, and J. P. Bull. 2003. Responses of periphyton and insect consumers to experimental manipulation of riparian buffer width along headwater streams, unpublished manuscript. Kiffney, P. M . , J. S. Richardson, and M . C. Feller. 2000. Fluvial and epilithic organic matter dynamics in headwater streams of southwestern British Columbia, Canadian Archive Fur Hydrobiologia 149:109-129. Kohler, S. L. 1984. Search mechanism of a stream grazer in patchy environments: the role of food abundance. Oecologia 62:209-218. Krebs, C. J. 1994. Ecology: The experimental analysis of distribution and abundance, Fourth edition. Harper Collins College Publishers, New York. Krebs, C. J. 1999. Ecological Methodology, 2nd edition. Addison Wesley Longman, Menlo Park, California. Lamberti, G. A. , and V. H. Resh. 1983. Stream periphyton and insect herbivores: an experimental study of grazing by a caddisfly population. Ecology 64:1124-1135. Likens, G. E., F. H. Bormann, N . M . Johnson, D. W. Fisher, and R. S. Pierce. 1970. Effects of forest cutting and herbicide treatment on nutrient budgets in the Hubbard Brook Watershed-Ecosystem. Ecological Monographs 40:23-47. 106 Lowe, R. L. , S. W. Golladay, and J. R. Webster. 1986. Periphyton response to nutrient manipulation in streams draining clearcut and forested watersheds. Journal of the North American Benthological Society 5:221-229. Lynch, J. A. , G. B. Rishel, and E. S. Corbett. 1984. Thermal alteration of streams draining clearcut watersheds: Quantification and biological implications. Hydrobiologia 111:161-169. Magurran, A. E. 1988. Ecological Diversity and its Measurement. Princeton University Press, Princeton, New Jersey. McGurk, B. J. 1988. Predicting stream temperature after riparian vegetation removal. Pages 157-164 in California Riparian Systems Conference. U.S. Department of Agriculture: Forest Service, Davis, California. Merritt, R. W., and K. W. Cummins. 1996. An introduction to the aquatic insects of North America, Third edition. Kendall/Hunt, Dubuque, Iowa. Minshall, G. W., C. T. Robinson, and D. E. Lawrence. 1997. Postfire responses of lotic ecosystems in Yellowstone National Park, U.S.A. Canadian Journal of Fisheries and Aquatic Sciences 54:2509-2525. Montgomery, D. R., and J. M . Buffington. 1997. Channel-reach morphology in mountain drainage basins. Geological Society of America Bulletin 109:596-611. Mulholland, P. J., and W. R. Hil l . 1997. Seasonal patterns in streamwater nutrient and dissolved organic carbon concentrations: Separating catchment flow path and in-stream effects. Water Resources Research 33:1297-1306. Mulholland, P. J., J. D. Newbold, J. W. Elwood, L . A. Ferren, and J. R. Webster. 1985. Phosphorus spiraling in a woodland stream: seasonal variations. Ecology 66:1012-1023. Munn, N . L. , and M . L. Meyer. 1990. Habitat-specific solute retention in two small streams: an intersite comparison. Ecology 71:2069-2089. Murphy, M . L. , and J. D. Hall. 1981. Varied effects of clear-cut logging on predators and their habitat in small streams of the Cascade Mountains, Oregon. Canadian Journal of Fisheries and Aquatic Sciences 38:137-145. Murphy, M . L. , and W. R. Meehan. 1991. Influences of forest and rangeland management on salmonid fishes and their habitats. American Fisheries Society 19:17-46. Newbold, J. D., J. W. Elwood, R. V . O'Neill, and A. L . Sheldon. 1982. Nutrient spiralling in streams: implications for nutrient limitation and invertebrate activity. American Naturalist 120:628-652. Newbold, J. D., J. W. Elwood, R. V . O'Neill, and W. V . Winkle. 1981. Measuring nutrient spiralling in streams. Canadian Journal of Fisheries and Aquatic Sciences 38:860-863. Noel, D. S., C. W. Martin, and C. A. Federer. 1986. Effects of forest clearcutting in New England on stream macroinvertebrates and periphyton. Environmental Management 10:661-670. 107 Perrin, C. J., M . L . Bothwell, and P. A. Slaney. 1987. Experimental enrichment of a coastal stream in British Columbia: Effects of organic and inorganic additions on autotrophic periphyton production. Canadian Journal of Fisheries and Aquatic Sciences 44:1247-1256. Peterson, B. J., L . Deegan, J. Helfrich, J. E. Hobbie, M . Hullar, B. Moller, T. E. Ford, A. Hershey, A. Hiltner, G. Kipphut, M . A . Lock, D. M . Fiebig, V . McKinley, M . C. Miller, J. R. Vestal, R. Ventullo, and G. Volk. 1993. Biological responses of a tundra river to fertilization. Ecology 74:653-672. Peterson, B. J., W. M . Wollheim, P. J. Mulholland, J. R. Webster, J. L . Meyer, J. L . Tank, E. Marti, W. B. Bowden, H. M . Valett, A . E. Hershey, W. H. McDowell, W. K. Dodds, S. K. Hamilton, S. Gregory, and D. D. Morrall. 2001. Control of nitrogen export from watersheds by headwater streams. Science 292:86-90. Quinn, J. M . , A. B. Cooper, M . J. Stroud, and G. P. Burrell. 1997. Shade effects on stream periphyton and invertebrates: an experiment in streamside channels. New Zealand Journal of Marine and Freshwater Research 31:665-683. Rempel, R. S., and J. C. H. Carter. 1986. An experimental study on the effect of elevated temperature on the heterotrophic and autotrophic food resources of aquatic insects in a forested stream. Canadian Journal of Zoology 64:2457-2466. Richards, C , and G. W. Minshall. 1988. The influence of periphyton abundance on Baetis bicaudatus distribution and colonization in a small stream. Journal of the North American Benthological Society 7:77-86. Richardson, J. S., and M . J. Jackson. 2000. Aquatic Invertebrates, in A . J. Davy, editor. Handbook of Ecological Restoration: Principles of Restoration. Cambridge, New York. Roby, K. B., and D. L. Azuma. 1995. Changes in a reach of a northern California stream following wildfire. Environmental Management 19:591-600. Rosemond, A . D. 1994. Multiple factors limit seasonal variation in periphyton in a forest stream. Journal of the North American Benthological Society 13:333-344. Rosemond, A . D., P. J. Mulholland, and S. H. Brawley. 2000. Seasonally shifting limitation of stream periphyton: response of algal populations and assemblage biomass and productivity to variation in light, nutrients, and herbivores. Canadian Journal of Fisheries and Aquatic Sciences 57:66-75. Shortreed, K. S., and J. G. Stockner. 1983. Periphyton biomass and species composition in a coastal rainforest stream in British Columbia: effects of environmental changes caused by logging. Canadian-Journal of Fisheries and Aquatic Sciences 40:1887-1895. Silsbee, D. G., and G. L . Larson. 1983. A comparison of streams in logged and unlogged areas of Great Smoky Mountains National Park. Hydrobiologia 102:99-111. Stone, M . K., and J. B. Wallace. 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. 108 Swank, W. T., and D. A . Crossley. 1988. Forest hydrology and ecology at Coweeta. Springer-Verlag, New York. SYSTAT. 1998. SYSTAT 9.0 for Windows. SPSS Inc., Chicago. Taylor, B. W., A. R. Mcintosh, and B. L . Peckarsky. 2002. Reach-scale manipulations show invertebrate grazers depress algal resources in streams. Limnology and Oceanography 47:893-899. Thedinga, J. F., M . L . Murphy, J. Heifetz, K. V. Koski, and S. W. Johnson. 1989. Effects of logging on size and age composition of juvenile coho salmon (Oncorkynchus kisutch) and density of pre-smolts in southeast Alaska streams. Canadian Journal of Fisheries and Aquatic Sciences 46:1383-1391. Toth, E. S. 1998. Watershed analysis as a tool for landscape management, monitoring, and restoration. Pages 45-54 in M . K . Brewin, editor. Forest-Fish Conference: Land management practices affecting aquatic ecosystems. Natural Resources Canada, Canadian Forest Service: Northern Forestry Centre. Tripathi, S. K., and K. P. Singh. 1994. Productivity and nutrient cycling in recently harvested and mature bamboo savannas in the dry tropics. Journal of Applied Ecology 31:109-124. Valentine, K. W. G., P. N . Sprout, T. E. Baker, and L . M . Lavkulich, editors. 1978. The soil landscapes of British Columbia. The Resource Analysis Branch, Ministry of the Environment, Victoria, British Columbia. Vannote, R. L. , G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. The river continuum concept. Canadian Journal of Fisheries and Aquatic Sciences 37:130-137. Vannote, R. L. , and B. W. Sweeney. 1980. Geographic analysis of 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., S. L . Eggert, J. L . Meyer, and J. R. Webster. 1997. Multiple trophic levels of a forest stream linked to terrestrial litter inputs. Science 277:102-104. Ward, J. V. , and J. A . Stanford. 1982. Thermal responses in the evolutionary ecology of aquatic insects. Annual Review of Entomology 27:97-117. Wootton, J. T., and M . E. Power. 1993. Productivity, consumers, and the structure of a river food chain. Proceedings of the National Academy of Sciences of the United States of America 90:1384-1387. Young, K . A . 2000. Riparian zone management in the Pacific Northwest: Who's cutting what? Environmental Management 26:131-144. Zweig, L . D., and C. F. Rabeni. 2001. Biomonitoring for deposited sediment using benthic invertebrates: a test on 4 Missouri streams. Journal of the North American Benthological Society 20:643-657. 109 APPENDIX 1 - ANOVA's (Undisturbed Streams - Dissolved ion concentrations): ANOVA results for A l , Ca, Mg, Na, N 0 3 and S i 0 2 concentrations in undisturbed streams. P-Value of <0.05 indicates significant results. Al Effect Sum-of-Sauares df Mean-Sauare F P R E A C H 0.00 4 0.00 0.18 0.94 E R R O R 1 0.00 20 0.00 0.38 0.98 W E E K 0.00 3 0.00 4.97 0.01 R E A C H * W E E K 0.01 29 0.00 0.45 0.98 E R R O R 2 0.02 40 0.00 Ca Effect Sum-of-Sauares df Mean-Sauare F P R E A C H 0.01 6 0.00 0.10 0.99 E R R O R 1 0.18 18 0.01 0.24 0.99 W E E K 0.83 4 0.21 16.78 <0.001 R E A C H * W E E K 0.34 28 0.01 0.30 0.99 E R R O R 2 1.63 40 0.04 Mg Effect Sum-of-Sauares df Mean-Sauare F P R E A C H 0.00 3 0.00 0.21 0.89 E R R O R 1 0.02 21 0.00 0.30 0.99 W E E K 0.04 4 0.01 21.35 <0.001 R E A C H * W E E K 0.01 28 0.00 0.16 1.00 E R R O R 2 0.13 40 0.00 Na Effect Sum-of-Sauares df Mean-Sauare F P R E A C H 0.00 3 0.00 0.02 0.99 E R R O R 1 0.27 20 0.01 0.55 0.92 W E E K 0.33 4 0.08 16.68 <0.001 R E A C H * W E E K 0.14 29 0.01 0.20 1.00 E R R O R 2 0.98 40 0.03 110 N0 3 Effect Sum-of-Sauares df Mean-Square F P R E A C H 0.00 3 0.00 0.22 0.88 E R R O R 1 0.07 20 0.01 0.90 0.59 W E E K 0.26 4 0.06 42.30 <0.001 R E A C H * W E E K 0.04 29 0.00 0.38 0.99 E R R O R 2 0.16 40 0.01 SiQ 2 Effect Sum-of-Sauares df Mean-Sauare F P R E A C H 0.92 5 0.18 0.86 0.52 E R R O R 1 3.84 18 0.21 0.33 0.99 W E E K 24.30 4 6.08 10.87 <0.001 R E A C H * W E E K 16.21 29 0.56 0.88 0.64 E R R O R 2 25.50 40 0.64 111 APPENDIX 2 - ANQVA's (Harvested Streams - Dissolved ion concentrations): ANOVA results for A l , Ca, K, Mg, Na, N0 3 , S i0 2 and S 0 4 concentrations in harvested streams. P-Value of <0.05 indicates significant results. Treatments (TRT) refers to the three areas in the harvested streams; above-harvested, harvested and below-harvested reaches. Al Effect Sum-of-Sauares df Mean-Square F P TRT 0.02 2 0.011 24.72 0.01 REACH(TRT) 0.00 3 0.000 0.19 0.90 E R R O R 1 0.03 14 0.002 0.27 0.99 W E E K 0.14 4 0.034 8.72 0.01 T R T * W E E K 0.03 8 0.004 8.95 <0.001 W E E K * R E A C H ( T R T ) 0.01 22 0.000 0.05 1.00 E R R O R 2 0.40 48 0.008 Ca Effect Sum-of-Sauares df Mean-Sauare F P TRT 1.37 2 0.69 2.26 0.22 REACH(TRT) 1.22 4 0.3 3.07 0.05 E R R O R 1 1.39 14 0.10 0.56 0.88 W E E K 7.57 4 1.89 8.19 0.01 T R T * W E E K 1.85 8 0.23 2.20 0.07 W E E K * R E A C H ( T R T ) 2.21 21 0.11 0.59 0.91 E R R O R 2 8.541 48 0.178 K Effect Sum-of-Sauares df Mean-Sauare F P TRT 0.11 2 0.06 128.40 <0.001 REACH(TRT) 0.00 3 0.00 0.01 0.99 E R R O R 1 0.46 14 0.03 1.93 0.05 W E E K 0.19 4 0.05 9.71 0.01 T R T * W E E K 0.03 7 0.01 1.35 0.27 W E E K * R E A C H ( T R T ) 0.08 23 0.00 0.22 1.00 E R R O R 2 0.81 48 0.02 Mg Effect Sum-of-Sauares df Mean-Sauare F P TRT 0.02 2 0.01 0.48 0.65 REACH(TRT) 0.08 5 0.02 2.35 0.11 E R R O R 1 0.08 12 0.01 0.80 0.65 W E E K 0.38 4 0.09 5.90 0.02 T R T * W E E K 0.11 7 0.02 1.98 0.10 W E E K * R E A C H ( T R T ) 0.19 23 0.01 0.99 0.50 E R R O R 2 0.39 48 0.01 112 Na Effect Sum-of-Sauares df Mean-Sauare F P TRT 0.00 1 0.00 1.50 0.27 REACH(TRT) 0.00 6 0.00 0.08 0.99 E R R O R 1 0.01 12 0.00 2.51 0.01 W E E K 0.00 4 0.00 1.04 0.45 T R T W E E K 0.00 8 0.00 1.31 0.29 W E E K * R E A C H ( T R T ) 0.01 22 0.00 1.00 0.48 E R R O R 2 0.02 48 0.00 N0 3 Effect Sum-of-Sauares df Mean-Sauare F P TRT 0.13 2 0.07 3.54 0.10 REACH(TRT) 0.11 6 0.02 1.05 0.44 E R R O R 1 0.21 12 0.02 0.74 0.71 W E E K 2.26 4 0.57 10.82 0.01 T R T * W E E K 0.37 7 0.05 3.51 0.01 W E E K * R E A C H ( T R T ) 0.33 22 0.02 0.62 0.89 E R R O R 2 1.16 48 0.02 SiO z Effect Sum-of-Sauares df Mean-Sauare F P TRT 0.02 2 0.01 1.69 0.48 REACH(TRT) 0.01 1 0.01 0.57 0.46 E R R O R 1 0.19 16 0.01 0.54 0.91 W E E K 0.44 4 0.11 5.96 0.02 T R T * W E E K 0.15 8 0.02 2.35 0.05 W E E K * R E A C H ( T R T ) 0.17 22 0.01 0.36 0.99 E R R O R 2 1.06 48 0.02 so 4 Effect Sum-of-Sauares df Mean-Sauare F P TRT 0.70 2 0.35 30.73 0.01 REACH(TRT) 0.05 4 0.01 0.10 0.98 E R R O R 1 1.55 14 0.11 0.52 0.91 W E E K 5.77 3 1.92 6.48 0.02 T R T * W E E K 2.37 8 0.30 3.43 0.01 W E E K * R E A C H ( T R T ) 1.90 22 0.09 0.40 0.99 E R R O R 2 10.29 48 0.21 113 APPENDIX 3 - Average dissolved ion concentrations over six-week study period: C o n c e n t r a t i o n 0.20 0.18 0.16 H 0.14 0.12 _i "5> 0.10 E 0.08 0.06 0.04 -0.02 -0.00 A l - A b o v e - H a r v e s t e d R e a c h e s A l - H a r v e s t e d R e a c h c 3 4 W e e k s Concent ra t ion 1.8 -i Figures: Average concentrations of A l and Na for different treatments within the harvested streams over the six-week study period. 114 Concent ra t ion 1.8 1.6 1.4 1.2 —I 1.0 £ 0.8 0.6 0.4 0.2 0.0 •o-\ s o 4 - Above-Harvested Reaches w o s o 4 - Harvested Reaches — - T — s o 4 - Below-Harvested Reaches \ 3 4 W e e k s Figures: Average concentrations of S i0 2 and S 0 4 for different treatments within the harvested streams over the six-week study period. 115 A P P E N D I X 4 - A N O V A ' s - Streamwater Temperature: ANOVA results for streamwater temperature of undisturbed streams. P-Value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F P R E A C H 9.63 6 1.61 2.67 0.05 E R R O R 1 10.81 18 2.60 0.41 0.98 W E E K 119.16 4 25.79 12.65 <0.001 R E A C H * W E E K 75.34 32 2.35 1.59 0.10 E R R O R 2 45.97 31 1.48 2.73 ANOVA results for streamwater temperature of harvested streams. P-Value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F P TRT 107.93 2 53.96 28.15 0.01 REACH(TRT) 7.67 4 1.92 2.41 0.09 E R R O R 1 11.15 14 0.80 0.50 0.92 W E E K 171.64 4 42.91 28.30 <0.001 T R T * W E E K 12.13 8 1.52 2.73 0.03 W E E K * R E A C H ( T R T ) 12.78 23 0.56 0.35 0.99 E R R O R 2 88.85 56 1.59 116 A P P E N D I X 5 - A N O V A ' s (Periphyton Measures): 5.1 ANOVA results for A F D M in the undisturbed streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F P R E A C H 0.68 6 0.11 0.30 0.93 E R R O R 1 6.89 18 0.38 0.92 0.56 W E E K 4.52 5 0.90 1.36 0.26 R E A C H * W E E K 26.59 40 0.67 1.60 0.06 E R R O R 2 19.57 47 0.42 5.2 ANOVA results for Chlorophyll a in the undisturbed streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F P R E A C H .59 4 0.148 1.04 0.41 E R R O R 1 2.86 20 0.143 0.69 0.82 W E E K 29.01 5 5.800 75.53 <0.001 R E A C H * W E E K 3.07 40 0.077 0.37 1.00 E R R O R 2 9.58 46 0.208 5.3 Anova results for AFDM in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F P TRT 6.37 2 3.18 9.04 0.02 REACH(TRT) 1.76 5 0.35 0.40 0.84 E R R O R 1 11.49 13 0.88 1.40 0.18 W E E K 4.60 4 1.15 4.70 0.02 T R T * W E E K 2.45 10 0.24 0.39 0.94 W E E K * R E A C H ( T R T ) 18.86 30 0.63 0.99 0.49 E R R O R 2 44.37 70 0.63 5.4 ANOVA results for Chlorophyll a in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F P TRT 9.63 2 4.81 15.21 0.01 REACH(TRT) 1.27 4 0.32 1.47 0.27 E R R O R 1 3.02 14 0.22 1.27 0.25 W E E K 21.63 4 5.41 102.51 <0.001 T F f T W E E K 0.53 10 0.05 0.60 0.80 W E E K * R E A C H ( T R T ) 2.62 30 0.09 0.51 0.98 E R R O R 2 11.73 69 0.17 117 A P P E N D I X 6 - A N O V A ' s (Undisturbed Stream (F7) - Invertebrate Density): 6.1 Anova results for invertebrate density in the undisturbed stream F7. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Region (upper-middle-lower) 11.36 2 5.679 1.89 0.17 Error 72.14 24 3.006 6.2 Anova results for Ephemeroptera density in the undisturbed stream F7. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Region (upper-middle-lower) 4.33 2 2.165 1.73 0.20 Error 30.00 24 1.250 6.3 Anova results for Plecoptera density in the undisturbed stream F7. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Region (upper-middle-lower) 1.52 2 0.762 0.71 0.50 Error 25.69 24 1.070 6.4 Anova results for Diptera density in the undisturbed stream F7. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Region (upper-middle-lower) 1.93 2 0.966 5.69 0.01 Error 4.07 24 0.170 118 APPENDIX 7 -ANOVA's (Undisturbed Stream (F7) - Invertebrate Diversity and Organic Matter): 7.1 Anova results for diversity of invertebrates (measured as Simpson's Reciprocal Index) at the order level in the undisturbed stream F7. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Region (upper-middle-lower) 0.27 2 0.136 0.40 0.69 Error 2.02 6 0.337 7.2 Anova results for diversity of Ephemeroptera genera (diversity measured as Simpson's Reciprocal Index) in the undisturbed stream F7. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Region (upper-middle-lower) 1.19 2 0.592 6.47 0.03 Error 0.55 6 0.092 7.3 Anova results for diversity of Plecoptera genera (diversity measured as Simpson's Reciprocal Index) in the undisturbed stream F7. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Region (upper-middle-lower) 0.39 2 0.196 5.38 0.046 Error 0.22 6 0.036 7.4 Anova results for diversity of Diptera genera (diversity measured as Simpson's Reciprocal Index) in the undisturbed stream F7. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Region (upper-middle-lower) 3.92 2 1.96 1.98 0.22 Error 5.96 6 0.99 7.5 Anova results for A F D M (Organic Matter) in the undisturbed stream F7. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Region (upper-middle-lower) 0.01 2 0.002 1.60 0.22 Error 0.04 24 0.002 119 A P P E N D I X 8 - A N O V A ' s (Harvested Streams - Invertebrate Density): 8.1 Anova results for invertebrate density in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 11.40 1 11.40 1.55 0.28 Reach (Treatment) 29.39 4 7.35 1.01 0.47 Stream 46.86 2 23.43 3.18 0.06 Stream * Treatment 44.25 1 44.25 6.01 0.02 Stream * Reach(Treatment) 43.88 6 7.31 0.99 0.45 Sampling Error 220.84 30 7.36 8.2 Anova results for Ephemeroptera density in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 10.77 1 10.77 2.69 0.20 Reach (Treatment) 12.02 3 4.01 1.48 0.30 Stream 15.84 2 7.92 1.80 0.18 Stream * Treatment 0.37 1 0.37 0.09 0.77 Stream * Reach(Treatment) 18.96 7 2.71 0.62 0.74 Sampling Error 132.03 30 4.40 8.3 Anova results for Plecoptera density in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 15.76 1 15.76 42.34 0.003 Reach (Treatment) 1.49 4 0.37 0.08 0.98 Stream 0.04 1 0.04 0.02 0.89 Stream * Treatment 0.91 2 0.46 0.23 0.80 Stream * Reach(Treatment) 26.48 6 4.41 2.20 0.07 Sampling Error 60.12 30 2.00 8.4 Anova results for Diptera density in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 44.75 1 44.75 10.59 0.05 Reach (Treatment) 12.67 3 4.23 0.42 0.74 Stream 9.09 2 4.55 2.67 0.09 Stream * Treatment 12.29 1 12.29 7.21 0.01 Stream * Reach(Treatment) 70.20 7 10.03 5.88 <0.001 Sampling Error 51.14 30 1.71 120 A P P E N D I X 9 - A N O V A ' s (Harvested Streams - Invertebrate Richness): 9.1 Anova results for Order level richness in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 1.76 1 1.76 11.32 0.028 Reach (Treatment) 0.62 4 0.16 0.29 0.87 Stream 0.65 2 0.32 0.61 0.58 Stream * Treatment 0.09 1 0.09 0.17 0.70 Error 3.20 6 0.53 9.2 Anova results for Ephemeroptera family level richness in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 0.11 1 0.11 0.34 0.59 Reach (Treatment) 1.34 4 0.34 4.46 0.052 Stream 0.30 1 0.30 3.95 0.09 Stream * Treatment 0.95 1 0.48 6:33 0.03 Error 0.45 6 0.08 9.3 Anova results for Plecoptera family level richness in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 0.05 1 0.05 0.15 0.72 Reach (Treatment) 1.18 4 0.30 1.52 0.31 Stream 0.32 1 0.32 1.64 0.25 Stream * Treatment 0.69 2 0.35 1.77 0.25 Error 1.17 6 0.20 9.4 Anova results for Diptera family level richness in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 0.00 1 0.00 0.01 0.94 Reach (Treatment) 0.41 4 0.10 0.47 0.76 Stream 0.78 2 0.39 1.82 0.24 Stream * Treatment 0.31 2 0.15 0.71 0.53 Error 1.29 6 0.22 121 APPENDIX 10 - ANOVA's (Harvested Streams - Invertebrate Diversity): 10.1 Anova results for diversity of invertebrate communities at the order level (diversity measured as Simpson's Reciprocal Index) in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 0.15 1 0.15 0.15 0.72 Reach (Treatment) 1.04 4 1.04 5.90 0.03 Stream 2.86 2 1.43 8.14 0.02 Stream * Treatment 1.30 1 1.30 7.37 0.04 Error 1.06 6 0.18 10.2 Anova results for diversity of Ephemeroptera genera (diversity measured as Simpson's Reciprocal Index) in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 0.01 1 0.01 0.03 0.88 Reach (Treatment) 1.32 4 0.33 5.35 0.04 Stream 0.30 1 0.30 4.81 0.07 Stream * Treatment 0.87 2 0.43 7.03 0.03 Error 0.37 6 0.06 10.3 Anova results for diversity of Plecoptera genera (diversity measured as Simpson's Reciprocal Index) in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 0.00 1 0.00 0.00 0.99 Reach (Treatment) 1.18 4 0.30 0.71 0.61 Stream 0.47 1 0.47 1.12 0.33 Stream * Treatment 0.90 2 0.45 1.08 0.40 Error 2.50 6 0.42 10.4 Anova results for diversity of Diptera genera (diversity measured as Simpson's Reciprocal Index) in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 0.19 1 0.19 7.65 0.05 Reach (Treatment) 0.01 4 0.03 0.13 0.97 Stream 0.01 2 0.01 0.04 0.84 Stream * Treatment 0.28 1 0.14 0.72 0.53 Error 1.16 6 0.19 122 APPENDIX 11 - ANOVA's (Harvested Streams - Organic Matter): Anova results for A F D M (Organic Matter) in the harvested streams. P-value of <0.05 indicates significant results. Effect Sum-of-Sauares df Mean-Sauare F-ratio P Treatment 0.01 1 0.01 0.13 0.73 Reach (Treatment) 0.27 4 0.07 0.30 0.87 Stream 0.02 1 0.02 0.18 0.67 Stream * Treatment 0.49 2 0.24 1.86 0.17 Stream * Reach(Treatment) 1.36 6 0.23 1.73 0.15 Sampling Error 3.92 30 0.13 123