Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effects of variable removal levels of the sea urchin, Strongylocentrotus franciscanus, on near-shore… Mooney, Robert C. 2001

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

Item Metadata

Download

Media
831-ubc_2001-715086.pdf [ 8.44MB ]
Metadata
JSON: 831-1.0090721.json
JSON-LD: 831-1.0090721-ld.json
RDF/XML (Pretty): 831-1.0090721-rdf.xml
RDF/JSON: 831-1.0090721-rdf.json
Turtle: 831-1.0090721-turtle.txt
N-Triples: 831-1.0090721-rdf-ntriples.txt
Original Record: 831-1.0090721-source.json
Full Text
831-1.0090721-fulltext.txt
Citation
831-1.0090721.ris

Full Text

T H E EFFECTS OF V A R I A B L E R E M O V A L L E V E L S OF T H E SEA URCHIN, Strongylocentrotus franciscanus, O N N E A R - S H O R E R O C K Y COMMUNITIES IN T H E TRADITIONAL TERRITORY OF T H E HESQUIAT FIRST NATION by ROBERT C. M O O N E Y B.Sc, California State Polytechnic University, Pomona, 1994 A THESIS SUBMITTED IN PARTIAL F U L F I L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF DOCTOR OF PHILOSOPHY in T H E F A C U L T Y OF G R A D U A T E STUDIES (Faculty of Forestry; Department of Forest Sciences; Centre for Applied Conservation Biology) We accept this thesis as conforming to the required standard  T H E UNIVERSITY OF BRITISH C O L U M B I A August 2001 © Robert C. Mooney, 2001 All rights reserved. This work may not be reproduced in whole or in part, by photocopy or other means, without the permission of the author.  In  presenting this  degree  at the  thesis  in  University of  partial  fulfilment  of  of  department  this thesis for or  by  his  or  requirements  British Columbia, I agree that the  freely available for reference and study. I further copying  the  representatives.  an advanced  Library shall make  it  agree that permission for extensive  scholarly purposes may be her  for  It  is  granted  by the  understood  that  head of copying  my or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department  of  The University of British Columbia Vancouver, Canada  DE-6 (2/88)  Abstract The shallow subtidal regions near Hot Springs Cove, Vancouver Island, British Columbia are characterised by large rocky areas dominated by the red sea urchin, Strongylocentrotus franciscanus. S.franciscanuswere removed at three sites with four levels of urchin removal per site. Manipulations of urchin density were maintained throughout the experiment and monitored seasonally for two seasons pre-treatment and seven seasons post-treatment. The manipulations resulted in increased gonad indices of remaining red sea urchins and caused the conversion of urchin dominated subtidal regions into kelp dominated communities with greater fish abundance. As well, a feeding experiment illustrated that the depressed gonad indices of field-collected urchins resulted from limited food resources in areas of high urchin density. The removal of sea urchins at all levels resulted in a rapid increase in the presence of laminarialean algal species (kelps). Study plots where all urchins were removed developed a dense understory and seasonal canopy of kelps with little bare rock remaining. Intermediate levels of removal resulted in a mosaic of smaller urchin-dominated and kelp-dominated patches. Control plots tended to maintain the urchin-dominated barrens-state throughout the study period. Sea urchins that were fed M. integrifolia, during the feeding experiment, showed 2.9 and 2.4 times greater gonadal development by weight than urchins collected from the field before and after the feeding trial, respectively. These results suggest that food limitation is an important factor in the gonadal development of this urchin population. Findings suggest that with supplemental, feeding, the resource base of sea urchins could be expanded to include barren habitats.  ii  The removal of S. franciscanus at all levels resulted in an increase in fecundity (measured as gonad index) for the sea urchins that remained, as well as for the urchins that reinvaded the total removal plots. All study plots showed an increase in gonad index over time, but the increase in gonad index was statistically greater for urchins in removal plots versus the control plots. The results indicate that small isolated urchin removals can have measurable effects on the fishery value of nearby urchins. Of the seven fish species monitored, pile perch, striped seaperch, kelp perch and black rockfish were most associated with kelp forest habitat. Kelp greenling, lingcod and copper rockfish showed no association with kelp forest habitat. The experimental approach taken indicates that sea urchin removal, and subsequent kelp growth, determined relative fish abundance. High densities of S.franciscanusappear to be responsible for the absence of kelp forest habitat in the region, the depressed fecundity (gonad index) of S. franciscanus, and the abundance of some fish species. The effect of urchin removal is discussed in regards to the implications for ecological theory as well as sea urchin fisheries management.  Table of Contents Abstract ii Table of Contents iv List of Tables vi List of Figures viii Acknowledgements xii Preface xiii 1. Introduction 1 Cultural context 1 Red sea urchin biology and ecology 12 Description 12 Distribution and patterns of abundance 15 Life history and reproduction 16 Ecology 18 Fishery management in British Columbia 19 Sea urchin fishery history and background •„ 19 Fishery methods and restrictions 24 Products and marketability 26 Thesis Overview 27 References 29 2. Sea Urchins and Kelp: Effects of Removing Urchins 38 Abstract : 38 Introduction '. 39 Methods and materials 43 Study area.; 43 Experimental design 44 Sea urchin surveys 44 Treatments - urchin density manipulation 45 Kelp surveys 47 Data analysis 47 Results : 49 Urchin Densities 49 Kelp Densities 57 Kelp Diversity 68 Urchins vs. Kelp 70 Discussion 74 Conclusions and implications 83 References 85 3. Gonadal enhancement of red sea urchins, Strongylocentrotus franciscanus, taken from barrens habitat in British Columbia 93 Abstract 93 Introduction 94 Materials and Methods -. 95 Results 97 Discussion 101 References 103 4. Urchins and Urchins: Effects of Removing Urchins on the Quality of Remaining Urchins 107 Abstract 107 Introduction 107 Methods and materials 109 Field 109 Laboratory 109 Analysis 110 Results 111  Discussion Conclusions and implications References 5. Urchins and Fish: The Influence of Urchin Removal on the Abundance of Fish Abstract Introduction Methods and materials Fish Surveys Data Analysis Results Embiotocids Hexagrammids Scorpaenids Discussion Embiotocids Hexagrammids Scorpaenids Conclusions and implications References 6. Summary Major findings management Implications References I. Appendix  :  122 126 128 131 131 131 134 134 135 136 136 137 138 147 147 149 150 152 153 156 157 159 165 170  List of T a b l e s Table 1-1. Previous experimental studies on urchin-algal interactions and the observed effects of removal, exclusion, or addition of sea urchins on algal abundance and composition. Experiment type codes are T= Total urchin removal, P= Partial urchin removal, A= Urchin addition, C= Caged (addition or exclusion of urchins).  9  Table 1-2. Analysis of red sea urchin roe nutritive components (percent of total roe weight). Modified after Kato and Schroeter (1985).  14  Table 1-3 Total allowable catch (TAC) for the entire B.C. coast and actual landings of red sea urchins (tonnes) from Clayoquot Sound and the entire coast between 1971 and June 1998.  23  Table 2-1. Split-plot analysis of variance for logio (data +1) of urchin densities. Treatment effects tested using the pseudo f-test. Error d.f. for treatment effects is 5.43 (refer to Hicks 1982).  54  Table 2-2. Species of kelp encountered throughout the study listed as a proportion of occurrence in samples over the total of all observations in all time periods.  61  Table 2-3. Split-plot analysis of variance for log (data +1) of kelp densities. Treatment effects tested using the pseudo f-test. Error d.f. for treatment effects is 6.23 (refer to Hicks 1982).  62  10  Table 2-4. Kelp diversity (Simpson 1949) for each site and mean for all sites at termination of the experiment. Numbers in parentheses are the total number of sub-samples taken from the total kelp encountered at each plot. Cells marked ' — ' represent an absence of kelp for that species. Letter codes are: Cs = C. costata, L m = L. bongardiana, M c = M integrifolia, N r = N. luetkeana, Pt = P. californica. Treatment labels 'total', 'low', 'med', and 'control' refer to the amount of urchin removal as total, low, medium, or no removal, respectively. 69 Table 4-1. Split-plot analysis of variance of urchin gonad indices (acrsin transformed). Treatment effects tested using the pseudo f-test. Error d.f. for treatment effects is 2.30 (refer to Hicks 1982).  113  Table 4-2. Two-way A N O V A for the mean increase in gonad index (arcsin transformed) between summer 1997 and summer 1998 for the three treatment levels. 115 Table 4-3. Two-way A N O V A for the mean increase in gonad index (arcsin transformed) between summer 1997 and summer 1999 for the three treatment levels. 116 Table 4-4. Fisher L S D tests for treatment differences observed in the A N O V A results of Table 4-2 and Table 4-3. Mean increase per treatment + standard error of the mean. Lower case letters denote means that are the same. Upper means are increases between summer 1997 and summer 1998 (t .i) and the lower means are increases between summer 1997 and summer 1999 (t .!).  117  Table 4-5. Split-plot analysis of variance for dissected urchin test diameters. Treatment effects tested using the pseudo f-test. Error d.f. for treatment effects is 3.41 (refer to Hicks 1982).  118  2  3  Table 4-6. Contingency table analysis of observations of urchins displaying a starved condition as noted by the presence of dark brown gonads for all dissected urchins before and after implementation of treatments. Yates corrected Chi-square = 34.44, p < 0.001. 121 Table 5-1. Classification of fish enumerated during this experiment by family name, common name, and genus and species.  140  Table 5-2. Split-plot analysis of variance of fish density (+ 0.001 logi transformed) for the three families of fish encountered during the study. Treatment effects tested using the pseudo f-test. Error d.f. for treatment effects is 4.42,4.45 and 5.50 for the Embiotocidae, Hexagrammidae and Scorpaenidae respectively (refer to Hicks 1982). 0  141  Table 5-3. Calculated mean density of fish ± standard error of the mean per 100 m for each of the families and species enumerated during the experiment for every sample period except for fall 1998. Rockfish have been subdivided into young of year (yoy) (<7 cm) as well as the two size classes small (8-19 cm) and large (>19 cm). Cells with "—" contained no fish for that group by time combination. 142 2  Table 5-4. Chi-squared goodness of fit between numbers of fish observed in kelp containing vs. non-kelp containing study plots for the summers of 1998 (year 2) and 1999 (year 3). Kelp plots were all plots where some level of urchin removal was conducted and non-kelp plots were the three control plots.  146  Table I-1. Definitions used for substrate classification.  174  Table 1-2. Urchin densities (m" ) and respective removal levels in parenthesis (cm) for all time periods. Removal levels were not necessary for Summer 1997 because treatments were not implemented until after Fall 1997. Removal levels were also not necessary at the termination of the experiment (time 33). Removal size is the minimum size for removal to achieve the required treatment density following the sampling period. N / A used for total removal because the minimum size is 0.0. N / A used for control plots as no urchins were removed. The term "none" refers to cases where no action was taken because the sample density was close to or below the required treatment density.  179  Table 1-3. Sampling seasons and dates samples were collected during the study.  188  2  Table 1-4. Biomass (wet-g /m ) + standard error of the mean for each site by treatment combination at the termination of the experiment (September 1999). In parenthesis is density (stipes / m ) ± standard error of the mean (n = 36 for each experimental unit). " — " denotes absence within the appropriate experimental unit for this time period. 2  2  189  List of F i g u r e s Figure 2-1. S. franciscanus density as a function of time for each treatment. Series represent each of the three study sites. Bars are ± one standard error of the mean (n = 18 for Total and Control; n= 36 for Intermediate). Vertical lines after Fall 1997 represents the implementation of treatments. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3).  53  Figure 2-2. S.franciscanusdensity as a function of time averaged across sites (N = 3). Bars are ± one standard error of the mean (n = 54 for Total and Control; n = 108 for Intermediate). Vertical line represents the time of treatment implementation. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 55 Figure 2-3. S. franciscanus density as a function of site with treatments plotted as series. Bars are ± one standard error of the mean (n = 18 for Total and Control; n = 36 for Intermediate). Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively.  56  Figure 2-4. Kelp density as a function of time for combined kelp species in each treatment. Series represent each of the three study sites. Bars are ± one standard error of the mean (n = 18 for Total and Control; n = 36 for Intermediate). Vertical lines after fall 1997 represents the implementation of treatments. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding each sampling season is given in the Appendix (Appendix; Table 1-3).  63  Figure 2-5. Kelp density as a function of time for all kelp species averaged across sites. Bars are ± one standard error of the mean (n = 54 for Total and Control; n = 108 for Intermediate). Vertical line represents the time of treatment implementation. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 64 Figure 2-6. Kelp density as a function of time for understory kelp species averaged across sites. Bars are ± one standard error of the mean (n = 54 for Total and Control; n = 108 for Intermediate). Vertical line represents the time of treatment implementation. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 65 Figure 2-7. Kelp density as a function of time for canopy kelp species averaged across sites. Bars are ± one standard error of the mean (n = 54 for Total and Control; n = 108 for Intermediate). Vertical line represents the time of treatment implementation. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 66 Figure 2-8. Density as a function of site for all kelp species with treatments plotted as series. Bars are ± one standard error of the mean (n = 18 for Total and Control; n = 36 for Intermediate). Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively.  67  viii  Figure 2-9. Kelp density as a function of urchin density per site for post-treatment data collection beginning six months after implementation of treatments and continued seasonally until project completion (n = 24). Vertical bars represent a visual estimate of the occurrence of an urchin threshold whereby removal of urchins beyond this point results in the recruitment of kelp. Note that the scatterplot for Site 1 varies in scale from Sites 2 and 3.  72  Figure 2-10. Login (data+1) of kelp density as a function of the log (data+1) of urchin density per site for post-treatment data collection beginning six months after implementation of treatments and continued seasonally until project completion (n = 24). Solid line represents the least squares linear regression. Dotted lines are 95% confidence intervals for the regression lines. Note that the scatterplot for Site 1 varies in scale from Sites 2 and 3.  73  Figure 2-11. Illustration of potential connections between the alternate community types. As urchin abundance increases, the system may or may not flip from kelp forest to urchin barren dependent upon the type and intensity of perturbations. Once urchin barrens are established they become stable, preventing kelp recruitment even if densities drop below the prior levels when the system was a kelp forest. Therefore, kelp forests can be maintained at a variety of urchin densities, but can only be initiated at relatively low urchin densities.  79  Figure 3-1. Results of feeding experiment compared to field populations before (March; n = 45) and after (June; n = 30) the supplemental feeding (n = 43). Data are presented as means ± 1 SE for horizontal test diameter, total wet mass, gonad wet mass, and gonad index. Values above error bars in the gonad index plot are the proportions of individuals with dark brown gonads in each sample.  99  10  Figure 3-2. Gonad index (left) and gonad wet mass in grams (right) versus horizontal test diameter (cm) for each of the urchin groups. Lines represent the fitted regression equations with 95% confidence intervals. Sample sizes are 45, 30, and 43 for the March, June and supplemental feeding urchin groups respectively. 100 Figure 4-1. Gonad index of S. franciscanus as a function of time for each of the three treatments averaged across study sites. Bars are ± one standard error of the mean (sample size varies between 5 and 9 for Total and Control, and between 12 and 18 for Intermediate). Vertical bar represents the time at which urchin removals were initiated (December 1997). Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 114 Figure 4-2. Average urchin test diameters for dissected urchins from each treatment throughout the duration of the experiment. Bars are ± one standard error of the mean (sample size varies between 5 and 9 for Total and Control, and between 12 and 18 for Intermediate). Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 119 Figure 4-3. Site by treatment interaction observed for test diameters of dissected urchins. Bars are ± one standard error of the mean (sample size varies between 66 and 152). Treatments were total urchin removal (Total), intermediate levels of urchin removal (Intermediate) and no removal of urchins (Control).  120  Figure 5-1. Calculated density of fish per 100 m for combined members of the Embidtocidae for each time period sampled. Bars are ± one standard error of the mean (sample size is 27 for Kelp and 9 for Control). A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3).  143  2  ix  Figure 5-2. Calculated density of fish per 100 m for combined members of the Hexagrammidae for each time period sampled. Bars are ± one standard error of the mean (sample size is 27 for Kelp and 9 for Control). A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 2  144  Figure 5-3. Calculated density of fish per 100 m for combined members of the Scorpaenidae for each time period sampled. Bars are ± one standard error of the mean (sample size is 27 for Kelp and 9 for Control). A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 145 2  Figure 1-1. Map of Vancouver Island, British Columbia showing the distribution of the Nuu-Chah-Nulth group of First Nations within the dashed line.  171  Figure 1-2. Study site locations near Hot Springs Cove, Vancouver Island, Canada. Latitude and longitude for Barney Rocks is 49° 20' 44" N , 126° 16' 41" W. Inset shows placement of Hot Springs Cove (HSC with arrow) on Vancouver Island. 172. Figure 1-3. Diagram of experimental unit sampling design. A l l sampling components in the diagram were conducted on each of the three sample transects within a study plot. The design is systematic with random starts utilised for the side of the line to begin kelp sampling as well as the initial bearing for transects. A t each line urchins were sampled within belt transects that were subdivided into three 4-m sections on the left and right side of each line. Kelp was sampled within 0.1-m quadrats alternating to the left and right of the transect line at each meter point. Treatments were carried out within a 15-m diameter treatment zone with sampling occurring within a 3-m buffer. Drawing not to scale. 173 2  Figure 1-4. Sample calculation of sea urchin removal size necessary to achieve a given treatment level. The highlighted size (8.2 cm) is the minimum size limit for removal that would result in 1 urchin / m remaining after removal.  2  175  Figure 1-5^. Urchin size class distributions for each treatment by time combination at Site 1. Histograms marked " N D " represent plots for which urchin diameters were not collected during that season. Remaining histograms without observations were due to an actual lack of sampled urchins.  176  Figure 1-6. Plot layout for Site 1 indicating treatment allocation, approximate depth at plot centre and proportion of availability for the substrate categories. Plots measure 30 m in diameter. Layout not to scale. For substrate category definitions see Table I-1.  180  Figure 1-7. Plot layout for Site 2 indicating treatment allocation, approximate depth at plot centre and proportion of availability for the substrate categories. Plots measure 30 m in diameter. Layout not to scale. For substrate category definitions see Table I-1.  181  Figure 1-8. Plot layout for Site 2 indicating treatment allocation, approximate depth at plot centre and proportion of availability for the substrate categories. Plots measure 30 m in diameter. Layout not to scale. For substrate category definitions see Table I-1.  182  Figure 1-9. Wet weight as a function of total length for Costaria costata. Weights were logio transformed to linearise the exponential relationship between total length and weight. Length measured to nearest centimetre by use of a ruler and weight measured to nearest tenth of a gram by use of a spring scale. Solid line represents the least-squares regression and dotted lines are the 95% confidence intervals for the regression (n = 21). Specimens collected in August arid September 1999. 183 Figure I-10. Wet weight as a function of stipe length for Laminaria bongardiana. Lengths and weights were logio transformed to linearize the relationship between stipe length and weight. Length measured to nearest 0.5-centimeter by use of a ruler and weight measured to nearest tenth of a gram by use of a  spring scale. Solid line represents the least-squares regression and dotted lines are the 95% confidence intervals for the regression (n = 326). Specimens collected in August and September 1999. 184 Figure I-11. Wet weight as a function of stipe length for Pterygophora californica. Lengths and weights were logi transformed to linearize the relationship between stipe length and weight. Length measured . (stipe base to meristem band) to nearest 0.5-centimeter by use of a ruler and weight estimated to nearest tenth of a gram by use of a spring scale. Solid line represents the least-squares regression and dotted lines are the 95% confidence intervals for the regression (n = 118). Specimens collected in August and September 1999. 185 0  Figure 1-12. Wet weight as a function of stipe length for Nereocystis luetkeana. Lengths and weights were logio transformed to linearize the relationship between stipe length and weight. Length measured to nearest 1.0-centimeter by use of a ruler and weight measured to nearest 10-grams by use of a spring scale. Solid line represents the least-squares regression and dotted lines are the 95% confidence intervals for the regression (n = 85). Specimens collected in August and September 1999. 186 Figure 1-13. Wet weight as a function of frond length for Macrocystis integrifolia. Length measured to nearest 1.0-centimeter by use of a ruler (to stipe terminus) and weight measured to nearest 10-grams by use of a spring scale. Solid line represents the least-squares regression and dotted lines are the 95% confidence intervals for the regression (n = 39). Specimens collected in August and September 1999. 187 Figure 1-14. General location of dives to assess the impacts of range expansion of sea otters in summer 1999. Inset shows location of the Hesquiat Peninsula (HP) in relation to Vancouver Island, B C .  190  Acknowledgements This work was conducted as a joint venture between the Centre for Applied Conservation Biology at the University of British Columbia and the Hesquiat First Nation's Management for a Living Hesquiat Harbour. I thank my academic advisor; Dr. Fred Bunnell and committee members Drs. Robert De Wreede and Scott Hinch for help in a variety of ways. I thank Sophie Boizard, Rufus Charleson, Diana Dobson, Joanna Dojillo, Josh Fehr, Laura Froc, Scott and Georgie Harrison, Kim Lisgo, Paul Lucas, James McCormick, Russ Markel, Pam Mickey, Renel Mitchell, Susan Paczek, Tom Reid, Al Sabas, Rob Serrouya, Nyree Sharp, and Michelle Tung for field and lab assistance. I thank Drs. Tony Kozak, Valerie LeMay, Peter Marshall, and David Moriarty for statistical advice. I also thank Bernard Charleson and Sam Mickey who donated space and equipment for the experiment. Steven Charleson, Sue Charleson, and Jackie Johnson provided support through administrative functions. I thank Sennen Charleson for helping to develop my relationship and research with the Hesquiat. My wife, Joanna, allowed me to miss important occasions while I was busy in the field. Dr. Keith Arnold and Scott Harrison supplied much needed prodding to help me when I fell down. I thank the membership of the Hesquiat First Nation for making me at home in their community. In addition to my academic committee, I thank Dr. Megan Dethier, Dr. Scott Harrison, James McCormick, and Dr. Stephen Schroeter for taking the time to review and discuss my dissertation. This work was partly funded by the Henry P. Kendall foundation through a grant to the Centre for Applied Conservation Biology, by a grant to conduct community-based research from the Long Beach Model Forest, Ucluelet, BC, and by economic development funds supplied by the province of British Columbia and administered by the Hesquiat First Nation. Did I remember to thank Kim Lisgo? She sure did help out a lot.  Preface During the research and writing of this dissertation I learned much more than the material presented in the following pages. I had a unique opportunity to combine the best of Hesquiat knowledge with the techniques of western science. In communicating this experience, I have failed. I can never convey to other scientists what it was like working and living with the Hesquiat. They have a culture that is deeply rooted within the area and the resources with which they live. Their culture is living, evolving, and able to incorporate new technologies and ways of thinking without disrupting its foundations. Their oral tradition and teachings reflect their kinship with nature and present a model from which the rest of the world could benefit. The Hesquiat have helped me to grow and mature as an individual, a member of society, and as a researcher. The friendship and actions of some band members have become a permanent component of my life. Although I have done little to convey the above feelings in the thesis, this is a reflection of my talents rather than a lack of appreciation. I simply cannot appropriately enlighten others to the depth of learning and the feelings shared during this research. Any attempt at such would not be fair to the Hesquiat. They have a rich ecological history and a culture that is best experienced person to person. I have included the following story to help illustrate the cultural connection between the Hesquiat and the environment. Three young women were down on the beach drying salmon. Raven came along and wanted their salmon, so he kept asking them if they were afraid to be there by themselves—if they were afraid of bears, or wolves of other such animals. They kept saying "no" to everything he asked them about until he said "owls." At this they said, "Oh, don't even talk about owls to us; we are afraid of owls." Raven went away, but hid in the bushes nearby and began to imitate owl sounds. The women were so frightened they ran away into the woods. They kept running until they came partway up the side of a hill. They were so tired, they decided to stop and rest. They said to themselves, "We'd  better stand here now on the side of the mountain; they will call us '.alhmapt'." 'And they turned into yellow-cedar trees. Raven snuck out and ate all their dried salmon. That is why yellow-cedars are always found on the mountainsides, and why they are such nice looking trees, with smooth trunks and few branches, because they used to be attractive young women with long shining hair. [Told by Alice Paul of Hesquiat and copied from Turner (1997, cited in Chapter 1) with the blessing of Larry Paul of Hesquiat]. 1  For me, the above story represents critical components of Hesquiat knowledge and morality of the Hesquiat. Within this short story we learn about a natural pattern explaining the occurrence of cedar trees as well as a why the trees look the way they do. Perhaps most importantly, we see a link between humans and the environment. By considering themselves part of the natural world with their ancestors being part of the plants and animals around them, the Hesquiat teach future generations to respect the natural world because all are related. The Hesquiat, through their resource management unit, Management for a Living Hesquiat Harbour, seek to become the lead agency when working with outside agencies on resource management issues in their traditional territory (S. Charleson pers. comm.; cited in Chapter 1). I hope that with this document and the years of working with Hesquiat fisheries and band members, I have contributed as much to them as they have to me. Through this research, I hope doors have been opened that will allow the Hesquiat to more effectively work with scientists and resource managers in the future. I know that I am now better equipped to tackle resource management issues at levels beyond the mere management of J  species.  xiv  1. Introduction This thesis grew out of a request by the Hesquiat First Nation, part of the Nuu-ChahNulth (see Appendix; Figure 1-1), for assistance in evaluating the potential for a red sea urchin (Strongylocentrotus franciscanus) fishery in Hesquiat traditional territory. Such an evaluation had to be respectful of Hesquiat ways of gathering knowledge while retaining the scientific credibility necessary to engage federal management agencies. This chapter provides an overview of three broad elements providing context for the research: cultural context, the biology and ecology of red sea urchins, and the approach to urchin fisheries management in British Columbia. After providing this broad context, I summarise the way in which majorfindingsare presented in the chapters following. CULTURAL CONTEXT Graziano and Raulin (1993) described six ways that knowledge might be obtained. These are tenacity, intuition, authority, rationalism, empiricism, and science. Each method differs in the amount of thought processing and the level of scepticism required before raw information is considered knowledge. With tenacity, information that has been around for a long time, in the form of stories or legends, is considered true and accepted as knowledge. Intuition uses extra-sensory perception, dreams, or visions to gain knowledge with no apparent sensory input. Authority enables one to consider information as knowledge because the information comes from a respected source in a position of authority, such as an elder, a priest, or a doctor. Rationalism derives knowledge through the reasoning process alone, with conclusions derived logically, e.g. if A=B and if B=C, then A=C. Empiricism requires one to make observations of the world and contrast alternative explanations through experiments or comparisons. Science is a process of gaining knowledge through the interplay of rational  1  thought and empiricism using critical thought and scepticism to propose alternative explanations. Historically, many First Nations, including the Hesquiat, used tenacity, intuition, and authority to gain knowledge about their surroundings. In Hesquiat culture, tenacity is reflected in their oral tradition, and the use of stories to explain natural patterns (see preface). Their oral tradition also is valuable for transferring ancestral and resource-based knowledge through generations. The Hesquiat use this technique to provide information on family histories, relationships and land holdings, as well as where people gathered food, what foods they ate, and the spiritual significance of various places. The Hesquiat accept this knowledge because it has been successfully used by their ancestors through centuries of close associations with the land. Dreams and visions also characterise the way many First Nations gain knowledge about the world. These components of intuition enable the Hesquiat spiritual connection , with the land and other species. Through dreams and visions many First Nations see connections to their ancestors in both the physical environment and the plants and animals around them. This perception is probably a strong influence in the Nuu-Chah-Nulth notion of "hishuk ish ts' awalk" or "everything is one" (Clayoquot Sound Scientific Panel 1995). This single idea alone has important implications for how scientists and First Nations frame research questions and is discussed further below. Finally, authority plays an integral role in the acquisition of knowledge within the Hesquiat culture. Elders and band leaders hold positions of high esteem, and the rest of the band commonly seeks their wisdom. Using the knowledge-gathering techniques of tenacity and intuition, elders impart Hesquiat traditions, knowledge, legends, and spiritual  relationships with the environment to the rest of the band. This type of knowledge transfer is particularly relevant today, as the more influential culture becomes a predominant feature in the lives of many young First Nation's individuals. Scientists have explored a variety of philosophies and techniques to gain knowledge (Popper 1963, Piatt 1964, Chamberlin 1965, Hilborn and Stearns 1982, Quinn and Dunham 1983, reviewed by Wenner 1989). Within science there is a broad division between scientists on the best way to achieving knowledge. Some argue that many of the great advances in science were due to the technique of exploring multiple working hypotheses, with advancement by rejection of most hypotheses, thus leading one to the ultimate 'truth' (Piatt 1964, Chamberlin 1965, Popper 1963). However, some ecologists have argued that this ideal method can rarely be used because multiple processes, working jointly and varying at different temporal and spatial scales, structure ecological communities (Hilborn and Stearns 1982, Quinn and Dunham 1983). Underwood (1997) attempted to bring the various views within ecology into focus by subsuming approaches taken within each group. He stressed that ecologists need to clearly state alternative models that incorporate multiple interacting processes, then test how the models can be mutually exclusive or combined to explain ecological observations. In science, testing, verification and re-examination, and rigorous experimental analysis are critical to gaining knowledge. No single study, regardless of the outcome, is sufficient to declare a problem solved (Underwood 1997). In comparison to the Hesquiat means of acquiring knowledge, scientists require more structured verification before accepting information as knowledge. The over-arching paradigm is that science combines rational thought and empiricism, using critical thought and scepticism to propose and investigate alternative explanations, or the degree of interaction  3  among differing explanations. Within this framework, tenacity, intuition, and authority play relatively insignificant roles. In science, the origin or person supporting an idea is less important than the testing and scrutiny the idea has gone through to gain acceptance (but see Kuhn 1970). Another distinction between the Hesquiat and western science is one of semantics. The notion of "hishuk ish ts' awalk" is congruent with the scientific study of ecology. Although the Hesquiat may elevate this term to a more literal and spiritual level, the fact remains that they believe in a complete connection throughout the natural environment and that impacts at any level can be felt throughout the system. In ecology, this idea is a generally accepted notion; however, the extreme viewpoints on this matter (Lovelock 1995) are still met with criticism. Semantics enter in scientists' ability to test hypotheses at the level of "hishuk ish ts' awalk". To obtain knowledge, ecologists often study single species or work within sub-components of the larger ecosystem. This dividing of the system seems inappropriate to the Hesquiat because they see components as only portions of a greater whole. The Hesquiat influence benefited this study by inspiring an increase in experimental spatial scale from tens to thousands of square meters and monitoring the effects of urchin removal on varying ecosystem components (e.g. urchins, kelp, and fish). At the same time, the Hesquiat learned about both the value and limitations of western scientific methods. Hopefully, this work will help the Hesquiat combine aspects of their culture with western science to improve resource management in British Columbia. Perhaps, over time the results of numerous scientific enquiries, encapsulated within the broad base of Hesquiat knowledge, will help the Hesquiat manage resources at the level of "hishuk ish ts' awalk".  The preceding distinctions on knowledge gathering-techniques and ecological perspectives are included because it is important to recognise the differences among the two cultures. It is also important to remember that individuals and cultures use a variety of techniques to obtain knowledge dependent upon time, place, and situation. It could be argued that in finding suitable food and medicinal items, the Hesquiat must have used rational thought and empircism. For instance, observation of resource use by animals would indicate potential human food items. Alternately, scientists are not always the unbiased practitioners of empirical science that many believe. The background and training of an individual may influence the formulation of scientific hypotheses and alter ones willingness to accept rational alternatives to a favored hypothesis (Wenner 1989). Although the previous discussion involved a view of our differences, it will be our shared visions and goals that enable the Hesquiat, scientists, and government to successfully manage resources across cultural boundaries in the future. Working with the Hesquiat has influenced the way that I conducted the research. There have been earlier studies on echinoderms similar to this study (Kitching and Ebling 1961, North 1964, Ebert 1977, Breen and Mann 1976, Duggins 1980, Keats et al. 1990, McClanahan and Kurtis 1991, Andrew and Underwood 1993, Leinaas and Christie 1996). I was asked to conduct research specifically relevant to the Hesquiat. That request meant conducting research within their traditional territory that would provide practical information on a species of cultural and potential economic importance to the Hesquiat. Specifically, they asked, how does the removal of sea urchins by a commercialfisheryor sea otters influence the coastal resources of the traditional ecological territory of the Hesquiat First Nation? Previous studies were deemed inadequate to answer their question for three reasons.  5  First, the Hesquiat wanted data that were specific to their territory. Second, they wanted to observe the process of obtaining scientific knowledge. Finally, the Hesquiat were interested in the effects of partial urchin removal, which has been looked at in only a few studies (Andrew and Underwood 1993, North 1964, Leighton et al. 1966). These prior studies either relied on unsuccessful total removals (North 1964, Leighton et al. 1966), added urchins to cages at scales which are arguably of little use in this situation (Kitching and Ebling 1961), or were conducted in regions unrepresentative of the Pacific Northwest (Andrew and Underwood 1993, Kitching and Ebling 1961). The Hesquiat have a long history within the natural environment. Their knowledge has allowed them to persist with and in their environment for thousands of years (Woodcock 1990) and left undisturbed, would likely allow them to persist for an equally long time into the future. What, then, is the relevance of science in this context? The relevance lies in a world that has changed greatly since First Nations' contact with Europeans. This contact has resulted in the loss of many First Nations' traditions and lands in the face of the more influential culture (western culture). Western culture tends to rely on scientific methods and information to manage resources. Some of these same resources are an integral component of Hesquiat culture and welfare. Thus, many First Nations view science as a language that will enable them to communicate with the members of the more influential culture. The Hesquiat hope that scientific information and methods can be used to verify what they already know and believe, gaining them acceptance as resource managers. Science is a means by which the Hesquiat can obtain information that will aid in current treaty negotiations as well as current and future participation in land management decisions. Ultimately, the Hesquiat strive to have varying degrees of autonomy and co-management  6  with the more influential culture. Combining the best of Hesquiat ecological knowledge with western science may give the Hesquiat the tools required for achieving successful examples of sound resource stewardship. I used a manipulative approach representative of empirical science to answer the Hesquiat question, how does the removal of sea urchins by a commercial fishery or sea otters influence the coastal resources of the traditional ecological territory of the Hesquiat First Nation? Although previous work was not directly relevant to Hesquiat interests, it helped frame my questions and complements this study (e.g. Table 1-1). The manipulation involved setting and maintaining sea urchins at different densities in study areas within Hesquiat territory. I have kept the research relevant to the Hesquiat by studying the effects of this manipulation from multiple levels. I monitored the effects of urchin removal on sea urchins themselves, on the abundance and diversity of kelp species, and on the abundance of some fish species in the manipulated sites. The study of these effects was important to the Hesquiat because the total urchin removal plots would indicate how future sea otter reestablishment might affect their coastal environment. Sea otters are known to have exceptional impacts on Pacific Northwest kelp communities because of their near-total consumption of sea urchins in nearshore communities (Estes and Palmisano 1974, Duggins 1980, Estes and Duggins 1995, see next section). Intermediate urchin removal levels helped the Hesquait understand the ramifications of a local, intensively managed sea urchin fishery. I believe that this work has provided benefits to both cultures. I have expanded on the earlier studies by both manipulating urchins at biologically relevant scales (nearing the level of the urchin bed) and through herbivore removals beyond the all or nothing approach common in past sea urchin studies (Table 1-1). This research has practical significance to the Hesquiat.  7  They now understand the ramifications of removing sea urchins for commercial harvest, and have seen what much of their coast may look like if sea otters repopulate the Hesquiat traditional territory. Moreover, the Hesquiat are better prepared to engage federal management agencies on sea urchin and kelp forest management issues that affect their cultural and economic heritage.  Table 1-1. Previous experimental studies on urchin-algal interactions and the observed effects of removal, exclusion, or addition of sea urchins on algal abundance and composition. Experiment type codes are T= Total urchin removal, P= Partial urchin removal, A= Urchin addition, C= Caged (addition or exclusion of urchins). Reference Leinaas and Christie 1996  Urchin Species Strongylocentrotus droebachiensis  Location Vega Island, northern Norway  Type T  Andrew and Underwood 1993  Centrostephanus rodgersii  Cape Banks, New South Wales, Australia  T,P  Keats et al. 1990  Strongylocentrotus droebachiensis  Conception Bay, Newfoundland  T  Dean etal. 1988  Lytechinus anamesus  San Onofre, southern California  C  Fletcher 1987  Centrostephanus rodgersii  Cape Banks, New South Wales, Australia  T  Himmelman et al. 1983  Strongylocentrotus droebachiensis  Lower St. Lawrence Estuary, Canada  T  Andrew and Choat 1982  Evechinus chloroticus  Goat Island Marine Reserve, northeastern New Zealand  P  Observed Effects Total urchin removal resulted in colonisation by filamentous algae replaced by Laminaria saccharina within a few weeks. After 3-4 years, Laminaria hyperborea became increasingly dominant. Total, 66%, and 33% removal of urchins resulted in increased cover of filamentous algae. Only total removal allowed recruitment of foliose algae with species assemblages varying among replicates Total removal of urchins resulted in dominance of Alaria esculenta at 2 to 3 m and Desmarestia aculeata at 6 to 9 m. Urchin exclusion experiments showed that L. anamesus inhibited laminarian alga recruitment by killing gametophyte or microscopic sporophyte life-stages. Removal of urchins and/or limpets from boulders resulted in a rapid increase in non-crustose algae. 100% cover by algae took 18-24 months after urchin removal and only 12 months after urchin and limpet removal. Limpet-only removals were not sufficient to maintain algal communities. Total removal resulted in the establishment of Alariadominated algal assemblage within 2 years, dependent upon depth. Experimental removal of all adults (>30mm) resulted in increased abundance of Sargassum sinclairii, Eklonia radiata, and turf coralline algae, and decreased abundance of herbivorous gastropods  9  Table 1-1. Continued. Ayling 1981  Evechinus chloroticus  Goat Island Marine Reserve, northeastern New Zealand  C  Duggins 1980  Strongylocentrotus Torch Bay, franciscanus southeast Alaska, StrongylocentrotusU S A droebachiensis Strongylocentrotus purpuratus  T  Vance 1979  Centrostephanus coronatus  Ebert 1977  Strongylocentrotus Mission Bay, San Diego, southern franciscanus Strongylocentrotus California, U S A purpuratus  A  Breen and Mann 1976  Strongylocentrotus St. Margaret's droebachiensis Bay, Nova Scotia, Canada  T, A  Bird Rock, Catalina Island, southern California, U S A  T  Cages excluding urchins provided for increased cover of ephemeral algae and coralline turf algae, but no change in species composition Total removal of urchins resulted in colonisation by both annual and perennial algae in the first year, with complete dominance by the annual Laminaria groenlandica by the second year. Exclusion of urchins from grazed areas resulted in increased fleshy and erect organisms and a decrease of calcareous and encrusting forms. Observations of fenced urchins on a rock jetty resulted in extrapolations of 42 S. purpuratus or 41 S. franciscanus 1 nT (2520 g or 12,300 g wet weight, respectively) being required to maintain total barrenness. Total removal resulted in a sequence of colonising algal species with Laminaria longicruris dominating in three months. Alternately, the addition of up to 400 urchins within 0.5 / m sampling stations was ineffective at reducing kelp biomass as urchins migrated or were predated. Removal of urchins resulted in lowered distribution of ulvoids in fall and winter, and kelps in the following spring. Sea urchin exclusions resulted in abundant algal growth confounded by the exclusion of herbivorous fish. 2  Low 1975  StrongylocentrotusPuget Sound, northeastern franciscanus Pacific  T  Leighton 1971  Strongylocentrotus Point Loma, franciscanus California, U S A Strongylocentrotus purpuratus Lytechinus anamesus Strongylocentrotus Mukkaw Bay and Friday Harbor, franciscanus Strongylocentrotus Washington, U S A purpuratus  C  Paine and Vadas 1969  T  Hedophyllum sessile dominated tide pools where S. purpuratus were removed. Subtidal removals of S. franciscanus resulted in the dominance of Laminaria spp.  10  Table 1-1. Continued Port Erin, Isle of Man, U . K .  Kain and Jones 1966, and Jones and Kain 1967  Echinus esculentus  North 1964, and Leighton etal. 1966  Strongylocentrotus Abalone Cove and franciscanus Point Loma, Strongylocentrotus southern purpuratus California, U S A Lytechinus anamesus  T  T,P  Total removal resulted in decreasing the lower extent of Laminaria hyperborea Killing sea urchins resulted in algal re-establishment of previously barren areas. Suggested (from failed total removals) one S. franciscanus, 10 S. purpuratus, or 10 L. anamesus 1 m required to maintain the barrens state. Total removal resulted in 50% cover of Enteromorpha within two months and 100% after one year. Cage experiments resulted in 33-50%, 30%, and 0% cover of algae with one, three, and six urchins respectively within two months. One square-yard enclosure with 16 Diadema not sufficient to prevent growth of algae in the absence of herbivorous fish. 2  Kitching and Ebling 1961  Paracentrotus lividus  Lough Ine, Ireland  T,C  Randall 1961  Diadema antillarum  Virgin Islands  C  RED SEA URCHIN BIOLOGY AND ECOLOGY Description Red sea urchins are echinoderms. Other closely related members of this group include starfish, sea cucumbers, brittle stars, and sand dollars. Sea urchins have radially symmetrical, spherical bodies contained within a calcareous test that is covered with a thin epithelium and armed with spines. Red sea urchins are relatively large, achieving test diameters in excess of 16 cm and spine lengths of up to 8 cm (Sloan 1991). The colour of red sea urchins is variable, but uniform within individuals, varying between bright red and dark purple. The oral opening is located on the underside and contains a calcareous feeding structure known as "Aristotle's lantern". The lantern is composed of five plates with teeth which come together to scrape the substratum for food. The mouth leads to the digestive tract, which coils through the centre of the body cavity and empties at the anus located on the aboral surface of the test. The most prominent internal feature of sea urchins is the presence of five rows of gonadal tissue under the interambulacral plates of the test. The gonadal tissue varies in colour based primarily on food availability, ranging from dark brown in starved individuals to bright yellow and orange in fed animals (see Chapter 4). In between the gonads, under the ambulacral plates of the test, are clear gelatinous structures that are the primary component of the water vascular system. This system is important for gas exchange as well as movement, attachment, and food manipulation via the tube feet. Currently, the entire fisheries value of sea urchins lies in the gonads. Unlike other roe fisheries where eggs are the products consumed, consumers of sea urchins eat the entire gonad. There is no differentiation in marketing urchin gonads based on sea urchin sex. It is  primarily the urchins' use of the gonads as an energy and nutrient storage organ that makes it an appealing food item. During spawning, the gonads swell with excess fluid, soften, and leak gametes. This condition prevents the marketing of urchins during the spawning season. Female urchin gonads are composed of variable levels of reproductive cells based on the stage of the oogenetic cycle. The components are oogonia, primary oocytes, secondary oocytes, late oocytes, ova, and relict ova. In males the corresponding components are spermatogonia, primary spermatocytes, secondary spermatocytes, spermatids, spermatozoa and relict spermatozoa (as per Bernard 1977). In both sexes, nutritive phagocytes ingest gametes and function as glycogen repositories alternating between globulated and deglobulated phases dependent upon the phase of the oogenetic or spermatogenetic cycle or on the general fitness of the individual. It is the abundance and condition of the nutritive phagocytes that largely determines the gonadal size, colour and taste of sea urchins. Traditional Hesquiat knowledge considers that urchin roe is a quality food source, rich in energy and vitamins with an almost medicinal quality. This information is supported by scientific analysis of the gonads of red sea urchins. Sea urchin roe contains calcium, phosphorus, iron, Vitamins A, Bi, B2, B12, nicotinic acid, pantothenic acid, folic acid, and carotenes (Kato and Schroeter 1985, citing Higashi et al. 1959, 1965). The major components of red sea urchin roe are presented in Table 1-2.  Table 1-2. Analysis of red sea urchin roe nutritive components (percent of total roe weight). Modified after Kato and Schroeter (1985). Source Component  Greenfield et al. 1958  Kramer and Nordin 1979  Moisture  70l)  708  Protein  7.7  9.6  Lipid  7.6  8.3  Ash  1.6  1.5  Glycogen  1.3  Nonprotein nitrogen  0.1  0.5  Distribution and patterns of abundance Red sea urchins occur along shores of the northern Pacific, from the tip of Baja California in the south to Sitka and Kodiak, Alaska in the north and back south to the southern tip of Hokkaido Island, Japan on the Asiatic coast (Kato and Schroeter 1985). Red urchins are abundant and widely distributed throughout coastal British Columbia except in sheltered habitats and on unsuitable, non-rocky substrates (Breen 1980). They occupy depths ranging from the low-intertidal zone to as deep as 125 m (McCauley and Carey 1967). Maximal abundance of 30 urchins / m are typically found in tidal passes (Breen 1980). 2  Red sea urchins often exist in high-density groups (Low 1975). Grouping may occur for various reasons. Sea urchins have been shown to cluster around algal food sources (Pearse et al. 1970, Mattison et al. 1977, Tegner and Dayton 1981) with grouping likely being the result of urchins sensing the presence of preferred food items via chemoreception (Vadas 1977). Given that urchins arefreespawners, shedding gametes into the water column for external fertilisation, increased group size and degree of aggregation has been shown to significantly increase fertilisation success (Levitan et al. 1992). Juvenile red sea urchins are known to associate with larger individuals (Low 1975, Tegner and Dayton 1977, Cameron and Schroeter 1980). Juveniles most likely experience decreased predation under the spine canopy of adults (Tegner and Dayton 1977, Cameron and Schroeter 1980, Duggins 1981). Patterns of predation or juvenile behaviour are likely driving the association with adults as opposed to site selection by larval urchins (Cameron and Schroeter 1980, Rowley 1989). Moreover, for urchins too large to be protected by direct association with an adult, general association with other adults may protect medium sized urchins by association with larger  15  urchins that have achieved a size refuge from predation (Duggins 1981, Tegner and Dayton 1981).  Life history and reproduction Female sea urchins may release between 100,000 and 2,000,000 eggs into the water column for external fertilisation by the males' sperm (Mottet 1976, but see Low 1975 for even higher estimates of production). Fertilised eggs sink during subsequent cell division stages. After passing through the blastula stage, the developing embryo begins to drift in the water column and eventually passes through the gastrula stage to become an echinoplutei larva. This process takes approximately 86 hours. The larvae are bilaterally symmetrical (as opposed to radial) and bear no resemblance to the adult form. The larvae eat microorganisms from the water column. Late in larval development the juvenile organs begin to develop; the oral half of the juvenile appears and almost "seems like a parasite absorbing nutrientsfroma host" (Mottet 1976). Once the juvenile is sufficiently developed, the larva begins seeking a suitable substratum for the juvenile stage. Sea urchin dispersal occurs primarily during the larval stage. Larvae are capable of subsisting on phytoplankton for several months while seeking suitable substrate (Johnson 1930, Bernard and Miller 1973), the latter consisting of rock surfaces covered with bacterial films (Cameron and Schroeter 1980), red algal turf or coralline algae (Rowley 1989). After settlement, metamorphosis occurs and a juvenile sea urchin with spines and tubefeet emerges within an hour. Red sea urchin recruitment levels are low in British Columbia, as compared to other populations (Bernard and Miller 1973, Breen et al. 1976, 1978, Campbell et al. 1999a). Moreover, recruitment levels are variable both among local (Tegner and Dayton 1981) and over broader spatial scales (Harrold and Pearse 1987); the factors influencing recruitment in  BC require study (Campbell et al. 1999a). Recruitment success is likely determined by fineand broad-scale current patterns, settlement site characteristics, and survival to juvenile size classes (Sloan et al. 1987). In British Columbia, red sea urchins demonstrate some gametogenic activity at the end of the first year, or about 30-mm test diameter. However, there is no indication of spawning at this size, and gamete release probably does not occur until the second year, at about 50-mm test diameter (Bernard and Miller 1973). Campbell et al. (1999a) noted that 100% of individuals over 70-mm test diameter were sexually mature. Spawning seems to follow a latitudinal gradient with spawning occurring later in northern latitudes (Bernard 1977). In British Columbia, spawning can potentially occur between March and September but is most likely during the summer months (Bernard 1977). Red sea urchin growth (reviewed in Kato and Schroeter 1985) has been studied in the laboratory (Leighton 1971), in the field via the use of cages (Swan 1961, Schroeter 1978), and in uncaged field investigations (Baker 1973, Bernard and Miller 1973, Ebert and Russell 1992). The results vary but indicate yearly growth rates rangingfrom13 to 25 mm of test diameter. Growth in sea urchins is not linear, and smaller urchins show the highest growth rates (Bernard and Miller 1973, Ebert and Russell 1992). On average, red sea urchins attain a size of between 90 and 100 mm in approximately 4-5 years (Kato and Schroeter 1985) (but see lower maximal sizes observed by Ebert and Russell 1992). Growth rate and maximum size in sea urchins has been shown to be positively affected by increased food availability (Ebert 1968, Edwards and Ebert 1991, Leinaas and Christie 1996) and food value of algae (Lemire and Himmelman 1996), and negatively by competition (Ebert 1968, 1977, Schroeter 1978, but see Duggins 1981 for interspecific facilitation among urchins). Physical site characteristics are also known to affect urchin growth rates. Decreased temperature within  an appropriate range promotes growth (Baker 1973), and spine damage due to waves and surge decreases growth (Ebert 1968).  Ecology To ecologists, sea urchins are best known because of their voracious appetite for fleshy algae (Leighton 1971, Lawrence 1975, Duggins 1980, Hughes et al. 1987, Ebeling and Laur 1988, Tegner and Dayton 1991, Leinaas and Christie 1996). The feeding activities of dense concentrations of sea urchins results in total removal of such algae, leaving behind a relatively bare rocky substratum termed "barren grounds" (Pearse et al. 1970). There are few predators of sea urchins in British Columbia and the most effective predators, capable of regulating urchin abundance, are humans (see Chapter 2) and sea otters (Breen 1980). Local extirpation of sea otters after the intrusion of Europeans in the late 18 century meant the th  removal of a "keystone" predator sensu Paine (1969) (see Watson 1993 for an historical account of European influence and the sea otter trade). The importance of sea otters in kelp forest dynamics was made evident in studies following the subsequent return of sea otters (Estes and Palmisano 1974, Laur et al. 1988, Watson 1993). Just as the arrival of Europeans had dramatic consequences on the distribution of kelp forests, so did the first people likely alter the marine ecosystems on which they depended for food (Simenstad et al. 1978). Simenstad et al. (1978) presented evidencefromkitchen middens of the Aleuts indicating that overharvesting of sea otters led to population explosions of herbivorous invertebrates such as sea urchins. They found that the presence of otter and nearshore fish remains in middens were inversely proportional to the presence of , urchin remains. The removal of sea otters in the 18 and 19 centuries has permitted th  th  18  localised population explosions of sea urchins leading to a reduced abundance of kelp forest communities (Druehl 1978, Duggins 1980, Breen et al. 1982). Kelp forests are important to physical and biological processes of nearshore environments. Kelp forests add structural complexity to the water column and provide food and habitat for a variety of vertebrate and invertebrate species (Quast 1968, Leighton 1971, Duggins et al. 1989). The increased abundance of invertebrates in kelp forests provides an important trophic link between the abundant plankton and larger consumers that feed on plankton-consuming invertebrates (Quast 1968). Physically, kelps alter and reduce currents and waves (Jackson and Winant 1983, Foster and Schiel 1985, Koehl and Alberte 1988), decrease light intensity (Pearse and Hines 1979, Reed and Foster 1984) and increase sedimentation (Eckman et al. 1989). Each of these factors can influence the recruitment of other kelp forest inhabitants (Duggins et al. 1990). Duggins et al. (1990) reported that flow velocity, sedimentation, reduced light intensity / microalgal cover all have important but variable effects on recruitment, dependent upon the invertebrate species investigated. Although the effects of kelp forest structure may vary by the species investigated, the general rule seems to be that kelps provide for increased habitat and species diversity as opposed to outside kelp forests. The loss of diversity when urchins are limiting the presence of kelp can be further magnified because sea-urchins consume benthic invertebrates when algal foods are limited (Briscoe and Sebens 1988, personal observation).  FISHERY M A N A G E M E N T IN BRITISH C O L U M B I A  Sea urchin fishery history and background In 1991, the Hesquiat First Nation implemented the program "Management for a Living Hesquiat Harbour" (MLHH). The mission of M L H H is "to protect and restore the  living resources and ecosystems of the Hesquiat territory to allow sustainable levels of harvest in the future" (Hesquiat First Nation 1992). The terms most critical to the intent of MLHH are "protect and restore" and "future". The Hesquiat are committed to avoiding the lure of immediate economic gain from resource extraction in exchange for maintaining a healthy ecosystem that can provide for the well being of future generations. I was asked to join the MLHHfisheriescrew as a research scientist in 1996, and began performing research on red sea urchins, S. franciscanus, in 1997. Sea urchins were an important and sought after food for the native people of the west coast of Vancouver Island (Ellis and Swan 1981). Purple sea urchins were known to the Nuu-Chah-Nulth as "hiix" and were collected from the subtidal and in tide pools along exposed outer coasts. Green sea urchins, or "nuuschi", were collectedfromrocks in the low intertidal or from canoe in the subtidal. Red sea urchins, "t'uts'up", were probably most prized and were collected in the very low intertidal zone and from the subtidal. Subtidal collections were usually made from a canoe and urchins were skewered with spears (Clayoquot Sound Scientific Panel 1995). The edible portions of the sea urchin are its gonads. Sea urchin gonads were eaten raw by First Nations' peoples and are still enjoyed today by them and others, particularly in Japan where they are known as uni. In global terms, the current market for sea urchins exists primarily in Japan. Numerous countries serve this market, including Canada, the United States, South Korea, Chile, North Korea, Peru, and China (Muse 1998). Within Japan, large-scale production began in the 1950s and reached production levels between 22,000 and 27,000 landed tonnes per year (fresh wet mass of whole urchins). However, Japanese production began to decline  after 1987, with demand being sufficient to increase market opportunities for other countries (Muse 1998). The red sea urchin is one of three sea urchin species commercially fished in British Columbia (DFO 1999a). However, it was not long ago that they were considered the only species of fishery potential (Breen 1980). A commercial dive fishery for green sea urchins (S. droebachiensis) was introduced in 1987 and still exists today (Muse 1998, D. Macey pers. comm.). Purple sea urchin harvests were conducted under scientific permitfrom1990 to 1992, however the current assessment of stocks indicates abundance is insufficient for a commercial fishery (DFO 1999a). Small-scale harvesting of red sea urchins with sale to Japan began on the west coast of Vancouver Island in October 1970 (Bernard and Miller 1973). Growth of the fishery was slow and remained small until 1978. The first large commercial landings, 75 landed tonnes, occurred in 1978 along the south coast of British Columbia (Muse 1998). Exceptional increases in landings occurred in the early 1980s, from 312 landed tonnes in 1982 to 3,334 landed tonnes in 1984, and continued to rise into the early 1990s. During this time period, entry into the fishery was not limited. As well, there were no consistent catch limits placed on the fishery until 1985 for the south coast, and not until 1993 for the north coast (Muse 1998). In 1990, one year before limited entry into the fishery was implemented, the fishery produced 3,158 tonnes (Muse 1998). Limited entry did not slow the catch with a maximum of 12,983 tonnes being landed in 1992. Individual quotas, implemented first by fishermen in 1994 and becoming part of the official management plan in 1996, decreased the catch but improved product quality, prices, and fishing conditions (Muse 1998). Unfortunately, data  do not exist for First Nations' sea urchin catches, however 2% of the total allowable catch is reserved for First Nations' use. Commercial catch statistics are reported in Table 1-3. Ultimately, the rise of sea otter populations in British Columbia (Watson 1993) may lead to the demise of the sea urchin fishery (see previous section on ecology). However, from a broad view, losses to this fishery may be replaced by fisheries benefits due to the revitalisation of kelp forest habitats (see Chapter 5). Moreover, although kelp harvesting has had few successes in British Columbia (R. De Wreede pers. comm.), a collapsing urchin fishery may make the industry viable.  Table 1-3 Total allowable catch (TAC) for the entire B.C. coast and actual landings of red sea urchins (tonnes) from Clayoquot Sound and the entire coast between 1971 and June 1998.  Year  TAC (tonnes)  Coast-wide Landings Clayoquot Sound (tonnes) Landings (tonnes) 474 254 1971-1973 0 139 1974-1977 121 0 1978 0 631 1079 664 0 1980 214 17 136 1981 5 312 1982 1,826 38 1983 3,334 103 1984 3,141 158 1,803 1985 3,549 285 1,500 1986 3,553 199 1,633 1987 3,334 250 1,678 1988 3,980 223 1,644 1989 4,318 215 1,668 1990 185 7,768 1,531 1991 200 14,071 1,554 1992 92 6,976 6,801 1993 111 6,568 7,440 1994 7,498 199 6,842 1995 7,184 122 6,624 1996 9,974 132 1997/1998 TAC in effect for the north coast; before this, reported TAC referred to only the south coast. Landings above the TAC before 1993 were due to unlimited north coast landings. 18-month fishing period reported due to change in fishery schedule. a  b  a  b  23  Fishery methods and restrictions Commercial sea urchin harvests are performed by crews typically consisting of two divers and a tender. Most crews utilise scuba gear, however some use hooka dive gear (surface-supplied air pumped to the diver). Divers harvest urchinsfromjust below zero tide level to depths of 15-20 m, but most harvesting occurs in less than 10 m of water. Harvesting usually begins by divers scouting for areas with suitable densities of quality sea urchins. Quality is determined by a diver opening the test of a few individuals and visually assessing the colour and size of gonads. Densities of five urchins / m or more are generally sought; the short diving time and large expense minimise economic returns at lower densities (Bernard and Miller 1973). Urchins are removed from the bottom with the use of a small aluminum rake that also helps divers estimate the size of removed urchins. Urchins are placed into mesh bags and floated to the surface where a dive tender retrieves bags and ensures that divers are continually supplied with empty bags. Dependent upon the crew and vessel size, urchins are either brought directly to port or else transferred to a larger vessel that then delivers the catch of multiple dive vessels to port. On BC's coast, acceptable landing ports as of the year 2000 were Masset, Port Edward, Queen Charlotte City, Klemtu, Bella Bella, Prince Rupert, Port Hardy, Campbell River, Kelsey Bay, Victoria, Nanaimo, Ucluelet, Coal Harbour, Sidney, Tofmo, Port McNeill and Sooke (DFO 1999b). Sea urchin harvesting is monitored and enforced through harvesters and the Department of Fisheries and Oceans Canada (DFO). Harvesters are organised via the Pacific Urchin Harvesters Association (PUHA). PUHA contracts D&D Pacific Fisheries Ltd. to provide notification, validation, biological sampling, and data services for the fishery. The DFO is responsible for determining regulations to prevent over-harvesting the resource and  for the enforcement of regulations. As of the 1999/2000 fishing season the DFO was using the following management approaches to stock management: limited entry, coast-wide and area quotas, individual quotas, and minimum size limits. Limited entry refers to the management practice by which the DFO limits the number of commercial licenses for the fishery. Currently, there are 110 licenses to harvest red urchins in British Columbia (DFO 1999a). By limiting the amount of licenses in the fishery, the DFO reduces competition among harvesters and can provide all harvesters with sufficient quantity and quality of sea urchins to support themselves. The fishery is currently considered data limited and so restrictive measures such as limited entry help to prevent over exploitation (Campbell et ai. 1999a, 1999b). Quotas also are implemented for the entire British Columbia coast. The quota system is designed to provide a conservative fixed exploitation rate. The rate is based on stock assessment information provided by scientists and fishers as well as biological information as it becomes available (Campbell et al. 1999b). Individual quotas divide the annual total allowable catch among the number of licenses in the fishery. This measure was implemented to decrease a "race" for urchins and allowed fishers to redirect attention towards fishing for quality rather than quantity. The measure has resulted in improved prices and a more manageable fishery (Muse 1998). Minimum size limits have been in effect since the initiation of the fishery in 1970; the designated size limit was 100 mm until the 2000/2001 fishing year. The 100-mm size restriction allows mature urchins approximately three spawning years before harvest (Breen 1984). Minimum size was lowered to 90 mm for the 2000/2001 fishing year, in response to market demand for smaller urchins with higher quality gonads and not necessarily to  diminishing returns. It is hoped that many larger urchins will remain as brooding stock because buyers demand smaller urchins and thus harvesters will leave larger urchins in the beds (D. Bureau pers. comm.). At this time, no upper size limit is in effect for the fishery. Working within the guidelines of the DFO, harvesting frequency and intensity is determined by buyers. Thus, there are currently no maximum size limits imposed to protect juveniles under spine canopies. The market demands, mentioned above, protect large urchins, providing smaller urchins the ability to shelter under spine canopies. Likewise, although harvesting is allowed year round, demand is low in the summer due to lower gonad quality during the spawning season. As well, the domestic Japanese market is producing quality gonads in the early summer and Japanese consumers prefer the domestic product. Thus, there is little need for enforced summer closures during the spawning season in British Columbia (D. Bureau pers. comm.) P r o d u c t s a n d marketability  Most sea urchin roe is soldfresh;however, a small amount is sometimes salted, steamed, baked, or frozen. Fresh roe requires immediate processing and shipping because the gonads can quickly spoil or deteriorate in appearance. Hence,freshroe is usually shipped by air to Japan (the largest urchin market) within 30 hours of being dropped at the dock by divers (Kato and Schroeter 1985). In Japan, the roe is generally sold to major wholesalers or through an auction system where it then makes its way throughout the country. In the first quarter of 2001 wholesale prices offreshred sea urchin roe in Japan were fluctuating between (900 - 2500 yen / 250-g tray; « $10.80-$30.00 CDN) but has been worth as much as 5,000 yen or $60.00 CDN (L. Chan pers. comm.). Harvesters during the same period received a range of $1.68-$ 1.90 CDN / landed kg of whole urchins. The large  discrepancy between harvesters and wholesalers lies in the fact that roe yield is typically between 10 and 20% of urchin weight. As well, there are considerable preparation, shipping, and tax expenses incurred by processors. Prices and demand for North American urchins are usually highest in late fall and early winter because Japanese domestic production is highest during the spring and summer while weather (monsoons) and the onset of reproduction in late summer depresses Japanese fishing during the fall. Domestically, red sea urchins have historically had little market value, although they were an important food item for First Nations in the past (Ellis and Swan 1981), and are still enjoyed as an occasional food item with important ceremonial status (personal observation). The present popularity of sushi bars (the traditional Japanese way of serving sea urchins and other delicacies) in North America is increasing the domestic market for the red sea urchin catch. However, Japan is by far the greatest consumer of North American red sea urchin roe. In the year 2000, Hi-To Fisheries in Vancouver shipped approximately 90% of its fresh urchin roe to Japan with the remaining 10% being split between Hawaii and local consumption (L. Chan pers. comm.).  THESIS OVERVIEW  In this dissertation, I examine the effects of variable sea urchin removal on kelp abundance, fish abundance, and on the fishery value of remaining urchins. The intent of the project was to explore probable impacts if a sea urchin fishery were conducted within the traditional territory of the Hesquiat First Nation. By studying their local resources before harvesting, the Hesquiat hope to develop management plans that allow for long-term, sustainable development of resources.  27  In the previous sections, I have given the background and contextual framework under which the experiments were conducted. In the following chapters, I present the results of my thesis research and discuss the theoretical and applied implications of my work. In Chapter 2,1 explore the relationship between variable urchin density and kelp diversity and abundance. The data summarised in Chapter 2 illustrate that sea urchins were responsible for the lack of kelp forest habitat in the study area. Total removal of urchins resulted in a quick and almost complete dominance of the rocky substrates by kelps. Intermediate levels of removal tended to produce a mosaic of kelp forest mixed with remnants from the previous barren state maintained by aggregations of remaining urchins. Control plots maintained the barren state throughout the study. By varying the level of urchin removal, I was able to develop regression equations that explain the functional relationship between sea urchin density and kelp density. Although the magnitude of the responses varied across sites, an exponential-decay function best fits the response of kelp density to urchin density. Such a response indicates that an urchin density threshold exists, above and below which the study area shifts towards either of two possible stable communities (urchin barren or kelp forest). Chapter 3 illustrates that sea urchins in the study area were food limited. The data in this chapter were obtained by conducting a feeding trial. For the trial, I collected and fed sea urchins for 83 days and compared gonad indices to urchins collected before and after the feeding trial. Gonad indices were greater for urchins in the feeding trial indicating that sea urchins in the study area were sufferingfromreduced fecundity due to food limitation. In Chapter 4,1 further explored the fecundity of sea urchins within the study area by comparing gonad indices across the sea urchin removal treatments throughout the duration of  the study. The results illustrate that at the beginning of the experiment, urchins in the study area were suffering from density-dependent competition for food resources. Through the removal of sea urchins, I reduced this competition and allowed for the successful recruitment of kelp. The provision of kelp, as a food source, improved the gonad indices of urchins in all treatments. I explain all increases as a result of the treatments, indicating that control plots (with little kelp) were affected by transport of kelp from urchin removal plots. These observations suggest that small, localised urchin removals can be an important enhancement tool for the sea urchin stocks in barren habitats. In Chapter 5,1 examine the relationship between members of three fish families with the sea urchin removal experiment. The kelp growth reported in Chapter 2, following removal of sea urchins, added structural complexity to the water column. Members of the Embiotocidae and Scorpaenidae were observed to associate with the presence of kelp while members of the Hexagrammidae did not. These data suggest that high densities of S. franciscanus are limiting the abundance of some fish species in the study area. The mechanism by which urchins influence fish abundance is by controlling the presence of kelp. In Chapter 6,1 conclude the thesis by discussing the implications of my work in terms of the Hesquiat and sea urchin management in general.  REFERENCES Andrew, N. L., and J. H. Choat. 1982. The influence of predation and conspecific adults on the abundance of juvenile Evechinus chloroticus (Echinoidea: Echinometridae). Oecologia 54:80-87. Andrew, N. L., and A. J. Underwood. 1993. Density-dependent foraging in the sea urchin Centrostephanus rodgersii on shallow subtidal reefs in New South Wales, Australia. Marine Ecology Progress Series 99:89-98. Ayling, A.M. 1981. The role of biological disturbance in temperate subtidal encrusting communities. Ecology 62:830-847. 29  Baker, S. L. 1973. Growth of the red sea urchin, Strongylocentrotus franciscanus (Agassiz), in two natural habitats. M.Sc. thesis. California State University, San Diego. Bernard, F. R. 1977. Fishery and reproductive cycle of the red sea urchin, Strongylocentrotus franciscanus, in British Columbia. Journal of Fisheries Resources Board of Canada 34:604-610. Bernard, F. R., and D. C. Miller. 1973. Preliminary investigation on the red sea urchin resources of British Columbia (Strongylocentrotus franciscanus (Agassiz)). Fisheries Research Board of Canada Technical Report 400:37p. Breen, P. A. 1984. Sea urchins: suitability of the present size limit. Canadian Technical Report of Fisheries and Aquatic Sciences 1774:25-51. Breen, P. A. 1980. The ecology of red sea urchins in British Columbia. Pages 3-12 in, International symposium on coastal Pacific marine life. Western Washington University, Bellingham, WA. Breen, P. A., and K. H. Mann. 1976. Destructive grazing of kelp by sea urchins in eastern Canada. Journal of the Fisheries Research Board of Canada 33:1278-1283. Breen, P. A., D. C. Miller, and B. E. Adkins. 1976. An examination of harvested sea urchin populations in the Tofino area. Fisheries Resources Board of Canada Manuscript Reports 1401:1-23. Breen, P. A., B. E. Adkins, and D. C. Miller. 1978. Recovery rate in three exploited sea urchin populations from 1972 to 1977. Fisheries and Marine Service Manuscript Report (Canada) 1446:1-27. Breen, P. A., T. A. Carson, J. B. Foster, and E. A. Stewart. 1982. Changes in subtidal community structure associated with British Columbia sea otter transplants. Marine Ecology Progress Series 7:13-20. Briscoe, C. S., and K. P. Sebens. 1988. Omnivory in Strongylocentrotus droebachiensis (Muller) (Echinodermata: Echinoidea): predation on subtidal mussels. Journal of Experimental Marine Biology and Ecology 115:1-24. Bureau, D. 2001. Stock Assessment Unit, Pacific Region. Department of Fisheries and Oceans Canada. Personal communication regarding urchin fishing schedules and market preferences based on sea urchin size. Cameron, R. A., and S. C. Schroeter. 1980. Sea urchin recruitment: effect of substrate selection on juvenile distribution. Marine Ecology Progress Series 2:243-247.  30  Campbell, A., J. Boutillier, and J. Rogers. 1999a. Discussion on a precautionary approach for management of the red sea urchin fishery in British Columbia. Department of Fisheries and Oceans Canada. Canadian Stock Assessment Secretariat Research Document 99/094. Campbell, A., W. Hajasj and D. Bureau. 1999b. Quota options for the red sea urchin fishery in British Columbia for fishing season 2000/2001. Department of Fisheries and Oceans Canada. Canadian Stock Assessment Secretariat Research Document 99/201. Chamberlin, T. C. 1965. The method of multiple working hypothesis. Science 148:754759. Chan, L. 2001. Hi-To Fisheries. Vancouver, British Columbia. Personal communication regarding sea urchin quality, wholesale pricing, processing, and shipping. Charleson, S.D. 2001. Member of the Hesquiat First Nation. Personal communication regarding mandates of Management for a Living Hesquiat Harbour. Clayoquot Sound Scientific Panel. 1995. First nations' perspectives relating to forest practices standards in Clayoquot Sound. Report 3. March, 1995. Victoria, B.C. Dean, T. A., F. R. Jacobsen, K. Thies, and S. L. Lagos. 1988. Differential effects of grazing by white sea urchins on recruitment of brown algae. Marine Ecology Progress Series 48:99-102. De Wreede, R.E., 2001. Professor, Department of Botany, University of British Columbia. Personal communication regarding the history of kelp harvesting in British Columbia. [DFO] Department of Fisheries and Oceans Canada. Pacific Region. 1999a. Pacific region 1999/2000 management plan, red sea urchin. Vancouver, B.C. 30p; Appendices. [DFO] Department of Fisheries and Oceans Canada. Pacific Region. 1999b. Conditions for transporting geoducks, horse clams, sea cucumbers, red sea urchins, and green sea urchins. Vancouver, B.C., 3p. Druehl, L. D. 1978. The distribution of Macrocystis integrifolia in British Columbia as related to environmental parameters. Canadian Journal of Botany 56:69-79. Duggins, D. O. 1980. Kelp beds and sea otters: an experimental approach. Ecology 61:447453. Duggins, D. O. 1981. Interspecific facilitation in a guild of benthic marine herbivores. Oecologia 48:157-163. Duggins, D. O., C. A. Simenstad, and J. A. Estes. 1989. Magnification of secondary production by kelp detritus in coastal marine ecosystems. Science 245:170-173.  31  Duggins, D. O., J. E. Eckman, and A. T. Sewell. 1990. Ecology of understory kelp environments. II. Effects of kelps on recruitment of benthic invertebrates. Journal of Experimental Marine Biology and Ecology 143:27-45. Ebeling, A. W., and D. R. Laur. 1988. Fish populations in kelp forest without sea otters: effects of severe storm damage and destructive sea urchin grazing. Pages 169-189 in G.R. VanBlaricom and J.A. Estes, editors. The community ecology of sea otters. Springer-Verlag, Berlin. Ebert, T. A. 1968. Growth rates of the sea urchin Strongylocentrotus purpuratus related to food availability and spine abrasion. Ecology 49:1075-1091. Ebert, T. A. 1977. An experimental analysis of sea urchin dynamics and community interactions on a rock jetty. Journal of Experimental Marine Biology and Ecology 27:1-22. Ebert, T. A., and M. P. Russell. 1992. Growth and mortality estimates for red sea urchin Strongylocentrotusfranciscanusfrom San Nicolas Island, California. Marine Ecology Progress Series 81:31-41. Eckman, J. E., D. O. Duggins, and A. T. Sewell. 1989. Ecology of understory kelp environments. I. Effects of kelps on flow and particle transport near the bottom. Journal of Experimental Marine Biology and Ecology 129:173-187. Edwards, P. B., and T. A. Ebert. 1991. Plastic responses to limited food availability and spine damage in the sea urchin Strongylocentrotus purpuratus (Stimpson). Journal of Experimental Marine Biology and Ecology 145:205-220. Ellis, D. W., and L. F. Swan. 1981. Teachings of the tides: uses of marine invertebrates by the Manhousat People. Theytus Books, Nanaimo, B.C. Estes, J. A., and D. O. Duggins. 1995. Sea otters and kelp forest in Alaska: generality and variation in a community ecological paradigm. Ecological Monographs 65:75-100. Estes, J. A., and J. F. Palmisano. 1974. Sea otters: their role in structuring nearshore communities. Science 185:1058-1060. Fletcher, W. J. 1987. Interactions among subtidal Australian sea urchins, gastropods, and algae: effects of experimental removals. Ecological Monographs 57:89-109. Foster, M. S., and D. R. Schiel. 1985. The ecology of giant kelp forests in California: a community profile. U.S. Fish and Wildlife Service. Biological report 85(7.2). Graziano, A. M., and M. L. Raulin. 1993. Research methods: a process of inquiry. 2 edition. Harper Collins College Publishers, New York.  nd  32  Greenfield, L., A. C. Giese, and A. Farmanfarmaian. 1958. Cyclic biochemical changes in several echinoderms. Journal of Experimental Zoology 139:507-524. Harrold, C , and J. S. Pearse. 1987. The ecological role of echinoderms in kelp forests. Pages 137-233 in M. Jangoux and J. M. Lawrence, editors. Echinoderm studies 2. A.A. Balkema, Rotterdam. Hesquiat First Nation. 1992. Management for a Living Hesquiat Harbour research program proposal. Unpublished research proposal. Summer, 1992. Higashi, H., R. Kikuchi, and K. Tabei. 1959. Nutritive elements in salted marine appetizers. Nutrition and Foods 11:62-65. Higashi, H., T. Yamakawa, M. Yanase, T. Kinumaki, Y. Shimizu, K. Sugii, and H. Iida. 1965. Vitamins in salted sea urchin gonad: Nutrition and Foods 18:66-69. Hilborn, R., and S. C. Stearns. 1982. On inference in ecology and evolutionary biology: the problem of multiple causes. Acta Biotheoretica 31:145-164. Himmelman, J. H., A. Cardinal, and E. Bourget. 1983. Community development following removal of urchins, Strongylocentrotus droebachiensis, from the rocky subtidal zone of the St. Lawrence Estuary, Eastern Canada. Oecologia 59:27-39. Hughes, T. P., D. C. Reed, and M. Boyle. 1987. Herbivory on coral reefs: community structure following mass mortalities of sea urchins. Journal of Experimental Marine Biology and Ecology 113:39-59. Jackson, G. A., and C. D. Winant. 1983. Effect of a kelp forest on coastal currents. Continental Shelf Research 2:75-80. Johnson, M. W. 1930. Notes on the larval development of Strongylocentrotus franciscanus. Publications Puget Sound Biological Station 7:401-407. Jones, N. S., and J. M. Kain. 1967. Subtidal algal colonization following the removal of Echinus. Helgolander Wissenschaftliche Meeresuntersuchungen 15:460-466. Kain, J.M., and N.S. Jones. 1966. Algal colonization after removal of Echinus. Pages 139140 in E.G. Young and J.L. McLachlan, editors. Proceedings of the fifth international seaweed symposium. Pergamon Press, Oxford. Kato, S., and S. C. Schroeter. 1985. Biology of the red sea urchin, Stongylocentrotus franciscanus, and its fishery in California. Marine Fisheries Review 47:1-20.  Keats, D. W., G. R. South, and D. H. Steele. 1990. Effects of an experimental reduction in grazing by green sea urchins on a benthic macroalgal community in eastern Newfoundland. Marine Ecology Progress Series 68:181-193. Kitching, J. A., and F. J. Ebling. 1961. The ecology of Lough Ine: XI. The controlof algae by Paracentrotus lividus (echinoidea). Journal of Animal Ecology 30:373-383. Koehl, M. A. R., and R. S. Alberte. 1988. Flow, flapping, and photosynthesis of Nereocystis luetkeana: a functional comparison of undulate and flat blade morphologies. Marine Biology 99:435-444. Kramer, D. E., and D. M. A. Nordin. 1979. Studies on the handling and processing of sea urchin roe. I. Fresh product. Canadian Fisheries and Marine Services, Technical Report 870, 47p. Kuhn, T. S. 1970. The structure of scientific revolutions, 2nd enlarged edition. University of Chicago Press, Chicago. Laur, D. R., A. W. Ebeling, and D. A. Coon. 1988. Effects of sea otter foraging on subtidal reef communities off central California. Pages 151-168 in G.R. VanBlaricom and J.A. Estes, editors. The community ecology of sea otters. Springer-Verlag, Berlin. Lawrence, J. M. 1975. On the relationships between marine plants and sea urchins. Oceanography and Marine Biology 13:213-286. Leighton, D. L. 1971. Grazing activities of benthic invertebrates in southern California. In, The biology of giant kelp beds (Macrocystis) in California, edited by W.J. North, Beihfte zur Nova Hedwigia, Vol. 32, pp. 421-453. Leighton, D. L., L. G. Jones, and W. J. North. 1966. Ecological relationships between giant kelp and sea urchins in southern California. Pages 141-153 in E.G. Young and J.L. McLachlan, editors. Proceedings of the fifth international seaweed symposium. Pergamon Press, Oxford. Leinaas, H. P., and H. Christie. 1996. Effects of removing sea urchins (Strongylocentrotus droebachiensis): stability of the barren state and succession of kelp forest recovery in the east Atlantic. Oecologia 105:524-536. Lemire, M., and J. H. Himmelman. 1996. Relation of food preference to fitness for the green sea urchin, Strongylocentrotus droebachiensis. Marine Biology 127:73-78. Levitan, D. R., M. A. Sewell, and F. S. Chia. 1992. How distribution and abundance influence fertilization success in the sea urchin Strongylocentrotus franciscanus. Ecology 73:248-254. Lovelock, J. E. 1995. Gaia: a new look at life on earth. Oxford University Press, Oxford.  34  Low, C. J. 1975. The effect of grouping of Strongylocentrotusfranciscanus,the giant red sea urchin, on its population biology. Ph.D. dissertation. University of British Columbia, Vancouver. Macey, D. 2001. D&D Pacific Fisheries Ltd. Gibsons, British Columbia. Personal communication regarding urchin demand cycles and current landed values. Mattison, J. E., J. D. Trent, A. L. Shanks, T. B. Akin, and J. S. Pearse. 1977. Movement and feeding activity of red sea urchins (Strongylocentrotusfranciscanus)adjacent to a kelp forest. Marine Biology 39:25-30. McCauley, J. E., and A. G. Carey. 1967. Echinoidea of Oregon. Journal of the Fisheries Research Board of Canada 24:1385-1401. McClanahan, T. R., and J. D. Kurtis. 1991. Population regulation of the rock-boring sea urchin Echinometra mathaei (de Blainville). Journal of Experimental Marine Biology and Ecology 147:121-146. Mottet, M. G. 1976. The fishery biology of sea urchins in the family Strongylocentrotidae. Technical report no. 20. State of Washington Department of Fisheries Shellfish Division. Muse, B. 1998. Management of the British Columbia sea urchin fisheries. Alaska Commercial Fisheries Entry Commission, Juneau, Alaska. CFEC 98-2N. 27p. North, W. J. 1964. Ecology of the rocky nearshore environment in Southern California and possible influences of discharged wastes. In, Advances in water pollution research. E.A. Pearson, ed. Pergamon Press, New York. Paine, R. T. 1969. A note on trophic complexity and community stability. The American Naturalist 103:91-93. Paine, R. T., and R. L. Vadas. 1969. The effects of grazing by sea urchins, Strongylocentrotus spp., on benthic algal populations. Limnology and Oceanography 14(5):710-719. Pearse, J. S., and A. H. Hines. 1979. Expansion of a central California kelp forest following the mass mortality of sea urchins. Marine Biology 51:83-91. Pearse, J. S., M. E. Clark, D. L. Leighton, C. T. Mitchell, and W. J. North. 1970. Kelp habitat improvement project, 1969-70, California Institute of Technology, Appendix, 87 pp. Piatt, J.R. 1964. Strong inference. Science 146:347-353.  35  Popper, K. R. 1963. Conjectures and refutations: the growth of scientific knowledge. Harper and Row, New York. Quast, J. C. 1968. Fish fauna of the rocky inshore zone. Pages 35-55 in W. J. North and C. L. Hubbs, editors. Utilization of kelp-bed resources in Southern California. California Department of Fish and Game, Sacramento. Quinn, J. F., and A. E . Dunham. 1983. On hypothesis testing in ecology and evolution. The American Naturalist 122:602-617. Randall, J. E . 1961. Overgrazing of algae by herbivorous marine fishes. Ecology 42:812. Reed, D. C , and M . S. Foster. 1984. The effects of canopy shading on algal recruitment and growth in a giant kelp forest. Ecology 65:937-948. Rowley, R. J. 1989. Settlement and recruitment of sea urchins (Strongylocentrotus spp.) in a sea-urchin barren ground and a kelp bed: are populations regulated by settlement or post-settlement processes? Marine Biology 100:485-494. Schroeter, S. C. 1978. Experimental studies of competition as a factor affecting the distribution and abundance of purple sea urchins, Strongylocentrotus purpuratus (Stimpson). Ph.D. dissertation. University of California, Santa Barbara. Simenstad, C. A., J. A . Estes, and K. W. Kenyon. 1978. Aleuts, sea otters, and alternate stable-state communities. Science 200:403-411. Sloan, N . A . 1991. Underwater world: red sea urchin. Department of Fisheries and Oceans Canada, Ottawa. 4pp. Sloan, N . A . , C. P. Lauridsen, and R. M . Harbo. 1987. Recruitment characteristics of the commercially harvested red sea urchin Strongylocentrotusfranciscanusin southern British Columbia, Canada. Fisheries Research 5:55-69. Swan, E. F. 1961. Some observations on the growth rate of sea urchins in the genus Strongylocentrotus. Fisheries Bulletin 120:420-427. Tegner, M . J., and P. K. Dayton. 1977. Sea urchin recruitment patterns and implications of commercial fishing. Science 196:324-326. Tegner, M.J. 1981. Population structure, recruitment and mortality of two sea urchins (Stongylocentrotus franciscanus and S. purpuratus) in a kelp forest. Marine Ecology Progress Series 5:255-268. Tegner, M . J . 1991. Sea urchins, El Ninos, and the long term stability of Southern California kelp forest communities. Marine Ecology Progress Series 77:49-63.  36  \  Turner, N.J. 1997. Traditional ecological knowledge. Pages 275-298 in P. K. Schoonmaker, B. von Hagen and E. W. Wolf, editors. The rain forests of home. Island Press, Washington, D.C. Underwood, A. J., editor. 1997. Experiments in ecology: their logical design and interpretation using analysis of variance. Cambridge University Press, Cambridge. Vadas, R. L. 1977. Preferential feeding: an optimization strategy in sea urchins. Ecological Monographs 47:337-371. Vance, R. R. 1979. Effects of grazing by the sea urchin, Centrostephanus coronatus, on prey community composition. Ecology 60:537-546. Watson, J. C. 1993. The effects of sea otter (Enhydra lutris) foraging on shallow rocky communities off northwestern Vancouver Island, British Columbia. Ph.D. dissertation. University of California Santa Cruz, Santa Cruz. Wenner, A. M. 1989. Concept-centered versus organism-centered biology. American Zoologist 29:1177-1197. Woodcock, G. 1990. British Columbia: a history of the province. Douglas & Mclntyre, Vancouver.  2 . Sea Urchins and Kelp: Effects of Removing Urchins ABSTRACT The shallow subtidal regions near Hot Springs Cove, Vancouver Island, British Columbia (49° 21' N, 126° 16' W) are characterised by large rocky areas populated by the red sea urchin, Strongylocentrotusfranciscanus.The resident Hesquiat were interested in learning the effects of removals of sea urchins by harvesting or by sea otters on the coastal resources of the traditional ecological territory of the Hesquiat First Nation. S. franciscanus were removed at three sites with four levels of urchin removal per site. Manipulations of urchin density were maintained throughout the experiment and monitored seasonally for two seasons pre-treatment and seven seasons post-treatment. The manipulations resulted in the conversion of urchin-dominated, subtidal regions into kelp-dominated algal communities. The removal of sea urchins at all levels resulted in a rapid increase in the presence of laminarialean algal species (kelps). Study plots where all urchins were removed developed a dense understory (dominated by Laminaria bongardiana and Pterygophora californica) and canopy (Nereocystis luetkeana) of kelps with little bare rock remaining. Intermediate levels of removal resulted in a mosaic of smaller urchin-dominated and kelp-dominated patches. Control plots maintained the urchin-dominated barren state throughout the study period. The density of S.franciscanusappears to be responsible for the absence of kelp forest habitat in the region. The effects of urchin removal on these communities, and the implications for both ecological theory and sea urchin management are discussed. Results indicate that removal of urchins below approximately one urchin per m stimulates conversion of urchin barrens to kelp forests. The balancing of these community  compositions has implications for the management of the sea urchin fishery in British Columbia.  INTRODUCTION Herbivore foraging can dramatically influence community structure, either increasing and decreasing plant diversity (Vadas 1968, Harper 1969, Paine and Vadas 1969, Lubchenco 1978). The positive or negative association between primary producers and their consumers can have dramatic effects on other members of the community. The term "trophic cascade" describes scenarios where productivity is regulated hierachically through both abiotic and biotic mechanisms. Abiotic factors such as nutrient supply establish the potential productivity, however actual productivity is determined by food web structure (Carpenter et al. 1987). Strong (1992) has further refined the concept by indicating that trophic cascades generally occur when the standing crop of autotrophs is greatly reduced by one or a few species of "potent herbivores" in the absence of predators. Consequently, studies of herbivore-plant interactions need to be sensitive to determining the critical points where a diverse community can begin a cascade to a virtually barren landscape. Sea urchins (Echinoidea) have been shown to control algal communities through intensive grazing in many parts of the world (Lawrence 1975, Duggins 1980, Harrold and Pearse 1987, Hughes et al. 1987, Keats et al. 1990, Tegner and Dayton 1991, Watanabe and Harrold 1991, Valentine and Heck 1991, Andrew and Underwood 1993, Leinaas and Christie 1996). In temperate waters, the results of intensive grazing by sea urchins have been termed "barren-grounds" (Pearse et al. 1970, "Isoyake" (Noro et al. 1983), and "coralline flats" (Ayling 1981). Each of these terms attempts to describe effects of intensive sea urchin grazing on algal communities. The general trend is that such grazing removes the fleshy  39  macroalgae, leaving tougher crustose and upright coralline algae forms. These forms have calcium carbonate in their cell walls and are less preferred food items as compared to many alga species (Vadas 1977). Intensive sea urchin grazing has been attributed to storms (Harrold and Reed 1985), poor kelp recruitment (Dean et al. 1984) urchin density (Low 1975, Mattison et al. 1977, Dean et al. 1984), and predator removal (Mann 1977, Garnick 1989, Estes and Duggins 1995). Regardless of the proximal cause, the critical component inducing grazing seems to be the availability of drift algae. For instance, when storms remove kelp plants, the resulting loss of drift algae produced by mature plants induces urchins to begin actively foraging (Harrold and Reed 1985). This active foraging prevents the establishment of new plants as urchins scrape the bottom in search of food. Perhaps the most compelling observations of feeding by urchins are those of Leighton (1960) who described a grazing front consisting of a 30-m wide band of S. purpuratus and S. franciscanus moving through a kelp bed in southern California. The urchins did not directly consume all of the algae, but much was lost as grazing on holdfasts and stipes caused the plants to weaken and wash away. Decreased predation is often cited as causing increases in density and biomass of sea urchins (Estes and Palmisano 1974, Mann 1977, Simenstad etal. 1978, Garnick 1989, Levitan 1992, Estes and Duggins 1995, McClanahan and Sala 1997). Increased urchin abundance then leads to the formation of barren areas as the limited supply of drift kelp induces more active grazing by sea urchins (Lees 1970, Harrold and Reed 1985) Lobster (Homarus americanus) harvesting (Mann 1977, Hagen and Mann 1992), on the east coast, sea otter (Enhydra lutris) harvesting (Estes and Palmisano 1974, Estes and Duggins 1995) on the west coast of North America, and fishing intensity in the  Mediterranean (McClanahan and Sala 1997), are proposed causes of overgrazing by sea urchins. In all cases, the decrease in predators allows sea urchin populations to increase. In many cases, the abundance of drift algae becomes limiting and urchins begin to actively forage, deforesting kelp forests. The barrens state is maintained by urchins subsisting on drift kelp from sources beyond the urchin barren (Lees 1970, Mattison et al. 1977, Duggins 1980), generalist feeding habits (Briscoe and Sebens 1988, Duggins 1981), through absorption of dissolved organic matter (Pearse et al. 1970), and an ability to survive on low nutritional inputs by modifying resource allocation to body components (Ebert 1967, 1968, Lawrence 1970, Lang and Mann 1976, Johnson and Mann 1982). Paine (1966) found that by preying on a space-monopolising organism (mussels), sea stars (Pisaster ochraceus) controlled local species diversity (Paine 1966, 1969, Paine et al. 1985). In 1969, Paine coined the term "keystone predator" to describe such a relationship. A similar keystone role has been ascribed to sea otters in North America. In the past, sea otters played a critical role in structuring kelp-forest ecosystems on the west coast of North America (Estes and Palmisano 1974, Estes and Duggins 1995). By feeding on sea urchins, sea otters prevent over grazing by sea urchins. Areas inhabited by sea otters have been shown to have higher biomass and diversity of algae than otter-free areas or the same areas before otter presence (Estes and Palmisano 1974, Watson 1993, Estes and Duggins 1995). The lost primary productivity, in the absence of sea otters, has implications for organisms that reside on kelps, feed on kelps, live within kelp forests, or depend on kelp by-products (Edwards 1980, Wing and Clendenning 1971, Lowry et al. 1974, Hobson and Chess 1976, Ebeling and Laur 1984, Duggins et al. 1989). The elimination of sea otters, in the late 1800s,  41  from many Vancouver Island habitats, resulted in the loss of historic community structure from those habitats (Watson 1993). Most studies investigating the effects of sea urchins on community structure have focused on observational evidence. In the temperate eastern Pacific, much research has focused on sea otters as the agent of urchin removal. Studies have either relied on observations of sea otter and sea otter-free areas (Estes and Palmisano ,1974) or natural experiments following the reintroduction of sea otters (Watson 1993). Others have relied on natural fluctuations in urchin density or die-off events (Foreman 1977, Pearse and Hines 1979, Hughes et al. 1987). Those researchers who conducted manipulative experiments tended to rely on total removals (Paine and Vadas 1969, Duggins 1980, Fletcher 1987, Keats et al. 1990, Leinaas and Christie 1996), or on the use of cages to prevent access by sea urchins to a small area (Paine and Vadas 1969, Dean et al. 1988). Few researchers have attempted to study the effects of urchins through manipulations beyond all or nothing removals at biologically meaningful scales. McClanahan and Kurtis (1991) manipulated levels of Echinometra mathaei through removal as well as addition, of two and three times natural levels, to determine the potential impacts of E. mathaei. Andrew and Underwood (1993) performed the experiment closest to the one conducted here. They removed 0, 33%, 66%o, and 100% of Centrostephanus rodgersiifromnaturally occurring barrens habitat off New South Wales, Australia. Because the current study addressed potential influences of fishing pressure on sea urchin populations, the all or nothing removals, sea otter literature, and natural experiments studying the effects of sea urchins provided incomplete information. To date, there have been no adequately controlled experimental manipulations to study the effects of variable urchin density in the northeastern Pacific Ocean (Table 1-1).  The goals of this study were to determine the spatial and temporal relationships between the sea urchin Strongylocentrotus franciscanus and members of the Laminariales (Phaeophyta), the order of brown algae commonly known as kelp. By experimentally removing sea urchins, it was possible to test the hypothesis that sea urchins are responsible for limiting the recruitment of kelps thereby mediating a top down trophic cascade. By manipulating urchins at variable densities, it was possible to determine the functional relationship between urchin and kelp abundance at the scale of the experiment. Through the manipulations performed in this experiment I can reject, modify or support some of the ecological theories closely associated with previous work on mechanisms of community structure. Moreover, managers can use the results of this work to assess appropriate harvesting levels and strategies for S. franciscanus.  METHODS AND MATERIALS Study area The study site is located on the west coast of Vancouver Island, British Columbia (49° 21' N, 126° 16' W) within the traditional territory of the Hesquiat First Nation (see Appendix; Figure 1-2). The local waters are bordered by a convoluted rocky shoreline with numerous rocky reefs and small islands. As shoreline, islands, and reefs give way to deeper water, one encounters large boulder fields and rocks which ultimately lead to soft gravel and sand bottom. Sea state varies dramatically among (and to a lesser degree within) the study sites because shore aspect, islands, and reefs act to subdue or enhance the effects of waves. In water depths ranging from 3 to 20 meters (perhaps even deeper) S. franciscanus  dominates  the local rocky-bottom habitats ranging from 0 to 27 urchins / m . 2  43  Experimental design To test the effect of varying urchin densities on kelp abundance, a repeated-measures, randomised block design was used. Three sites were chosen as experimental replicates. Each site contained four treatments of different levels of urchin density. The treatments were each confined to circular plots with radii of 15-m. Each site was sampled during two pretreatment seasons (Summer and Fall 1997) and seven post-treatment seasons (Winter 1998 through Summer 1999) for a total of nine repeated measures. I selected the study sites opportunistically within the study area with the following criteria in mind. Sites needed to contain enough urchin-dominated rocky-bottom habitat for the placement of the four circular 15-m radius plots. If large portions of a site were broken up by soft-bottom habitats, then that excluded the site from the study. Within each site, treatment plots were placed such that they did not overlap or contain significant areas of non-rock habitat or shallow reef. Treatments were randomly assigned to study plots within each site.  Sea urchin surveys To assess the pre-treatment levels of sea urchin abundance and monitor the success of removals, I conducted sea urchin surveys in all study sites and plots. Sea urchin surveys were conducted with the use of SCUBA. Surveys began in August 1997 and continued seasonally until the project completion in September 1999. All surveys for the study, utilised transect lines radiatingfromthe centre of study plots. Plot centres were marked by rope and a buoy, fastened to the bottom via a stainless steel anchor bolt (Hilte 1/2x6 inches). Transect lines were lead-core rope laid along the bottom for 12 m. The first transect was placed from the plot centre along a randomly selected bearing with two more placed systematically at 120-degree intervals. Transects were marked to divide the sample area into three equal 4-m segments (see Appendix; Figure 44  1-3), and were used to delineate the centre of a belt transect. All red sea urchins were counted within the belt transects and their placement noted in regards to side of the line, and position within clusters marked as 0 to 4 m, 4 to 8 m, and 8 to 12 m. Urchin test diameters were measured and the substrate of attachment was recorded as sand, shell, rock, boulder or reef (see Appendix; Table 1-1 for definitions) for all urchins in the belt transects or until the constraints of SCUBA prevented further sampling. I purposely varied the width of belt transects within each experimental unit and time periodfrombetween 1 m and 2 m based on previous notes on urchin densities within the plot. The variation allowed for equal sampling per unit of effort, and often prevented a need for multiple dives to estimate the density of urchins in plots of higher urchin density. Some juvenile urchins may have been missed because the sampling technique did not involve moving large urchins to expose juveniles that may have been sheltering under adults. Treatments - urchin density manipulation To describe the kelp response to urchin density, I began experimental manipulations of S. franciscanus densities in November / December 1997. The four treatment levels were total removal of urchins (total), removal of urchins to 0.4 urchins / m (medium), removal of 2  urchins to 1.0 urchins / m (low), and a control where no urchins were removed (control). 2  These treatment levels were chosen because natural densities of approximately two urchins per m were observed to maintain the barrens condition. Thus, a variety of densities below 2  this level were chosen to best elucidate relationship of urchin density with the study measurables. Manipulations were achieved by removing urchins within a 15-m radius from the study plot centres. The 15-m radius included an extra 3 m beyond the sampled area to allow a buffer to account for some movement of urchins towards the plot between sampling  periods. Removed urchins were released in suitable habitats. When oceanic conditions were limiting, some urchins were crushed instead of removed. Whenever possible crushing was restricted to total removal plots because crushed urchins have been shown to cause a flight response in remaining urchins (Watson 1993). To achieve the appropriate density, data collected just prior to urchin removal were used. Urchin densities were calculated from the surveys discussed above. From the density data, the ratio of the desired treatment density to the sampled density was calculated. For instance, if a plot contained an average of five urchins / m and the assigned treatment was 2  one urchin / m ,1 would need to leave behind 1/5 of the existing urchins to achieve the 2  treatment level. Because the Hesquiat were concerned with the potential impacts of a fishery, only the largest urchins from the treatment plots were removed. This strategy is similar to the current urchin management practice of setting a lower size limit for commercial fishing (DFO 1999). The minimum urchin size was selected such that removal above this size would result in the desired density. To select the removal size, the sampled urchin test diameters were sorted from low to high. The size for removal was the point in the sorted sample where the cumulative frequency of the observations equalled to the ratio of desired treatment density to the sampled density. This point established the size of urchins such that removal of urchins above that size would theoretically result in the desired treatment level (see Appendix; Figure 1-4 for a sample calculation and Figure 1-5 for urchin size histograms for each experimental unit by time combination). Removals were continued as necessary following each round of seasonal post-treatment data collection (see Appendix; Table 1-2 for treatment history and Figure 1-6 through Figure 1-8 for study plot layout and physical characteristics within each site).  ^  46  Kelp surveys Kelp surveys were conducted to assess the implications of variable removal levels of sea urchins on kelp resources. Kelp surveys were conducted concomitantly with sea urchin surveys, spanning the same time periods as the sea urchin surveys. The same transect line used to sample urchins was used to place quadrats for kelp sampling. 0.1 -m quadrats (33-cm x 33-cm) were placed in an alternating pattern to each 2  side of the line at each meter mark along the 12-m transect line. Placement of the first as either left or right side of the line was chosen at random (see Appendix; Figure 1-3). Occasionally, when kelp density was low and time permitted, all possible quadrats (left and right at each point) were sampled. Data collected for kelp included species, estimated length, substrate of attachment, presence of blades, number of stipes, and occasionally, holdfast diameter. In August 1999, during the final field season, I sampled as usual but also removed all sampled kelp by cutting just above the holdfast. Removed kelp were weighed and measured for stipe and overall length. The regression equations for calculating biomass from length for each of the encountered species are included in the Appendix (Appendix; Figure 1-9 through Figure 1-13).  Data analysis Sea urchin and kelp density data were analysed as a repeated measures analysis of variance and presented as a split-plot ANOVA. The between-subjects experimental design is a randomised complete block design. The three study sites were treated as blocks and the four levels of urchin removal were the treatments. Treatments were randomly assigned to study plots within each block creating 12 experimental units. The within-subjects design consists of the nine repeated measurements (sampling seasons) of urchin and kelp density 47  within each experimental unit. Within each experimental unit, multiple observations (kelp quadrats and urchin transects) were collected. Quadrats were created from the sampled belt transects for urchin density. A quadrat consisted of transect line segments (0 to 4 m, 4 to 8 m or 8 to 12 m) combined with the side of the line (left or right). The three transect lines, each with three line segments, and two sides provided 18 quadrats. For kelp, the 0.1-m sampling quadrats were combined within each of the urchin sampling quadrats to create 18 coinciding observations per experimental unit. These "pseudoreplicates" (sensu Hurlbert 1984) are only used where appropriate to partition variance into sampling error. Experimental replication for the purposes of the data analyses are three (n = 3). The mean density for all plots replaced missing data for the Site 2 treatments "medium and "low" in fall 1998. Because repeated measures (time) could not be randomly assigned within the experimental units, I have chosen to present all time series analyses as split-plot ANOVAs. Study sites (blocks) are equivalent to the whole plot, and the experimental units (treatment plots) are equivalent to the whole-plot treatment factor in the split-plot design. The repeated measures on experimental units are analogous to the subplot in the split-plot design (Kuehl 1994). Treatments were considered a fixed effect with all other factors analysed as random effects. I conducted linear regression analyses to determine the degree of the relationship between urchin and kelp density after manipulating urchin density. Observed urchin and kelp densities for each sample plot were paired for post-treatment time periods except for the sampling period immediately following thefirstremoval of sea urchins (n = 24 / site). Data were Login(x+l) transformed before analysis and plotted. This transformation made the data linear and more robust in meeting the assumptions of normality. The data were analysed and  48  graphed using Statistica software for Windows . An alpha probability of 0.05 was chosen for all statistical tests. Simpson's diversity index (Simpson 1949) was used to assess the species diversity at the termination of the experiment. The kelp samples from each site were standardised to prevent misrepresentation of the data based on increased probability of encountering rare species as sample size increases (Ricklefs and Miller 2000). Standardisation was established by having Microsoft Excel™ sub-sample the kelp data to obtain equal sample sizes of species composition among treatments within each site. The plot with the least number of kelp observations was used to determine the number of sub-samples taken for each treatment within a site. The index was then manually calculated as  where p is the proportion of the observed number of each given species to the total observations across all species in the sample  RESULTS Urchin Densities Removal of S.franciscanusfromthe study plots significantly affected the spatial and temporal abundance of urchins. However, maintaining the medium and low levels of urchin removal as distinct treatments was more difficult than anticipated. During the study, it soon became evident that the low and medium levels of urchin removal contained urchin densities too close together to obtain biologically or statistically insightful data. In many cases, the density of urchins in the medium-level removal plots was higher at the sampling period following removal than the low-level removal plots and vice versa. This happened for two 49  reasons. Sampling variability was too high to statistically separate the two intermediate levels of removal, and it was difficult to maintain specific numbers of urchins for three months without some means of restraining movement (e.g. cages). Thus, the low and medium levels of urchin removal were combined into a single treatment level, termed 'intermediate' Figure 2-1 illustrates the effectiveness of urchin removal in obtaining the desired densities. The figure displays the treatment and site effects of the study. The control plots (no urchin removal) display the general trend in urchin density among the three areas. Site 1 had fewer urchins than either of Sites 2 and 3. Total and intermediate levels of urchin removal show a clear drop in density after the removal of urchins for all sites. Two-way repeated measures ANOVA for urchin densities over the duration of the study show all effects and interactions are statistically significant (Table 2-1). The variability in urchin numbers through time is likely the result of urchins invading the study plots in between sampling periods. It is important to note that the differing amounts of urchins removed to obtain the given densities affected the total feeding in a plot per unit time, creating a temporal refuge for kelp. This refuge is likely beneficial to kelp recruitment regardless of the observed urchin density noted at subsequent sampling periods. The effect of varying urchin removal was apparent in the amount of kelp found in the study plots, and is addressed in the kelp density results. Treatment effects are responsible for both the treatment by time interaction and the significant time factor (Figure 2-2). Both components of the ANOVA model are significant because of the drop in urchin densities after treatments were implemented. The interaction term is significant because of the extreme deviation in pattern between the removal and  50  control sites. I attribute the significant time factor to the differences in pre-treatment and post-treatment urchin densities. The averaged trends in urchin density are relatively stable before and after the treatments were implemented. However, the post-treatment densities for urchins are much lower than pre-treatment levels. The observed densities of urchins among the treatments illustrate the difficulties in maintaining separate classes of urchin removal (Figure 2-2). While it may not be possible to separate the intermediate and total urchin removals from each other in terms of urchin density, the resulting kelp growth is quite different (next section). The differences in the response of kelp suggest that the small differences in urchin density between total and intermediate removals have dramatic implications for kelp recruitment. Treatment by block interaction is apparent in Figure 2-3. The interaction results from the varying amounts of urchins in the control plots among the three sites. While the intermediate and total removal sites were forced into low and relatively stable urchin densities the controls were not. The fact that the controls then varied greatly among the three sites creates a divergence in pattern among the three treatments. This interaction further supports the biological significance of the treatments because while controls varied among sites the treatments did not. Figure 2-3 again illustrates the difficulty in separating the total removal plots from the intermediate removals with regards to urchin density. The lack of a clear distinction between removal levels is due to urchins re-invading the study plots. It should be recognised that the abundance of urchins sampled does not reflect the amount of feeding that occurs in a plot. The re-invasion and subsequent sampling of urchins in the total removal plots makes it difficult to demonstrate differences in urchin numbers among the two treatments. However, the kelp response to urchin removal is not only a function of observed  51  urchin densities but also of the impact of the actual treatments. The treatments also create temporal refuges for kelpfromthe urchins (sensu Duggins 1983). A more appropriate measure of urchin presence would be a measure of urchin-days spent feeding within a plot. The sampling methodology and resources available prevented such a measure.  Urchin Density Over Time Mean urchin densities and standard errors 11 10 9  E  Total  Intermediate  Control  Figure 2-1. S. franciscanus density as a function of time for each treatment. Series represent each of the three study sites. Bars are ± one standard error of the mean (n = 18 for Total and Control; n= 36 for Intermediate). Vertical lines after Fall 1997 represents the implementation of treatments. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3).  Table 2-1. Split-plot analysis of variance for logio (data +1) of urchin densities. Treatment effects tested using the pseudo f-test. Error d.f. for treatment effects is 5.43 (refer to Hicks 1982). Effect  D.F.  MS  Error M S  F  p  Block (Site)  2  6.34  0.40  15.81  0.000  Treatment  2  22.97  2.75  8.36  0.025  Error 1 (BlockTreatment)  4  2.32  0.40  5.79  0.001  Time  8  4.21  0.40  10.49  0.000  Time*Treatment  16  0.83  0.40  2.07  0.027  Error 2  48  0.40  —  —  ...  1656  0.04  —  ...  ...  Sampling Error  54  Time by Treatment Interaction M e a n urchin density and standard error  Total Intermediat S u m 97 Fall 97 W i n 98 S p r 98 S u m 98 Fall 98 W i n 99 S p r 99 S u m 99  Control  TIME  Figure 2-2. S. franciscanus density as a function of time averaged across sites (N = 3). Bars are ± one standard error of the mean (n = 54 for Total and Control; n = 108 for Intermediate). Vertical line represents the time of treatment implementation. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3).  Site by Treatment  Interaction  Urchin densities a n d standard errors  8  Figure 2-3. S. franciscanus density as a function of site with treatments plotted as series. Bars are ± one standard error of the mean (n = 18 for Total and Control; n = 36 for Intermediate). Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively.  Kelp Densities Removal of S.franciscanushad an obvious effect on the post-treatment recruitment of kelp. The kelp species encountered during sampling are listed and categorised as canopy or understory species in Table 2-2. Table 2-3 summarises the results of the two-way repeated measures ANOVA for kelp densities where all individuals are counted regardless of species over the duration of the study. In the model, all effects and interactions are statistically significant. Plots of the interactions and main effects show that it is the removal of sea urchins that drives the response in kelp numbers present (e.g. Figure 2-5). Figure 2-4 illustrates replicates of the experiment for each individual treatment through time. As with urchin densities, the kelp response varied with each replicate of the experiment. Site 1 had the least amount of kelp present, while Sites 2 and 3 tended to have much higher levels of kelp recruitment after removal of urchins. The pattern in control plots is striking. Plots experiencing no removal of urchins showed almost total exclusion of kelp during the study (Figure 2-4). The significant time by treatment interaction and the significant time factor from the analysis are clearly displayed in the density response of kelp averaged by treatment over time (Figure 2-5). The treatment by time interaction is primarily the result of the treatments. Treatments began in November and December of 1997. The subsequent two sampling periods (Win 1998 & Spr 1998) show an increase in kelp abundance while the controls remained stable with no change in kelp or urchin densities. The inclusion of before and after data create a time by treatment interaction because the control plots and two treatment levels differ in their response through time. Given that this interaction is adequately explained though the treatment response, it is appropriate to look at the response of the treatment plots independent of time. There was a clear trend between the level of urchin removal and kelp 57  density. Total removal plots showed a strong response in the magnitude of kelp density following removal. Sites 2 and 3 were the most strongly affected and had the greatest kelp densities. Site 1 contained less kelp, but the functional response was similar to Sites 2 and 3 while very different from the controls. Effects of intermediate levels of urchin removal were harder to discern and from Figure 2-5 appear similar to controls. Given that the average values for control plots hovers very close to zero while the intermediate plots fluctuate around 10 stipes per square meter, there is biological significance. I noted that urchins reinvaded total removal plots and the resulting urchin densities were often difficult to discern from the intermediate levels of urchin removal (Figure 2-2 and Figure 2-3). However, the kelp data show a clear difference between these treatment levels in terms of the kelp response (Figure 2-5). The kelp response suggests the impact of re-invading urchins was minimal and likely negligible. I also divided the kelp response into understory and canopy groups because of their structural and ecological differences (Estes and Duggins 1995), and because their responses vary (Figure 2-6 and Figure 2-7). The figures illustrate that understory species account for the greatest numbers of stipes per square meter. Laminaria bongardiana (formerly known as L. groenlandica (Gabrielson et al. 2000)) and Pterygophora californica make up the majority of this response (Table 2-2). P. californica tends to be the dominant understory species at Site 1 while L. bongardiana tends to dominate at Sites 2 and 3 (see Table 2-4). Canopy kelps show peaks in abundance during spring 1998 and spring 1999 (Figure 2-7) and explain the peaks in total kelp in Figure 2-5. I observed large recruitment events for the kelp Nereocystis luetkeana in the spring of both post treatment years. N. luetkeana is an annual species that recruits in late winter, grows rapidly to the surface, and dies back in early fall. Macrocystis  58  integrifolia is the other canopy-forming species encountered within the study. M. integrifolia was found in very low numbers (Table 2-4). However, it is interesting to note that M. integrifolia was best established at Site 1. Site 1 showed a lower response by N. luetkeana and had a few large and well-developed M. integrifolia. A typical M. integrifolia plant at Site 1 would have about 13 stipes coming from a large holdfast (approximately 50-cm at widest point) and weigh on the order of 9 kg without the holdfast. A correction for biomass would indicate a better response by Site 1 to removal of urchins because of the few M. integrifolia present. However, given the low numbers of M. integrifolia and the related sampling concerns, such a correction is not presented. However, I did run the same model ANOVA as in Table 2-1 substituting biomass values obtained by utilising the regression equations relating the size of kelp plants to mass (Appendix; Figure 1-9 through Figure 1-13). This approach did not alter my interpretation of the data (see Appendix; Table 1-4 for biomass by species for each plot at termination of the experiment). Plot by treatment interaction for kelp (Figure 2-8) can be interpreted similarly to that for urchins. However, here it is the total removal plots not the controls that cause the interaction. Sites 2 and 3 had a higher magnitude response to total urchin removal than did Site 1. The deviation in response among replicates creates the statistical interaction. Ultimately, it is the treatment that created this response disparity, so the interaction is further evidence for the importance of urchin removal to the kelp response. Figure 2-8 also illustrates the blocking effect in the ANOVA model. Similar replicates of this experiment will vary depending on where they are performed. Abiotic factors, such as substratum type, sand scouring, shading, wave exposure and turbulence all potentially contribute to the local conditions that affect kelp recruitment and growth. However, given that the kelp response to  59  urchin density varies among sites in magnitude but not in the mathematical function, I predict, that replicates of this experiment would produce similar results in any temperate waters where kelp and urchins are found.  Table 2-2. Species of kelp encountered throughout the study listed as a proportion of occurrence in samples over the total of all observations in all time periods. Species  Common Name  Form  Laminaria bongardiana  Laminaria  Understory  51.30  Nereocystis luetkeana  Bull kelp  Canopy  23.98  Pterygophora californica  Walking kelp  Understory**  11.66  Unknown*  N/A  Understory  10.47  Macrocystis integrifolia  Giant kelp  Canopy  1.28  Costaria costata  Ribbed kelp  Understory  1.19  Hedophyllum sessile  Sea cabbage  Understory / Intertidal  0.04  Alaria sp.  Winged kelp  Understory  0.04  Egregia menziesii  Feather boa kelp  Understory / Intertidal  0.04  % of observations  ""Unknown were too small or too damaged to tell apart. In all cases the choice was between L. bongardiana and P. californica. ** P. californica often forms an understory or middle canopy due to its upright stature. Individuals in this study were typically similar in size to L. bongardiana and are thus classified as understory.  Table 2-3. Split-plot analysis of variance for log (data +1) of kelp densities. Treatment effects tested using the pseudo f-test. Error d.f. for treatment effects is 6.23 (refer to Hicks 1982). 10  Effect  D.F.  MS  Error M S  F  P  Block  2  22.64  2.04  11.09  0.000  Treatment  2  90.04  10.07*  8.95  0.012  Error 1 (Treatment x Block)  4  7.74  2.04  3.79  0.009  Time  8  20.29  2.04  9.94  0.000  Time x Treatment  16  4.36  2.04  2.14  0.022  Error 2  48  2.04  —  —  —  Sampling Error  1656  0.17  —  —  —  62  Kelp Density Over Time Mean kelp densities with standard errors  Figure 2-4. Kelp density as a function of time for combined kelp species in each treatment. Series represent each of the three study sites. Bars are ± one standard error of the mean (n = 18 for Total and Control; n = 36 for Intermediate). Vertical lines after fall 1997 represents the implementation of treatments. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding each sampling season is given in the Appendix (Appendix; Table 1-3).  T i m e by Treatment Interaction M e a n kelp densities a n d standard errors 70  Figure 2-5. Kelp density as a function of time for all kelp species averaged across sites. Bars are ± one standard error of the mean (n = 54 for Total and Control; n = 108 for Intermediate). Vertical line represents the time of treatment implementation. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3).  Understory Kelp Density Over Time Mean density and standard errors o co  Figure 2-6. Kelp density as a function of time for understory kelp species averaged across sites. Bars are ± one standard error of the mean (n = 54 for Total and Control; n = 108 for Intermediate). Vertical line represents the time of treatment implementation. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3).  Canopy Kelp Density Over Time Means and standard errors  35  Figure 2-7. Kelp density as a function of time for canopy kelp species averaged across sites. Bars are ± one standard error of the mean (n = 54 for Total and Control; n = 108 for Intermediate). Vertical line represents the time of treatment implementation. Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3).  Site by Treatment Interaction Mean kelp densities and standard errors  Figure 2-8. Density as a function of site for all kelp species with treatments plotted as series. Bars are ± one standard error of the mean (n = 18 for Total and Control; n = 36 for Intermediate). Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively.  Kelp Diversity Simpson's diversity index (Simpson 1949) was used to determine the effects of variable urchin density on the diversity of kelp (Table 2-4). The medium and low treatment levels were not combined with this analysis as the increase in area would affect the possible number of species encountered (Arrhenius 1921). As with other components of the experiment, the trend in diversity was not consistent among all three sites. Again, Site 1 tends to stand out as different from Sites 2 and 3. Site 1 clearly shows diversity increasing with increased urchin removal. Site 2 diversity indices were relatively equal among all three removal levels with medium removal resulting in the highest diversity. Within Site 3, intermediate removal plots were most diverse. In all cases, the controls were devoid of kelp and therefore have no kelp diversity. The one exception to this was the Site 2 control plot, which did contain some kelp. Those data are omitted because they represented only a single shallow-depth point within the plot. This shallow point in the reef provides an obvious refuge for kelp as strong wave action excludes urchins (Schroeter 1978). I also calculated the mean proportion for each species among the three sites and used these values to calculate a mean diversity index for each of the treatments. This approach shows that intermediate levels of removal created the highest level of kelp diversity. That finding supports my field observations. Intermediate levels of removal created a mosaic of kelp- and urchin-dominated zones within their respective plots. This mosaic allowed for more than just kelp diversity, allowing for the presence of both large, fleshy kelps as well as crustose and coralline red algae, which were not counted in the surveys.  Table 2-4. Kelp diversity (Simpson 1949) for each site and mean for all sites at termination of the experiment. Numbers in parentheses are the total number of sub-samples taken from the total kelp encountered at each plot. Cells marked '—' represent an absence of kelp for that species. Letter codes are: Cs = C. costata, Lm = L. bongardiana, Mc = M. integrifolia, Nr = TV. luetkeana, Pt = P. californica. Treatment labels 'total', 'low', 'med', and 'control' refer to the amount of urchin removal as total, low, medium, or no removal, respectively.  Proportion of Species Cs  Lm  Mc  Nr  Pt  Diversity  Total (10)  —  0.40  0.10  0.20  0.30  3.33  Low (10)  —  —  0.50  0.40  0.10  2.38  Med (10)  —  —  0.10  —  0.90  1.22  Control (10)  —  —  —  ...  ...  0.00  Total (56)  —  0.71  —  0.23  0.05  1.76  Low (56)  0.05  0.77  —  0.09  0.09  1.64  Med (56)  —  0.70  —  0.18  0.13  1.88  Control (56)  —  —  —  —  ...  0.00  Total (29)  —  0.83  —  0.07  0.10  1.43  Low (29)  —  0.48  —  0.24  0.28  4.02  Med (29)  —  0.59  —  0.38  0.03  2.05  Control (29)  —  —  —  —  ...  0.00  Total  —  0.65  0.03  0.17  0.15  2.12  Low  0.02  0.42  0.17  0.24  0.16  3.50  Med  —  0.43  0.03  0.19  0.35  2.91  Control  —  —  —  ...  Site 1  .  Site 2  Site 3  •  Mean  0.00  '  Urchins vs. Kelp I conducted regression analyses to determine the functional response of kelp density after manipulating urchin density. The differences between sites noted above (Figure 2-8) prevent the grouping of data into a single response. As noted in the ANOVA for the kelp response, it is the magnitude of the response and not the mathematical function that creates the differences between sites (Figure 2-9). Site 1 is again markedly different from Sites 2 and 3. At Site 1, kelp density reached only a maximum of 16 stipes/m . Sites 2 and 3 showed a much higher response with maximum kelp densities of 113 and 91 stipes / m , respectively. The relationship at each site follows an 2  exponential decay function (Figure 2-9). After taking the logarithms (data +1) for urchin and kelp densities the data were analysed as a simple linear regression. The fitted regression equations show that in all three cases urchin density explained a significant proportion of the variance in kelp densities (see Figure 2-10 for fitted log-log scatterplots with equations and p-values). Observation of the scatterplots (Figure 2-9) illustrates the biological importance of urchin removal. There is an apparent threshold of urchin density that releases kelpfromthe effects of grazing and allows for persistence of the kelp forest community. From visual estimation, the presence of this threshold lies at approximately 0.6 urchins / m for Site 1, 1.0 urchins / m for Site 2, and 1.3 urchins / m for Site 3. Although there are differences between the three sites in this study, the presence of the threshold still appears to lie within a narrow range. This range may be affected by urchin size in addition to abundance. However, examination of urchin test diameters indicates that there is little variation in the  proportional presence of urchin size classes among experimental units (see Appendix; Figure 1-5).  Plots containing densities of urchins higher than the proposed thresholds contain little if any kelp whereas below the urchin density threshold kelp was releasedfromherbivore pressure and density increased. Below the urchin threshold, kelp densities were limited by either the remaining urchins, inter- and intraspecific competition with kelp species, or by physical site factors. The total removal plots within each site allow for complete release from the effects of urchins and for much increased kelp densities. Within these plots, competition among kelps most likely sets the limit on kelp density. The intermediate urchin removal plots are the most complex. The urchin and kelp distributions tend to be clumped at these plots. The clumping of kelp comes about by the foraging of urchin groups, while within clumps the kelp density is probably governed by competition. At all plots, physical and environmental factors affect both kelp and urchins. Site 1 is an example of how physical factors may be working to affect urchins and kelp. The lower abundance of solid substrate and the presence of sand at Site 1 (Appendix; Figure 1-6) tends to decrease the amount of kelp and urchins within the site. The presence of less stable substrates means less attachment area for both kelp and urchins while the extra sand further enhances this effect by scouring rocks and preventing possible recruitment of available substrate by kelp gametophytes.  71  C/3  E © 03  u u  >! w  c« « e u "cL *> "C  B H >« H a,a — u iCU k> C/}CU « S !•  to  £  _  c 0)  .a s  V  £f  _  03 03  s  cu 3  Q)  CO  £  X. TJ C CD  .2 « _ 5 CO  O V O 1-  O i_  2 §• it, o o o o o o o o o o o o o rsi i - o o> ooh- co m f co CN T CO  £  w e (U  S I  <D  CO .Q  3  O  O tf)  08 cu  E « « .a  | _ CO  © ^i. i - <s  <2 li cu S w i_ o  o Q. L _  CD  o o o o o o o o o o o o o csii-ooooi-.tDin^rcocN't-  -•—» ^—»  o CO  ©  CO  Q  CO  ® £  S *S a. >-  (0  D TJ 0  cu .O  0) CO  n . <u  i § •o ** _ u s u is 'cT  U 1- o >, cu  .2 >, 5  B  u u s  CI-I  O "« o  a  cu i_Q. S >> cu "cu XJ u cu 03 OS S W  «s OJD  g 2  R£ U B 03 5 <S  72  CA B  2 "5  '2 .5 "2  ° — 5.  * « 2 « ts s  co  ii S © •3 Jj is  </)  w  c  <u «u  •§ « S  Q  Q CU *"*  ss» £  o  a . ij  8£5  Z .2  4>  » ~ s .O  O  _1  0  a.  «  CM  E  g  w c  o c  CD  CO  Lire w  w  Q  cu cu  .9 S JS J3 © ~ u u CU  3t3  p  ° i? . a S + - .2  (0  (3  <o  o  <D '•^ (7)  C  Q Q.  3  O  M A O B — O  L M cu  CU CA  1M  cu  cu  °  S  .2  ~  0  I  c  0  4—  o D) O  CA fi £ cs  cs  ^  >. CA W  •5 fi u •a a  g « S Ts c  u  "2 B  « ®  e> O V) ON cu 03  C*H  CNJ O CO CD rt CN O ^ ^ d d d d d  o  +  ( LU) Aiisuaa d|9» S  B  Bon  -  CU  IS  o _a >o t3 as o c S  •- £  J  " 2 »3 M  2  cu  CS -> O M  E S £  DISCUSSION Results show that S.franciscanusoccurs in densities high enough to defoliate many rocky bottom habitats in this portion of Clayoquot Sound, British Columbia. Although the potential annual production at these sites is unknown, observed urchin densities can reduce kelp density by up to 100%. Study data, and personal observations of urchin densities throughout much of Clayoquot Sound, indicate that sea urchins influence the distribution of the kelp forest habitat in this region. The control of kelp forest community structure by S.franciscanusdisplays the primacy of top-down community control. Sea urchin predation by either sea otters or humans seems necessary to prevent kelp habitats from cascading into non-vegetated urchin barrens. Trophic cascade is the term used to describe downward dominance through food chains. Strong (1992) argues convincingly that trophic cascades are relatively unusual and generally restricted to low-diversity systems where one or few species have great influence. Often these low diversity systems are aquatic with autotrophs lacking the well-developed defence systems found in terrestrial systems. Alternately, terrestrial systems are receiving more attention as potentially cascading with the loss of top carnivores (Schmitz et al. 2000). However, analysis of this system as a trophic cascade needs to be viewed with caution. Hunter and Price (1992) point out that top-down effects cannot exist without bottom-up effects. In other words, without factors such as nutrients there can be no autotrophs and without autotrophs there is no community. In this experiment, the differences between sites can be a function of physical processes as noted above or may also be the result of differences in bottom-up influences among the sites. Unfortunately, determination of bottom-up effects is not within the scope of this experiment. Regardless, the presence of  kelp forests is seemingly limited by a lack of predation on urchins. In treating this system as a trophic cascade, it is important to understand the point at which predation can control the cascade and the effects of varying levels of predation on community composition. From the results, it appears that disturbances that reduce urchin abundance to between 0.6 and 1.3 urchins / m are capable of reversing the trophic cascades induced by sea otter removal. The results of this investigation show that shallow rocky-bottom communities in the Hesquiat territory can exist as either kelp forests or urchin barrens. Intermediates between these states are likely to be unstable given the shape of the plots of urchin density and kelp density (Figure 2-9). The shape of the curve fitting these variables supports the idea that there is a threshold separating two domains of attraction (Holling 1973). As urchin densities approach the threshold, the system can quickly flipfromone community type to the other. Alternate community compositions occur in a variety of conditions based on historical conditions such as order of species recruitment (Sutherland 1974), herbivory (Lubchenco 1978), and natural disturbance (Dethier 1984). Many authors have referred to these alternate compositions as multiple stable states, or alternate stable points (Sutherland 1974, Dublin et al. 1990, Janse 1997). The application of this terminology is loose and often inappropriate with each author having toframewhat stability means within their study system. To demonstrate that multiple stable points exist, it must be shown that with similar community composition, under similar environmental conditions, and in the absence of artificial controls, two or more stable points exist and that each state can persist indefinitely (Krebs 1985, Dublin et al. 1990). In the real world, stochastic perturbations are the norm, preventing these alternate states from existing for extended periods. Thus, the ability to prove that multiple stable points occur in nature is often suspect (Connell and Sousa 1983).  I prefer to think of this system as one with multiple community compositions as opposed to multiple stable states. In this scenario, the community can exist in alternate forms with different species assemblages based on abiotic and biotic factors. In this study, the critical factor determining alternate community types was urchin removal. Thus, the two community types are boundary points (Lewontin 1969) with domains of attraction (Holling 1973) on either side of the urchin abundance threshold. Small disturbances within each domain of attraction will not affect the system composition and will allow it to return to the previous boundary point. However, at some point, no system is stable enough to withstand any and all punishment. As perturbations increase in intensity, stability of one boundary point becomes overwhelmed and the system flips into another domain of attraction ending in a new boundary point. Therefore, they cannot be alternate stable points in the truest sense of the term because they will ultimately convert to the opposite state under some set of natural conditions. These results are empirical evidence in support of observationsfromother studies. Estes and Duggins (1995) found a strongly hyperbolic relationship between strongylocentrotid urchin biomass and kelp density in the Aleutian Islands and southeast Alaska. In the Caribbean, Steneck (1993) observed a similar pattern between sea urchin abundance and reef algae following a massive die-off of Diadema antillarum.  However all  of these results are in contrast to Ebert (1977) who observed a linear relationship between urchin and algal abundance with 41 S. franciscanus I m being required to completely 2  remove fleshy macrophytes. Ebert's observations are likely differentfromthis study for two reasons. First, his use of a rock jetty as an experimental site increases the actual surface area measured by fixed quadrats. In this study, lead transect lines were used to estimate urchin  76  abundances and so a given size quadrat represents the same surface area among sites regardless of the amount of relief. The rock jetty in Ebert's study had high relief and his quadrats were likely encompassing a greater number of urchins per unit area than measured in this study. An adjustment for actual bottom area would only decrease the magnitude of his findings but not the relative relationship. A more likely explanation for his results involves the type of manipulation he performed. By adding instead of removing urchins, his study displays that once established, kelp production can maintain high densities of urchins. I argue that manipulations in the opposite direction would not allow kelp recruitment in a linear fashion (see Figure 2-11 for a model of alternation between community compositions). In non-manipulated sea urchin-kelp studies, the same conclusions can be drawn with the mechanisms flipping the system between two domains of attraction being different. The effects of climate (Tegner and Dayton 1991, Dayton et al. 1998), storms (Harrold and Reed 1985), poor algal recruitment (Dean et al. 1984), high algal recruitment (S. Schroeter pers. comm.), disease (Scheibling 1986, Tegner et al. 1995), and predation (Watson 1993, Estes and Duggins 1995) have all been shown to occur at magnitudes significant enough to flip the system from one community composition to another. In each of the former studies, the mechanisms causing the alternation between community types were different, but the observed patterns are similar. Interaction between a perturbation and urchin presence pushed the system toward one of two domains of attraction, each ending in an algal community or urchin-barren boundary point. In the study area, urchin barrens are likely more stable than kelp forests. The proof of this seems to be the widespread abundance of urchin barrens along the Hesquiat open coastline. Kelp forests only exist in isolated patches where urchin populations have likely  77  been perturbed by storms, fishing, or predators. However, unless the perturbations are extremely large, such islands of kelp habitat will likely be grazed by urchins with the system quickly returning to the urchin barren state (local stability). On the other hand, urchin barrens may not be globally stable. Imagine widespread harvesting of sea urchins or the large perturbations caused by sea otters. In these scenarios, a critical point is crossed where the abundance of urchins is not significant enough to push the system back to the barren state. Urchin densities will not necessarily continue to crash; the remaining urchins will coexist with kelp forests because increased primary production will provide plentiful drift kelp for the urchins. Once this state has been achieved, the urchin population can increase, perhaps even to previous levels, due to the high productivity of kelps. Perturbation of kelp forests can reverse the process. If kelps experience multiple years of poor recruitment or storms for instance, then the abundance of algal drift may be decreased. In this case, sea urchins will begin to actively graze the attached plants causing them to weaken and wash away creating the barrens state once again (see Figure 2-11). Although the mechanisms responsible for flipping an area between kelp forest and urchin barrens are generally known, the threshold about which such changes occur is little studied. The paucity of research utilising intermediate levels of urchin removal (Table 1-1) means that the generality of the kelp-urchin density relationship discovered in this study require confirmation by future research.  Kelp Forest  -a a 3  < a  Perturbations resulting in loss of kelp forests • High urchin recruitment • Storms Poor kelp recruitment Kelp harvesting Climate change  CO CJ  a  Perturbations resulting in loss of urchin barrens • Urchin harvesting • Disease • Predation  Increasing Urchin Abundance  Figure 2-11. Illustration of potential connections between the alternate community types. As urchin abundance increases, the system may or may not flip from kelp forest to urchin barren dependent upon the type and intensity of perturbations. Once urchin barrens are established they become stable, preventing kelp recruitment even if densities drop below the prior levels when the system was a kelp forest. Therefore, kelp forests can be maintained at a variety of urchin densities, but can only be initiated at relatively low urchin densities.  It has been hypothesised that intermediate levels of disturbance promote greater biological diversity than respectively higher or lower levels of disturbance (Connell 1978). Macrophytic brown algae (kelps) have great influence on the presence of other non-kelp species. Kelp plants provide food, substrate, and shelter for a variety of vertebrate and invertebrate species (Harrold and Pearse 1987). Edwards (1980) found 62 invertebrate species in just the holdfasts of Laminaria hyperborea in the northern Atlantic. Wheeler and Druehl (1986) calculated that up to 11 m of kelp surface is provided for every 1 m of ocean bottom inhabited by Macrocystis integrifolia. Wing and Clendenning (1971) calculated that 2  2  M. pyrifera provided 15.4 m of surface area for every m of substratum in a southern California kelp bed. They also found 2811 individuals per dm of kelp blade belonging to eight taxonomic groups (ostracods, copepods, amphipods, decapods, polychaetes, nematodes, turbellarians, and molluscs). The increased physical structure of kelp forests also provides habitat for numerous species offish(see Chapter 5). Based on this information and the results presented, it can be seen that removal of sea urchins (disturbance) results in increased abundance of kelp species and thus greater diversity of kelp and associated species can be achieved than without removal of urchins. The intermediate disturbance hypothesis predicts that there will be a point where increased levels of disturbance result in a lowering of biotic diversity. The results for kelp diversity do not completely support this portion of the theory (sensu Connell 1978). Intermediate levels of removal did not always create higher levels of kelp diversity than total urchin removal plots. However, when data were averaged across sites, the resulting mean diversity index was clearly higher for the intermediate levels of removal. Although the trend is not consistent for all replicates, on average these results tend to support the intermediate  80  disturbance hypothesis. It could be argued that the high diversity indices observed for total removal plots comefromone or both of two critical components. First, succession may not have had adequate time to convert the high disturbance areas to a system dominated by fewer climax kelp species such as was observed by Duggins (1980). Or secondly, the urchins in this study are not selecting the dominant kelp species when choice is available (Lubchenco 1978). If Duggins observations pertain to this study, they would help to support the intermediate disturbance hypothesis more so than this study alone. If the Lubchenco's hypothesis holds, then intermediate removals may decrease in diversity over time weakening this study's support of the intermediate disturbance hypothesis. Given that urchins often prefer to graze on large community-structuring species to other algal forms (e.g. N. luetkeana) (Irvine 1973, Vadas 1977), it is likely that intermediate removals of urchins would ultimately lead to greater diversity in the overall algal community, even if such diversity is not represented by only observing kelps. The barren space made available by partial herbivory in intermediate removal plots is seemingly open for more "fugitive" species (sensu Hutchinson 1951) such as C. costata and Alaria spp. (Duggins 1980). The fact that these species were rarely encountered is probably the result of low spore availability. Given more time, I would expect greater abundance of fugitive species in the intermediate removal plots. Further, general observations support the intermediate disturbance hypothesis. The conversion of barrens to kelp forest created different patterns between total and intermediate removals. Specifically, intermediate removal of sea urchins created a mosaic of kelp and non-kelp patches whereas total removal tended to create more continuous kelp forest. While the kelp forests resultingfromtotal urchin removal tended to attract many vertebrate and invertebrate species not under direct  81  study, the intermediate removals also provided such habitat but in lower supply. However, the intermediate removal plots also contained groups of urchins that prevented total monopolisation by kelp. This condition tended to allow for remnantsfromthe previous barren habitat such as coralline and crustose algae to coexist with the newly created kelp forest patches. In the absence of herbivores, numerous authors have shown that competition among algal species determines community composition (Dayton, 1975, Lubchenco 1978, Duggins 1980, Kim 1997). Often, when marine herbivores are completely removed, the associated algal community becomes dominated by one or few species competing for resources such as light (Pearse and Hines 1979, Reed and Foster 1984). Duggins (1980) performed an urchin manipulation similar to the total removal component of this study. Duggins performed total removals of sea urchins in southeast Alaska and observed recruitment of species similar to those found in this study. His observations on the competitive interaction between Laminaria groenlandica (probably the same species as L. bongardiana in this study) and Nereocystis luetkeana differ from the patterns I observed. Duggins found that in the first year following removal, Nereocystis (and other annuals) were most abundant and Laminaria was sparse (8% of total biomass). However, in the second year, Laminaria accounted for more than 65% of the total biomass with Nereocystis abundance being much reduced. Duggins attributed this shift to the competitive superiority of Laminaria, attained through the large blade of L. groenlandica, which he states does not significantly die back in the winter. The blade presumably limits the available light for recruiting Nereocystis and can abrade the substratum, killing the microscopic gametophyte stages of its competitors. The post-removal observations in this study are in disagreement with Duggins'  82  (1980) observations. Nereocystis recruitment was not prevented in the second year in either of Sites 2 or 3 (Nereocystis is the dominant kelp creating the response seen in Figure 2-7). Sites 2 and 3 had consistent levels of L. bongardiana throughout the study with slight losses in the fall replaced in the spring (L. bongardiana makes up the majority of the observations in Figure 2-6). Thus, the generality of Duggins' observations on competition between Nereocystis and Laminaria may not apply in this system or more time may be required for Laminaria to dominate. If the generality does not hold, it may be due to any or all of the following reasons. First, in southeast Alaska, kelp densities in excess of 300 stipes / m were observed in the first year after removal. In this study, the maximum observed average of all species combined was 91 stipes / m for Site 3. Perhaps some factor other than competition 2  is regulating kelp abundances below the point at which Laminaria might dominate. Alternatively, perhaps my study sites are more topographically complex, allowing for microsite refugia that protect Nereocystis from abrasion by Laminaria. Finally, and most likely, significant winter loss of blades were noted for L. bongardiana at these study sites. Perhaps this difference is due to greater waves and surge at this study site or to phenotypic plasticity in the species. Whatever the reason causing the observed differences, the generality of competitive exclusion of Nereocystis by Laminaria needs further investigation.  CONCLUSIONS AND IMPLICATIONS The results of this experiment have important implications for S. franciscanus harvesting strategies and contribute to our understanding of ecological theory. The urchin population studied here has been shown to be food limited (Chapter 3). As well, other researchers have found that under a variety of conditions supplementing sea urchins with additional food improves the quality of their reproductive organs and thus marketability  (gonads are the only edible portion of a sea urchin) (Lawrence et al. 1997, McBride et al. 1998). The strength and function of the kelp response to reduced urchin density can be used to help develop a more efficient and profitable sea urchin fishery while managing for the conservation of kelp forests and other associated fisheries (see Chapter 5). The results of this study support the hypothesis that sea urchins are limiting the presence of kelp forests on the exposed coast of Clayoquot Sound, BC. Moreover, these results highlight the relevance of some ecological theories. The effects of removing & franciscanus on this community support the idea of a top-down trophic cascade (Power 1992, Strong 1992). In this case, humans take on the role of a "keystone species" (sensu Paine 1969) and have top-down control of community composition by removing sea urchins. The parabolic relationship between urchin and kelp abundance results in a quick conversion of urchin barrens to kelp forest, illustrating that multiple community compositions can exist in the same spatial unit based on the management of S. franciscanus. Finally, the degree and frequency with which urchins are removed not only affects the presence of kelp forests, but can also influence the biological diversity within kelp forests. The above observations have implications for how we can sustainably and efficiently manage sea urchin harvesting for both economic gain and conservation of kelp forests. However, the results of this study are somewhat limited by the variable response across study sites. Future work should focus on larger scale experiments that encompass greater spatial variation. Such experiments can only be accomplished by actually harvesting urchins. By increasing the scale of the experiment we remove problems associated with urchins reinvading plots and can tend towards averaging out effects of micro-site differences in physical factors that can affect experiments that take place on smaller spatial scales.  84  Through large-scale experiments with the input of managers, resource users, and scientists we can improve harvests while meeting management objectives relating to habitat and fisheries conservation (see Chapter 6 for a thorough discussion of implications).  REFERENCES Andrew, N. L., and A. J. Underwood. 1993. Density-dependent foraging in the sea urchin Centrostephanus rodgersii on shallow subtidal reefs in New South Wales, Australia. Marine Ecology Progress Series 99:89-98. Arrhenius, O. 1921. Species and area. Journal of Ecology 9:95-99. Ayling. 1981. The role of biological disturbance in temperate subtidal encrusting communities. Ecology 62:830-847. Briscoe, C. S., and K. P. Sebens. 1988. Omnivory in Strongylocentrotus droebachiensis (Muller) (Echinodermata: Echinoidea): predation on subtidal mussels. Journal of Experimental Marine Biology and Ecology 115:1-24. Carpenter, S. R., J. F. Kitchell, J. R. Hodgson, P. A. Cochran, J. J. Elser, M. M. Elser, D. M. Lodge, D. Kretchmer, X. He, and C. N. von Ende. 1987. Regulation of lake primary productivity by food web structure. Ecology 68:1863-1876. Connell, J. H. 1978. diversity in tropical rain forests and coral reefs. Science 199:13021310. Connell, J. H., and W. P. Sousa. 1983. On the evidence needed to judge ecological stability or persistence. The American Naturalist 121:789-824. Dayton, P. K. 1975. Experimental studies of algal canopy interactions in a sea otterdominated kelp community at Amchitka Island, Alaska. Fishery Bulletin 73:230-237. Dayton, P. K., M. J. Tegner, P. B. Edwards, and K. L. Riser. 1998. Sliding baselines, ghosts, and reduced expectations in kelp forest communities. Ecological Applications 8:309-321. Dean, T. A., S. C. Schroeter, and J. D. Dixon. 1984. Effects of grazing by two species of sea urchins (Strongylocentrotus franciscanus and Lytechinus anamesus) on recruitment and survival of two species of kelp (Macrocystis pyrifera and Pterygophora californica). Marine Biology 78:301-313. Dean, T. A., F. R. Jacobsen, K. Thies, and S. L. Lagos. 1988. Differential effects of grazing by white sea urchins on recruitment of brown algae. Marine Ecology Progress Series 48:99-102. 85  Dethier, M. N. 1984. Disturbance and recovery in intertidal pools: maintenance of mosaic patterns. Ecological Monographs 54:99-118. [DFO] Department of Fisheries and Oceans Canada. Pacific Region. 1999. Pacific region 1999/2000 management plan, red sea urchin. Vancouver, B.C. 30p; Appendices. Dublin, H. T., A. R. E. Sinclair, and J. McGlade. 1990. Elephants and fire as causes of multiple stable states in the Serengeti-Mara woodlands. Journal of Animal Ecology 59:1147-1164. Duggins, D. O. 1980. Kelp beds and sea otters: an experimental approach. Ecology 61:447453. Duggins, D. O. 1981. Sea urchins and kelp: the effects of short term changes in urchin diet. Limnology and Oceanography 26:391-394. Duggins, D. O. 1983. Starfish predation and the creation of mosaic patterns in a kelpdominated community. Ecology 64:1610-1619. (  Duggins, D. O., C. A. Simenstad, and J. A. Estes. 1989. Magnification of secondary production by kelp detritus in coastal marine ecosystems. Science 245:170-173. Ebeling, A. W., and D. R. Laur. 1984. The influence of plant cover on surfperch abundance at an offshore temperate reef. Environmental Biology of Fishes 12:169-179. Ebert, T. A. 1967. Negative growth and longevity in the purple sea urchin Strongylocentrotus purpuratus (Stimpson). Science 157:557-558. Ebert, T. A. 1968. Growth rates of the sea urchin Strongylocentrotus purpuratus related to food availability and spine abrasion. Ecology 49:1075-1091. Ebert, T. A. 1977. An experimental analysis of sea urchin dynamics and community interactions on a rock jetty. Journal of Experimental Marine Biology and Ecology 27:1-22. Edwards, A. 1980. Ecological studies of the kelp, Laminaria hyperborea, and its associated fauna in south-west Ireland. Ophelia 19:47-60. Estes, J. A., and D. O. Duggins. 1995. Sea otters and kelp forest in Alaska: generality and variation in a community ecological paradigm. Ecological Monographs 65:75-100. Estes, J. A., and J. F. Palmisano. 1974. Sea Otters: their role in structuring nearshore communities. Science 185:1058-1060.  Fletcher, W. J. 1987. Interactions among subtidal Australian sea urchins, gastropods, and algae: effects of experimental removals. Ecological Monographs 57:89-109. Foreman, R. E. 1977. Benthic community modification and recovery following intensive grazing by Strongylocentrotus droebachiensis. Helgolander Wissenschaftliche Meeresuntersuchungen 30:468-484. Gabrielson, P. W., T. B. Widdowson, S. C. Lindstrom, M. W. Hawkes, and R. F. Scagel. 2000. Keys to the benthic marine algae and seagrasses of British Columbia, Southeast Alaska, Washington and Oregon, phycological contribution no. 5. University of British Columbia, Vancouver. Garnick, E. 1989. Lobster (Homarus americanus) population declines, sea urchins, and 'barren grounds': a space-mediated competition hypothesis. Marine Ecology Progress Series 58:23-28. Hagen, N. T., and K. H. Mann. 1992. Functional response of the predators American lobster Homarus americanus (Milne-Edwards) and Atlantic wolfish Anarhichas lupus (L.) to increasing numbers of the green sea urchin Strongylocentrotus droebachiensis (Muller). Journal of Experimental Marine Biology and Ecology 159:89-112. Harper, J. L. 1969. The role of predation in vegetational diversity. Brookhaven Symposia in Biology no. 22. pp. 48-62. Harrold, C , and J. S. Pearse. 1987. The ecological role of echinoderms in kelp forests. Pages 137-233 in M. Jangoux and J. M. Lawrence, editors. Echinoderm studies 2. A.A. Balkema, Rotterdam. Harrold, C , and D. C. Reed. 1985. Food availability, sea urchin grazing, and kelp forest community structure. Ecology 66:1160-1169. Hicks, C. R. 1983. Fundamental concepts in the design of experiments, 3rd edition. Holt, Rinehart, and Winston, New York. Hobson, E. S., and J. R. Chess. 1976. Trophic interactions among fishes and zooplankters near shore at Santa Catalina Island, California. Fishery Bulletin 74:567-598. Holling, C. S. 1973. Resilience and stability of ecological systems. Annual Review of Ecology and Systematics 4:1-23. Hughes, T. P., D. C. Reed, and M. Boyle. 1987. Herbivory on coral reefs: community structure following mass mortalities of sea urchins. Journal of Experimental Marine Biology and Ecology 113:39-59.  87  Hunter, M. D., and P. W. Price. 1992. Playing chutes and ladders: heterogeneity and the relative roles of bottom-up and top-down forces in natural communities. Ecology 73:724-732. Hurlbert, S. H. 1984. Pseudoreplication and the design of ecological field experiments. Ecological Monographs 54:187-211. Hutchinson, G. E. 1951. Copepodology for the ornithologist. Ecology 32:571-577. Irvine, G. V. 1973. The effects of selective feeding by two species of sea urchins on the structure of algal communities. M.Sc. thesis. University of Washington, Seattle. Janse, J. H. 1997. A model of nutrient dynamics in shallow lakes in relation to multiple stable states. Hydrobiologia 342/343:1-8. Johnson, C. R., and K. H. Mann. 1982. Adaptations of Strongylocentrotus droebachiensis for survival on barren grounds in Nova Scotia. Pages 277-283 in J. M. Lawrence, editor. Echinoderms: proceedings of the international conference, Tampa Bay. A.A. Balkema, Rotterdam. Keats, D. W., G. R. South, and D. H. Steele. 1990. Effects of an experimental reduction in grazing by green sea urchins on a benthic macroalgal community in eastern Newfoundland. Marine Ecology Progress Series 68:181 -193. Kim, J. H. 1997. The role of herbivory, and direct and indirect interactions, in algal succession. Journal of Experimental Marine Biology and Ecology 217:119-135. Krebs, C. J. 1985. Ecology: The experimental analysis of distribution and abundance. Harper & Row, New York. Kuehl, R. O. 1994. Statistical principles of research design and analysis. Duxbury Press, Belmont, CA. Lang, C , and K. H. Mann. 1976. Changes in sea urchin populations after the destruction of kelp beds. Marine Biology 36:321-326. Lawrence, J. M. 1970. The effect of starvation on the lipid and carbohydrate levels of the gut of the tropical sea urchin Echinometra mathaei (de Blainville). Pacific Science 24:487-489. Lawrence, J. M. 1975. On the relationships between marine plants and sea urchins. Oceanography and Marine Biology 13:213-286. Lawrence, J. M., S. Olave, R. Otaiza, A. L. Lawrence, and E. Bustos. 1997. Enhancement of gonad production in the sea urchin Loxechinus albus in Chile fed extruded feeds. Journal of the World Aquaculture Society 28:91-96.  Lees, D. C. 1970. The relationship between movement and available food in the sea urchins Strongylocentrotusfranciscanusand Strongylocentrotus purpuratus. M.Sc. thesis. San Diego State College, San Diego. Leighton, D. L. 1960. Studies of kelp-grazing organisms. In, Kelp investigation program, quarterly report, University of California, Institute of Marine Research, IMR Ref. 607, 13-22. Leinaas, H. P., and H. Christie. 1996. Effects of removing sea urchins (Strongylocentrotus droebachiensis): stability of the barren state and succession of kelp forest recovery in the east Atlantic. Oecologia 105:524-536. Levitan, D. R. 1992. Community structure in times past: influence of human fishing pressure on algal-urchin interactions. Ecology 73:1597-1605. Lewontin, R. C. 1969. The meaning of stability. Brookhaven Symposia in Biology 22:1324. Low, C. J. 1975. The effect of grouping of Strongylocentrotus franciscanus, the giant red sea urchin, on its population biology. Ph.D. dissertation. University of British Columbia, Vancouver. Lowry, L. F., A. J. McElroy, and J. S. Pears. 1974. The distribution of six species of gastropod molluscs in a California kelp forest. Biological Bulletin 147:386-396. Lubchenco, J. 1978. Plant species diversity in a marine intertidal community: importance of herbivore food preference and algal competitive abilities. The American Naturalist 112:23-39. Mann, K. H. 1977. Destruction of kelp-beds by sea urchins: a cyclical phenomenon or irreversible degradation? Helgolander Wissenschaftliche Meeresuntersuchungen 30:455-467. Mattison, J. E., J. D. Trent, A. L. Shanks, T. B. Akin, and J. S. Pearse. 1977. Movement and feeding activity of red sea urchins (Strongylocentrotusfranciscanus)adjacent to a kelp forest. Marine Biology 39:25-30. McBride, S. C , J. M. Lawrence, A. L. Lawrence, and T. J. Mulligan. 1998. The effect of protein concentration in prepared feeds on growth, feeding rate, total organic absorption, and gross assimilation efficiency of the sea urchin Strongylocentrotus franciscanus. Journal of Shellfish Research 17:1563-1570. McClanahan, T. R., and J. D. Kurtis. 1991. Population regulation of the rock-boring sea urchin Echinometra mathaei (de Blainville). Journal of Experimental Marine Biology and Ecology 147:121-146.  89  McClanahan, T. R., and E. Sala. 1997. A Mediterranean rocky-bottom ecosystem fisheries model. Ecological Modelling 104:145-164. Noro, T., T. Masaki, and H. Akioka. 1983. Sublittoral distribution and reproductive periodicity of crustose coralline algae (Rhodophyta, Cryptonemiales) in southern Hokkaido, Japan. Bulletin of the Faculty of Fisheries, Hokkaido University 34:1-10. Paine, R. T. 1966. Food web complexity and species diversity. The American Naturalist 100:65-75. Paine, R. T. 1969. A note on trophic complexity and community stability. The American Naturalist 103:91-93. Paine, R. T., and R. L. Vadas. 1969. The effects of grazing by sea urchins, Strongylocentrotus spp., on benthic algal populations. Limnology and Oceanography 14:710-719. Paine, R. T., J. C. Castillo, and J. Cancino. 1985. Perturbation and recovery patterns of starfish-dominated intertidal assemblages in Chile, New Zealand, and Washington State. The American Naturalist 125:111-119. Pearse, J. S., M. E. Clark, D. L. Leighton, C. T. Mitchell, and W. J. North. 1970. Kelp habitat improvement project, 1969-70, California Institute of Technology, Appendix, 87 pp. Pearse, J. S., and A. H. Hines. 1979. Expansion of a central California kelp forest following the mass mortality of sea urchins. Marine Biology 51:83-91. Power, M. E. 1992. Top-down and bottom-up forces in food webs: do plants have primacy. Ecology 73:733-746. Reed, D. C , and M. S. Foster. 1984. The effects of canopy shading on algal recruitment and growth in a giant kelp forest. Ecology 65:937-948. Ricklefs, R. E., and G. L. Miller. 2000. Ecology, 4th edition. W.H. Freeman, New York. Scheibling, R. 1986. Increased macroalgal abundance following mass mortalities of sea urchins {Strongylocentrotus droebachiensis) along the Atlantic coast of Nova Scotia. Oecologia 68:186-198. Schmitz, O. J., P. A. Hamback, and A. P. Beckerman. 2000. Trophic cascades in terrestrial systems: a review of the effects of carnivore removals on plants. The American Naturalist 155:141-153.  Schroeter, S.C. 2001. University of California, Santa Barbara. Personal communication regarding interactions between sea urchins and kelp. Schroeter, S. C. 1978. Experimental studies of competition as a factor affecting the distribution and abundance of purple sea urchins, Strongylocentrotus purpuratus (Stimpson). Ph.D. dissertation. University of California, Santa Barbara.  Simenstad, C. A., J. A. Estes, and K. W. Kenyon. 1978. Aleuts, sea otters, and alternate stable-state communities. Science 200:403-411. Simpson, E. H. 1949. Measurement of diversity. Nature 163:688. Steneck, R. S. 1993. Is herbivore loss more damaging to reefs than hurricanes? Case studies from two Caribbean reef systems (1978 - 1988). Pages 220-226 in R. N. Ginsburg, editor. Proceedings of the colloquium on global aspects of coral reefs: health, hazards and history. Rosenstiel School of Marine and Atmospheric Science, Miami. Strong, D. R. 1992. Are trophic cascades all wet? Differentiation and donor-control in speciose ecosystems. Ecology 73:747-754. Sutherland, J. P. 1974. Multiple stable points in natural communities. The American Naturalist 108:859-873. Tegner, M. J., and P. K. Dayton. 1991. Sea urchins, El Ninos, and the long term stability of Southern California kelp forest communities. Marine Ecology Progress Series 77:4963. Tegner, M. J., P. K. Dayton, P. B. Edwards, and K. L. Riser. 1995. Sea urchin cavitation of giant kelp (Macrocystis pyrifera C. Agardh) holdfasts and its effects on kelp mortality across a large California forest. Journal of Experimental Marine Biology and Ecology 191:83-99. Vadas, R. L. 1977. Preferential feeding: an optimization strategy in sea urchins. Ecological Monographs 47:337-371. Vadas, R. L. 1968. The ecology of Agarum and the kelp bed community. Ph.D. dissertation. University of Washington. Valentine, J. F., and K. L. Heck. 1991. The role of sea urchin grazing in regulating subtropical seagrass meadows: evidence from field manipulations in the northern Gulf of Mexico. Journal of Experimental Marine Biology and Ecology 154:215-230.  91  Watanabe, J. M., and C. Harrold. 1991. Destructive grazing by sea urchins Strongylocentrotus spp. in a central California kelp forest: potential roles of recruitment, depth, and predation. Marine Ecology Progress Series 71:125-141. Watson, J. C. 1993. The effects of sea otter {Enhydra lutris) foraging on shallow rocky communities off northwestern Vancouver Island, British Columbia. Ph.D. dissertation. University of California Santa Cruz, Santa Cruz. Wheeler, W.N., L.D. Druehl. 1986. Seasonal growth and productivity of Macrocystis integrifolia in British Columbia Canada. Marine Biology 90:181 -186. Wing, B. L., and K. A. Clendenning. 1971. Kelp surfaces and associated invertebrates. In, The Biology of Giant Kelp Beds (Macrocystis) in California, edited by W.J. North, Beihfte zur Nova Hedwigia, Vol. 32, pp. 319-341.  3. G o n a d a l e n h a n c e m e n t of red s e a u r c h i n s , Strongylocentrotus franciscanus, taken f r o m barrens habitat in British C o l u m b i a ABSTRACT  I collected red sea urchins, Strongylocentrotus franciscanus, in urchin barrens habitat with the intent of enhancing them for market. Urchins were placed in enclosures and provided with the kelp Macrocystis integrifolia for the 83-day period between 20 March 1998 and 12 June 1998. Sea urchins that were fed M. integrifolia had a final gonad index of 8.3, whereas the donor population had initial and final gonad indices of 2.9 and 3.5 respectively. In addition, the proportion of individuals displaying a state of starvation via dark brown gonads was significantly lower after feeding as compared to the donor population at the end of the experiment (2% and 17%, respectively). These results suggest that food limitation is an important factor in the gonadal development of this urchin population. Moreover, the findings suggest that the urchin resource can be expanded to include barren habitats. Urchins taken from barrens can be quickly enhanced before being sent to market.'  93  INTRODUCTION  Foreign demand for sea urchin gonads, called uni on the Japanese market, led to the development of the sea urchin fishery in British Columbia during the 1970's (Farr and Bunnell 1980). The harvest has grown rapidly since 1982 with consistent landings since 1994 (DFO 1998). In 1997, total landings for Clayoquot Sound (my study site) consisted of over 124,000 kg of red sea urchins, Stongylocentrotus franciscanus (J. K. Davidson, Department of Fisheries and Oceans, Canada, personal communication). Sea urchin feeding habits (Leighton 1960, Mann 1977, Mattison et al. 1977, Dean et al. 1984, Leinaas and Christie 1996) and preference for macrophytic algae species (Leighton 1966, 1971, Vadas 1977) can lead to the conversion of many marine rocky bottom habitats from kelp forest to "barren grounds" (Pearse et al. 1970, Farr and Bunnell 1980). The ability of sea urchins to survive on very low nutritional inputs (Ebert 1967, 1968) and the absence of effective urchin predators, such as sea otters, Enhydra lutris, (Estes and Palmisano 1974, Watson 1993, Estes and Duggins 1995), enables the persistence of barren grounds in some areas. Under these conditions of little food, S. franciscanus resorbs gametes, and the gonads become dark brown (Bernard 1977). The red sea urchin has the highest commercial potential of naturally occurring sea urchins in British Columbia (Breen 1980). The potential periods of spawning are between March and September in southern British Columbia (Bernard 1977). The fishery operates between October and March or April during the period when urchins are recovering from spawning but before the gonads become milky with gametes (Breen 1980). Unfortunately, much of the harvesting occurs before the annual kelp Nereocystis luetkeana has recruited. Thus, in years when over-wintering kelp species are scarce, the fisheries value of urchins will  94  be low because of depressed food supply (Breen 1980). Enhancement techniques can resolve many problems associated with poor fisheries quality of wild stocks (Tegner 1989), while providing the harvester with the ability to time sales in order to maximise and maintain stable profits during much of the year. Harvesters ignore many urchin barrens habitats because the urchins present often have low quality roe (poor colour and mass). Many researchers have shown through field and laboratory studies that the growth and reproduction of strongylocentrotid sea urchins is directly related to the quality and quantity of consumed food items (Ebert 1968, Lawrence 1975, Vadas 1977, Larson et al 1980, Thompson 1982, Lemire and Himmelman 1996, Meidel and Scheibling 1999). The primary objective of this experiment was to test the hypothesis that mass and colour of gonads in S.franciscanuswere food limited in this area of Clayoquot Sound. By studying food limitation, I could determine if food supplementation enhances urchins for the uni market. Under the hypothesis of food limitation, providing a supplement of a natural food source should increase the relative gonad size (measured as an index of gonad mass to total mass) and decrease the number of individuals with dark brown gonads (an indication of starvation). To test this hypothesis, urchins from barrens habitat were penned and provided with the kelp Macrocystis integrifolia.  MATERIALS AND METHODS  I collected S.franciscanusat Hot Springs Cove in Clayoquot Sound on the west coast of Vancouver Island, British Columbia (49° 21' N, 126° 16' W). The study area is part of a 0.5 by 2.5-km coastal, aboriginal fishing zone, designated by the Department of Fisheries and Oceans, Canada. That designation prevents commercial harvesting of the resources present,  95  allowing the resident Hesquiat First Nation to conduct a food fishery and study their environment without commercial fishing pressure. I collected 90 S.franciscanus(>80 mm) on 20 March 1998 and separated them into two groups. I sampled gonad mass of one group (n = 45) immediately, and penned the second group (n = 45) with 15 urchins in each of three chicken wire pens measuring 127 cm x 79 cm x 51 cm. Pens were 2-3 m below the ocean surface within Hot Springs Cove. The penned urchins were fed M integrifolia ad libitum for 83 days. The congeneric M. pyrifera has been shown to be a preferred algal food (Leighton 1966) and other studies on food preference in strongylocentrotids have shown that preferred foods increase gonad production (Larson et al. 1980, Lemire and Himmelman 1996). Moreover, M. integrifolia is a perennial canopy forming species making collection for this and future enhancements relatively simple. On 12 June 1998,1 dissected the supplemental feeding group and a second collection (n = 30) of free-ranging urchins (>80 mm). All urchins were collected from the same barren at approximately 8 m. I measured test diameter, wet body mass, wet gonad mass, and recorded gonad colour. Gonad indices were calculated as the percentage of the gonad mass to the total body mass. Since the urchins were not maintained as independent replicates (Hurlbert 1984), ANOVAs were not used to compare treatments. Rather, parameter means and standard errors for test diameter, wet body mass, wet gonad mass, and gonad index are presented. Comparisons are made among the initial donor group (March group), the final donor group (June group) and the experimental group (supplemental feeding). No effort was made to control for cage effects as the applied nature of the experiment makes such comparisons irrelevant. Statistica™ statistical software for Windows ™ version 5.5 (StatSoft Inc., Tulsa, OK, USA) was used to graph the data.  96  RESULTS  S.franciscanusremoved from urchin barrens and penned readily consumed the algae provided. The resulting data (Figure 3-1) show considerable differences in average gonad wet mass and gonad index for the feeding group as compared to either of the field collections. Moreover, there appear to be no biologically meaningful differences in gonad mass and gonad index among the March and June field collections. Urchins from the March, June, and supplemental feeding groups had average gonad masses of 11.0 ± 1.1 g, 12.9 ± 1.2 g, and 34.4 ± 1.5 g respectively. Gonad indices in the same order were 2.9 ± 0.3, 3.5 ± 0.3, and 8.3 ± 0.4. Horizontal test diameter and total wet mass varied little among groups and arguably display no biological differences among the three groups. Plots of gonad index and gonad wet mass versus test diameter illustrate different relationships between gonad mass and body size within the field and fed collections (Figure 3-2). No significant correlations were noted between gonad mass or index and test diameter for either of the field collections. However the supplemental feeding group displayed significant correlation between diameter and gonad index (r = -0.51; p < 0.001) as well as between diameter and gonad mass (r = 0.49; p < 0.01). I fitted all relationships as linear although gonad index has been shown to be non-linear in relation to test size (Gonor 1972). This decision to use linear correlation is due to the narrow size range of urchins sampled for this study. I noted the presence of dark brown gonads among the field collections and the supplemental feeding group. The March and June field samples had a high percentage of individuals with very dark brown gonads (March sample = 47%, June sample = 17%)  97  compared to the supplemental feeding group (2%) (Chi-squared = 25.29, p < 0.001). The importance of gonad colour is discussed below.  98  15 14 13 12 11 10 9 8 7 6 5| 4 1 0I  40 ,  S  32 24  f•a  16  a c  o O  March  June  March  Fed  June  10  Fed  2%  8 6 4  47%  17%  o  C3  March  June  Fed  2  March  June  Fed  Figure 3-1. Results of feeding experiment compared tofieldpopulations before (March; n = 45) and after (June; n = 30) the supplemental feeding (n = 43). Data are presented as means ± 1 SE for horizontal test diameter, total wet mass, gonad wet mass, and gonad index. Values above error bars in the gonad index plot are the proportions of individuals with dark brown gonads in each sample.  .s CU B  •J  "3 CU  CO  <M  "  I=  1 1  •  1  i  WD  1  B  i  1 1  •  ! 1  i i  1  i  1  1 CN  8 d  CO  >S  1  i  m  a)  O  • .< i ii»  •  o o  c cu E  1-9  I-  J3  cS  t.  w  •.L  cu  cu o  £  s  . cs  E i  b 1 1  CD  6) sse|/\| peuoo  CO  3 -a  •= 2 cu cu  t i  CO  i  •1 J •  11 II  ei  •g ™ —  E  3 c<o E a>  CL CL 13  >• 1 •  1  S £ -w £  IS o  g ^ CS  O  <S _<U 'w'  x  CM  g  8  W> ° ^  jr  CO  5  V3  S g  i i i i  i  ,  Q *-—<  .S.  T  r  •s °  •B  1 1  g  cu cs  -M  '§£  E ra  1  J2  0)  (A Q)  • ••  3  -M  rla  b  B  s2  i  f  E ro  JB U 03  « t  •  1.  t  g. B  ti—i CS O „  |  00"  cu "H. m  co cu _  f r. 1  g "  u  E  b  W> B  CO  xepu| peuog  cu  •o a  £  co B  O  2  s o" «  cs O  O  cu II  «  k. u 6JD  CU  . <u a . CO V _ cu  fNl  r> S  u >IM  3  feJD cu ©  E •£  6H  100  DISCUSSION I conducted the feeding trial using S.franciscanuscollected from severely grazed urchin barrens habitat. Gonad indices for the March and June collections were 2.9 + 1.9 (n = 45) and 3.5 ± 1.8 (n = 30)'respectively. The lack of a biologically meaningful difference between the two time periods is unusual because S.franciscanuswould normally be increasing in gonad index during this time, with a peak achieved in May (Bernard 1977). However, urchins are known to exhibit plastic responses in the allocation of resources to body components (Ebert 1980, Thompson 1982, Edwards and Ebert 1991, Levitan 1991). The lower percentage of individuals displaying a state of starvation in the June sample may mean that urchins are reallocating resources to prepare for spawning. The relatively low values for gonad indices (compare to McBride et al 1997) of thefree-rangingS. franciscanus reflects the poor quality of habitat at this site. The much increased gonad index for the supplemental feeding group, 8.3 + 2.3, suggests food availability was limiting. There are two interesting trends in the correlation data of Figure 3-2. First, the urchins sampledfromthe field (March and June) display no relationships between test diameter and gonad mass or index, while urchins in the supplemental feeding group do. These findings suggest that in starved urchins above 8.0 cm, gonad mass is relatively constant and probably at a minimum value that is independent of size. Conversely, urchins in the fed group show a positive correlation between size and gonad mass. Second, the trends between size and gonad index and between size and gonad mass are opposite. The positive correlation between size and gonad mass indicates that larger urchins increased their gonad mass significantly more than smaller urchins did. However, the negative association between size and gonad index indicates that as a proportion to total mass, larger urchins are  increasing their gonad mass less than smaller urchins. Given this diminishing return in gonad index, along with a market preference for smaller urchins (Sloan 1986, RogersBennett et al. 1995) and the importance of larger urchins for the recruitment of juveniles (Tegner and Dayton 1977, Breen et al. 1985, Rogers-Bennett et al. 1995), managers should consider setting upper size limits for any urchin harvesting or enhancement program. Understanding the relationship between food supply and gonad index is critical to the development of sea urchin aquaculture and harvesting programs (Minor and Scheibling 1997). Recent work evaluating aquaculture techniques for sea urchins (Klinger et al. 1986, Lawrence et al. 1997, McBride et al. 1997) has relied on the use of artificial feeds to improve the quality of fishery urchins. Klinger et al. (1986) and Lawrence et al. (1997) both reported enhancement after offering artificial feeds to S. droebachiensis and Loxechinus albus. McBride et al. (1997) showed no significant difference between S.franciscanusfed Nereocystis luetkeana versus an artificial diet. McBride et al. (1997) observed greater gonad indices for urchins fed N. luetkeana (17.3 ± 4.7, SD), than I obtained for urchins fed M. integrifolia (8.3 + 2.3, SD). Two factors may explain this difference. I fed urchins for a total of 83 days as opposed to 120 days reported in McBride et al. (1997). In addition, selection of urchinsfrombarrens habitat means that I used urchins of lower initial body condition than the urchins used by McBride et al. (1997). This lower body condition would result in more time being required to increase gonad mass. By supplementing the food available to &franciscanus,gonad mass and indices were increased above those found in field-collected specimens. Moreover, the number of individuals displaying a state of starvation through the presence of dark brown gonads (Bernard 1977) was reduced in the supplemental feeding group. This study demonstrates that  102  free ranging urchins of low commercial fisheries value can be quickly enhanced for market. With an expansion of the methods employed here, the enhancement of urchins from barrens habitat may be viewed as an economical alternative to pure aquaculture techniques (rearing from juvenile seed) (Tegner 1989). I suggest that urchins from barren habitats can be enhanced and sold to market, thereby placing less harvesting pressure on other areas. I suggest that before artificial feeds are accepted for use on this species, the cost / benefit ratio of readily available natural kelp be more closely examined.  REFERENCES Bernard, F. R. 1977. Fishery and reproductive cycle of the red sea urchin, Strongylocentrotusfranciscanus,in British Columbia. Journal of Fisheries Resources Board of Canada 34:604-610. Breen, P. A. 1980. The ecology of red sea urchins in British Columbia. Pages 3-12 in, International symposium on coastal Pacific marine life. Western Washington University, Bellingham, WA. Breen, P. A., W. Carolsfeld, and K. L. Yamanaka. 1985. Social behaviour of juvenile red sea urchins, Strongylocentrotusfranciscanus(Agassiz). Journal of Experimental Marine Biology and Ecology 92:45-61. Davidson, J.K. 1999. Catch Statistics Unit, Pacific Region. Department of Fisheries and Oceans Canada. Personal communication regarding sea urchin catch statistics. Dean, T. A., S. C. Schroeter, and J. D. Dixon. 1984. Effects of grazing by two species of sea urchins (Strongylocentrotusfranciscanusand Lytechinus anamesus) on recruitment • and survival of two species of kelp (Macrocystis pyrifera and Pterygophora californica). Marine Biology 78:301-313. [DFO] Department of Fisheries and Oceans Canada 1998. Pacific region 1998/99 red sea urchin management plan. Fisheries and Oceans Canada, 30pp. Ebert, T. A. 1967. Negative growth and longevity in the purple sea urchin Strongylocentrotus purpuratus (Stimpson). Science 157:557-558. Ebert, T. A. 1968. Growth rates of the sea urchin Strongylocentrotus purpuratus related to food availability and spine abrasion. Ecology 49:1075-1091.  103  Ebert, T. A. 1980. Relative growth of sea urchin jaws: an example of plastic resource allocation. Bulletin of Marine Science 30:467-474. Edwards, P. B., and T. A. Ebert. 1991. Plastic responses to limited food availability and spine damage in the sea urchin Strongylocentrotus purpuratus (Stimpson). Journal of Experimental Marine Biology and Ecology 145:205-220. Estes, J. A., and J. F. Palmisano. 1974. Sea otters: their role in structuring nearshore communities. Science 185:1058-1060. Estes, J. A., and D. O. Duggins. 1995. Sea otters and kelp forest in Alaska: generality and variation in a community ecological paradigm. Ecological Monographs 65:75-100. Farr, A.C.M., and F.L. Bunnell. 1980. The sea otter in British Columbia - a problem or opportunity? Pp. 110-128 in, R. Stace-Smith, L. Johns, and P. Joslin, eds. Threatened and endangered species and habitats in British Columbia and the Yukon. B.C. Ministry of Environment, B.C. Fish and Wildlife Branch, Victoria, British Columbia. Gonor, J. J. 1972. Gonad growth in the sea urchin, Strongylocentrotus purpuratus (Stimpson) (Echinodermata: Echinoidea) and the assumptions of gonad index methods. Journal of Experimental Marine Biology and Ecology 10:89-103. Hurlbert, S. H. 1984. Pseudoreplication and the design of ecological field experiments. Ecological Monographs 54:187-211. Klinger, T. S., H. L. Hsieh, R. A. Pangallo, C. P. Chen, and J. M. Lawrence. 1986. The effect of temperature on feeding, digestion, and absorption of Lytechinus variegatus (Lamarck) (Echinodermata: Echinoidea). Physiological Zoology 59:332-336. Larson, B. R., R. L. Vadas, and M. Keser. 1980. Feeding and nutritional ecology of the sea urchin Strongylocentrotus drobachiensis in Maine, USA. Marine Biology 59:49-62. Lawrence, J. M. 1975. On the relationships between marine plants and sea urchins. Oceanography and Marine Biology 13:213-286. Lawrence, J. M., S. Olave, R. Otaiza, A. L. Lawrence, and E. Bustos. 1997. Enhancement of gonad production in the sea urchin Loxechinus albus in Chile fed extruded feeds. Journal of the World Aqauculture Society 28:91-96. Leighton, D. L. 1960. Studies of kelp-grazing organisms. In, Kelp investigations program, quarterly report. University of California, Institute of Marine Research, IMR Ref. 607, J3-22. Leighton, D. L. 1966. Studies of food preference in algivorous invertebrates of Southern California kelp beds. Pacific Science 20:104-113.  104  Leighton, D. L. 1971. Grazing activities of benthic invertebrates in southern California. In, The biology of giant kelp beds (Macrocystis) in California, edited by W.J. North, Beihfte. zur Nova Hedwigia, Vol. 32, pp. 421-453. Leinaas, H. P., and H. Christie. 1996. Effects of removing sea urchins (Strongylocentrotus droebachiensis): stability of the barren state and succession of kelp forest recovery in the east Atlantic. Oecologia 105:524-536. Lemire, M., and J. H. Himmelman. 1996. Relation of food preference to fitness for the green sea urchin, Strongylocentrotus droebachiensis. Marine Biology 127:73-78. Levitan, D. R. 1991. Skeletal changes in the test and jaws of the sea urchin Diadema antillarum in response to food limitation. Marine Biology 111 :431-436. Mann, K. H. 1977. Destruction of kelp-beds by sea-urchins: a cyclical phenomenon or irreversible degradation? Helgolander Wissenschaftliche Meeresuntersuchungen 30:455-467. Mattison, J. E., J. D. Trent, A. L. Shanks, T. B. Akin, and J. S. Pearse. 1977. Movement and feeding activity of red sea urchins (Strongylocentrotus franciscanus) adjacent to a kelp forest. Marine Biology 39:25-30. McBride, S. C , W. D. Pinnix, J. M. Lawrence, A. L. Lawrence, and T. M. Mulligan. 1997. The effect of temperature on production of gonads by the sea urchin Strongylocentrotus franciscanus fed natural and prepared diets. Journal of the World Aquaculture Society 24:357-365. Meidel, S. K., and R. E. Scheibling. 1999. Effects of food type and ration on reproductive maturation and growth of the sea urchin Strongylocentrotus droebachiensis. Marine Biology 134:155-166. Minor, M. A., and R. E. Scheibling. 1997. Effects of food ration and feeding regime on growth and reproduction of the sea urchin Strongylocentrotus droebachiensis. Marine Biology 129:159-167. Pearse, J. S., M. E. Clark, D. L. Leighton, C. T. Mitchell, and W. J. North. 1970. Kelp habitat improvement project, 1969-70, California Institute of Technology. Appendix, 87 pp. Rogers-Bennett, L., W. A. Bennett, H. C. Fastenau, and C. M. Dewees. 1995. Spatial variation in red sea urchin reproduction and morphology: implications for harvest refugia. Ecological Applications 5:1171-1180.  105  Sloan, N. A. 1986. World jellyfish and tunicate fisheries, and the northeast Pacific echinoderm fishery. Canadian Special Publication of Fisheries and Aquatic Sciences 92:23-33. Tegner, M. J., and P. K. Dayton. 1977. Sea urchin recruitment patterns and implications of commercial fishing. Science 196:324-326. Tegner, M. J. 1989. The feasibility of enhancing red sea urchin, Strongylocentrotus franciscanus, stocks in California: an analysis of the options. Marine Fisheries Review 51:1-22. Thompson, R. J. 1982. The relationship between food ration and reproductive effort in the green sea urchin, Strongylocentrotus droebachiensis. Oecologia 56:50-57. Vadas, R. L. 1977. Preferential feeding: an optimization strategy in sea urchins. Ecological Monographs 47:337-371. Watson, J. C. 1993. The effects of sea otter (Enhydra lutris) foraging on shallow rocky communities off northwestern Vancouver Island, British Columbia. University of California Santa Cruz, Santa Cruz. Ph.D. dissertation.  4. U r c h i n s a n d U r c h i n s : Effects of R e m o v i n g U r c h i n s o n the Quality of R e m a i n i n g U r c h i n s  ABSTRACT The shallow subtidal regions near Hot Springs Cove, Vancouver Island, British Columbia (49° 21' N, 126° 16' W) are generally characterised by large rocky areas dominated by the red sea urchin, Strongylocentrotus franciscanus. S. franciscanus were removed at three sites with four levels of urchin removal per site. Manipulations of urchin density were maintained throughout the experiment and monitored seasonally for two seasons pre-treatment and seven seasons post-treatment. The manipulations allowed testing of the effects of competition-mediated food limitation in this population of S.franciscanus. The removal of S. franciscanus at all levels resulted in an increase in fecundity (measured as gonad index) for the sea urchins that remained, as well as for the urchins that reinvaded the total removal plots. All study plots showed an increase in gonad index over time, but the increase in gonad index was statistically greater for urchins in removal plots versus the control plots. The abundance of S. franciscanus could be responsible for both the absence of kelp forest habitat in the region (Chapter 2) and the depressed fecundity of urchins in this population. The effects of urchin removal on these communities and the implications for sea urchin management are discussed with regards to the resulting fecundity of S. franciscanus.  INTRODUCTION Foreign demand for sea urchin gonads, called uni on the Japanese market, led to the development of the sea urchin fishery in British Columbia during the 1970's (Farr and Bunnell 1980, Muse 1998). The harvest has grown rapidly since 1982 with consistent  landings since 1994 (DFO 1999). In 1997, total landings from Clayoquot Sound (my study site) consisted of over 124,000 kg of red sea urchins, Stongylocentrotus franciscanus (J. K. Davidson, Department of Fisheries and Oceans, Canada, personal communication). Sea urchin feeding habits (Leighton 1960) and preference for macrophytic algae species (Leighton 1966, 1971, Vadas 1977) can lead to the conversion of many marine rocky bottom habitats from kelp forest to "barren-grounds" (Pearse et al. 1970, Farr and Bunnell 1980). The barren grounds persist due to the loss of a keystone predator (sea otters)fromsea otter hunting in the late 1800s (Estes and Palmisano 1974, Watson 1993). The removal of sea otters meant the removal of an historically important community-structuring agent (see Chapter 1). The ability of sea urchins to survive on very low nutritional inputs (Ebert 1967, 1968) facilitates the persistence of barren grounds because urchins can survive at densities that limit the recruitment of kelp. Under these conditions of little food, S. franciscanus resorbs gametes, and the gonads become dark brown and of low fisheries value (Bernard 1977). Nuu-Chah-Nulth traditional ecological knowledge teaches that urchins were only collected from specific locations where seaweed was abundant (Clayoquot Sound Scientific Panel 1995). As well, harvesters ignore many urchin barren habitats because the urchins present often have low quality roe (poor colour and weight). In Chapter 3, from feeding trial data, I concluded that weight and colour of gonads from urchins in this population were food limited. The objective of this chapter is to evaluate the conclusion from Chapter 3, that weight and colour of gonads in S.franciscanusare food limited using in situ data from the traditional territory of the Hesquiat First Nation, and to test if food limitation is the result of intraspecific competition. Under the hypothesis of intraspecific competition, a reduction in  the population size of S. franciscanus should result in an increase in fecundity (measured as an index of gonad weight to total weight) for any remaining S. franciscanus. To test this hypothesis, I manipulated the density of urchins in the field and sampled the gonad quality of the remaining urchins by dissecting and collecting data on test diameter, wet weight, gonad weight, and gonad colour.  METHODS AND MATERIALS Field To test for competition-induced food-limitation in the field, I used a repeated measures, randomised block design. Sea urchins were manipulated to the four treatment levels within three study sites as described in Chapter 2. While sampling transects within study plots as discussed for urchin and kelp sampling (see chapter 2), nine urchinsfromeach study plot were selected to monitor gonad indices of urchins in each experimental unit throughout the study. Three points were randomly chosen on each transect line and the nearest urchin to the chosen point was collected and brought to the surface for dissection. In the case of total removal plots, urchins often had to be collected much further from the chosen point than desired. This often meant collecting from the periphery of the plot. If there was not an adequate supply of urchins within the area represented by a given transect line, then I collected fewer urchins leading to an unbalanced experimental design.  Laboratory Collected urchins were brought to a makeshift laboratory and dissected between 2 to 8 hours after collection. Attempts were made to keep urchins cool and shaded until dissection. If urchins were kept outdoors for extended periods, or could not be dissected right away, then they were kept in mesh bags and submerged in seawater until dissection. I  measured test diameter to the nearest millimetre by taking three or more separate measurements across the urchin with vernier callipers. I measured the wet weight of the urchin with an electronic balance to the nearest 0.1 g. Finally, the test was split, gonads extracted and wet weight measured. Gonad indices were calculated as the percentage of the gonad weight to the total body weight (Gonor 1972). Any urchins displaying a state of starvation via dark brown gonads were noted.  Analysis The two intermediate levels of urchin removal (low and medium) were combined into a single treatment (intermediate) for the reasons discussed in Chapter 2. To determine site and treatment differences with the inclusion of temporal variation, the data were analysed using repeated measures ANOVA. Closer examination of treatment effects was investigated with a two-way ANOVA comparing differences in gonad indices at the start and conclusion of the experiment. In both analyses, treatment group (Total, Intermediate and Control) was the independent variable and gonad index (arcsin transformed) was the dependent variable. Analyses were blocked by spatial replicates of the experiment (sites) and in the case of the repeated measures analyses, time was specified as the repeated measure. Test diameters were compared for all urchins dissected during the study with the same repeated measures approach used for gonad index. Treatments were considered a fixed effect and all other factors treated as random effects. All data met the assumptions of normality and homoscedasticity. Post-hoc Fisher-LSD multiple range tests were used to distinguish among statistically different means for the two-way ANOVAs. The data were analysed and graphed using Statistica™ software for Windows™. An alpha probability of 0.05 was chosen for all statistical tests.  RESULTS S. franciscanus collected from the study sites showed improvement in condition of the gonads after the implementation of the treatments. Two-way repeated measures ANOVA comparing levels of urchin removal resulted in significant differences in gonad index over time (F(g48) = 17.46, p < 0.001: Table 4-1). The repeated measures technique did not ;  distinguish among the three levels of urchin removal (F(2,2) = 15.41, p = 0.06: Table 4-1) but was likely confounded by temporal cycles in urchin gonad index as well as the inclusion of pre-treatment and post-treatment data points (e.g., Figure 4-1). Pre- and post-treatment data were included in the analysis because they highlight the effect of urchin removal by observing the temporal component of the ANOVA model. To more closely examine the treatment effects, two-way ANOVAs comparing changes in gonad index between the first and second year, and the first and final year were used. Both analyses found significant differences in mean increase in gonad index for the three treatment groups. Increases in gonad index were compared for the periods of summer 1997 through summer 1998 (F(2,4)  =  8.94, p = 0.033: Table 4-2) and for summer 1997 through summer 1999 (F , ) = 13.58, p = (2 4  0.016: Table 4-3). Post hoc Fisher LSD comparisons detected differences between the control plots compared with intermediate and total urchin removals for both of the analysed periods (Table 4-4). There were no observed differences between intermediate and total removals. The significant statistical differences between control plots and both removal levels for the two way ANOVAs (Table 4-2 & Table 4-3) and the consistently lower gonad indices for control plots (Figure 4-1) suggest that the significant temporal variation observed in the repeated measures ANOVA (Table 4-1) is the result of treatment implementation. Had the indices not been different among treatments, the data could be interpreted as having been influenced by factors unrelated to the manipulations.  Gonad index does not change in a linear fashion with regards to urchin size (Gonor 1972). Gonad index in small urchins increases at higher rates per unit size than larger urchins; larger urchins maintain a relatively stable gonad mass with increases in size (Gonor 1972). Thus, comparisons of gonad indices across treatments or areas need to consider the potential effects of different sized individuals in the populations being compared. Test diameters of urchins in this study did not vary between treatments (F(2,3> = 0.71, p = 0.61: Table 4-5; and Figure 4-2). Differences were noted among sites, and there was a significant site by treatment interaction (F(448) = 4.59, p< 0.001: Table 4-5). The interaction was due to ;  a slight decrease in test diameters for urchins in total removal and control plots in Sites 2 and 3 as compared to Site 1, while sizes of urchins in the intermediate removals were stable across all three sites (Figure 4-3). Although there is variability across sites and some interaction, the differences among treatments are slight and likely have no effect on interpretation of the gonad index data. Moreover, given that the mean sizes are all within larger adult size classes, gonad indices are likely stable over even greater test size differences than actually observed in this experiment (Gonor 1972). I also noted differences in gonad colour for all treatments grouped as pre-treatment and post-treatment. Dark brown gonads are an indication of starvation among S. franciscanus (Bernard 1977). In pre-treatment sampled groups of urchins, 13.4% of individuals displayed a state of starvation compared with only 2.6% for the post-treatment urchins (Yates corrected Chi-squared = 34.44, p < 0.001: Table 4-6).  Table 4-1. Split-plot analysis of variance of urchin gonad indices (acrsin transformed). Treatment effects tested using the pseudo f-test. Error d.f. for treatment effects is 2.30 (refer to Hicks 1982). Effect  D.F.  MS  Error MS  F  p  Block (Site)  2  0.0310  0.0036  8.51  0.00  Treatment  2  0.0698  0.0046  15.14  0.06  Error 1 (Block*Treatment)  4  0.0059  0.0036  1.61  0.19  Time  8  0.0637  0.0036  17.46  0.00  Time*Treatment  16  0.0024  0.0036  0.65  0.82  Error 2  48  0.0036  —  —  —  Sampling Error  423  0.0010  —  —  —  o  113  Gonad Index Over Time  - o - Total •••©•• Intermediate Sum 97 Fall 97 Win 98 Spr 98 Sum 98 Fall 98 Win 99 Spr 99 Sum 99  Control  TIME  Figure 4-1. Gonad index of S. franciscanus as a function of time for each of the three treatments averaged across study sites. Bars are ± one standard error of the mean (sample size varies between 5 and 9 for Total and Control, and between 12 and 18 for Intermediate). Vertical bar represents the time at which urchin removals were initiated (December 1997). Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3).  Table 4-2. Two-way ANOVA for the mean increase in gonad index (arcsin transformed) between summer 1997 and summer 1998 for the three treatment levels.  Effect  D.F.  MS  MS error  F  P  Site  2  0.00015  0.00024  063  058  Treatment  2  0.00217  0.00024  8.94  0.03  Error (Site * Treatment)  4  0.00024  Table 4-3. Two-way ANOVA for the mean increase in gonad index (arcsin transformed) between summer 1997 and summer 1999 for the three treatment levels. Effect  D.F.  MS  MS error  F  P  Site  2  0.00021  0.00019  Til  041  Treatment  2  0.00258  0.00019  13.58  0.02  Error (Site * Treatment)  4  0.00019  Table 4-4. Fisher LSD tests for treatment differences observed in the ANOVA results of Table 4-2 and Table 4-3. Mean increase per treatment ± standard error of the mean. Lower case letters denote means that are the same. Upper means are increases between summer 1997 and summer 1998 (t _i) and the lower means are increases between summer 1997 and summer 1999 (t _i). 2  3  Treatment (removal)  Total  Intermediate  Control  M e a n (t -i)  8.32 ± 0.85a  6.62 ± 0.79a  3.10 ± 0.57b  M e a n (t .i)  9.20 ± 0.74a  10.59 ± 0.51a  5.54 ± 1.03b  2  3  r  Table 4-5. Split-plot analysis of variance for dissected urchin test diameters. Treatment effects tested using the pseudo f-test. Error d.f. for treatment effects is 3.41 (refer to Hicks 1982). D.F.  MS  Error MS  F  P  Block (Site)  2  71.80  4.30  16.70  0.00  Treatment  2  13.05  18.28  0.71  0.61  Error 1 (Block* Treatment)  4  19.72  4.30  4.59  0.00  Time  8  17.69  4,30  4.11  0.00  Time*Treatment  16  2.86  4.30  0.67  0.81  Error 2  48  4.30  —  —  —  Sampling Error  496  2.02  —  —  —  Effect  Figure 4-2. Average urchin test diameters for dissected urchins from each treatment throughout the duration of the experiment. Bars are ± one standard error of the mean (sample size varies between 5 and 9 for Total and Control, and between 12 and 18 for Intermediate). Treatment labels 'total', 'intermediate' and 'control' refer to the amount of urchin removal as total, intermediate or no removal, respectively. A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3).  Test Diameter Site by Treatment Interaction  Figure 4-3. Site by treatment interaction observed for test diameters of dissected urchins. Bars are ± one standard error of the mean (sample size varies between 66 and 152). Treatments were total urchin removal (Total), intermediate levels of urchin removal (Intermediate) and no removal of urchins (Control).  Table 4-6. Contingency table analysis of observations of urchins displaying a starved condition as noted by the presence of dark brown gonads for all dissected urchins before and after implementation of treatments. Yates corrected Chi-square = 34.44, p < 0.001.  Before  After  Total  Starving  27  17  44  Feeding  175  636  811  Total  202  653  855  121  DISCUSSION This experiment was conducted to help guide management of a potential sea urchin fishery. By showing that this population is exhibiting density-dependent competition for limited food resources, I can support management practices to alleviate competition and thereby improve the fisheries quality of field-collected urchins. I began the experiment by selecting study sites within severely grazed urchin-barrens habitat. However, this is not an automatic indication that the urchins present would be suffering from density-dependent competition for a limited food resource. The presence of food (algae) and food limitation can be decoupled in this scenario as urchins can subsist on imports of pieces of drift algae from the intertidal zone (North 1964, Dayton et al 1992). However, early observations suggested algal imports were insufficient to prevent food limitation in this population. I observed little drift kelp in the pre-treatment sampling periods and a relatively large proportion of urchins displayed a state of starvation (Table 4-6) through the presence of depleted dark-brown gonads (Bernard 1977). Under the hypothesis of intraspecific exploitation competition for limited food resources, a reduction in the density of sea urchins should result in an increase in fecundity for any remaining individuals. The results of this experiment support the hypothesis. Increases in gonad index (a surrogate for fecundity) after implementation of the treatments was likely the result of increased kelp growth (Chapter 2). The kelp response to urchin removal took approximately two sampling periods (six months) after initiation of removals. The strongest increase in gonad index was between six and nine months after initiation of the urchin removals. The appearance of kelp sporophytes six months after urchin removals began was likely responsible for the increase in gonad index and indicates that this  122  population of urchins is food limited. Intraspecific competition was driving the food limitation in this population because increases in both kelp growth and gonad index resulted from a decrease in urchin density. Alternately, interference competition could produce the same results. If sea urchins actively prevent one another from obtaining food, or force each other into poor quality habitats where food is limited, then interference competition may be important in explaining these results. Schroeter (1978) induced 'spine fencing' in laboratory experiments with S. franciscanus. By starving urchins, then limiting food availability, he found that urchins would use their spines to interfere with each other's ability to gain access to the food resource. However, in the field Schroeter (1978) found that increasing red sea urchin abundance in cages did not affect the growth or gonad index of individuals. Moreover, Duggins (1981) found interspecific facilitation among sea urchins. Urchin grouping may actually improve the ability of the group to gather food particularly in wave exposed environments where capture of large pieces of drift algae may be difficult for single urchins (Duggins 1981). If urchin gonad index is depressed by density dependent food limitation, then why should urchins group? The facilitation of food gathering, protectionfrompredators, improved fertilisation success, and enhanced juvenile survival under spine canopies (Low 1975, Tegner and Dayton 1977, Duggins 1981, Levitan et al. 1992) all promote grouping in sea urchins. While the results of this experiment suggest that individual fecundity may be depressed when densities are high, the benefits of grouping may outweigh this cost. In fact, the improved fertilisation success alone (Levitan et al. 1992) may compensate for lower production of gametes under food-limiting conditions.  The issue then reduces to one of scale. If urchins are in high densities over large areas, then the population may become food limited resulting in lowered fecundity. If urchins exist in only localised groupings interspersed with kelp forests, then urchin groups will likely not be food limited and will also benefit from increases in fertilisation success, food capture, and predator protection. In this study, it appeared that urchin densities were high and consistent across large areas (as opposed to clumped) before manipulation. Therefore, individual urchins were suffering a density-dependent drop in fecundity. Although the population may not suffer as a result, the implications for a sea urchin fishery are clear. High urchin densities over large areas will result in poor quality roe yield for fishers. There has been much recent work in applying aquaculture techniques to sea urchins (Lawrence et al. 1997, McBride etal. 1997, McBride et al. 1998). McBride era/. (1997) varied food type and temperatures, and did not find significant differences in gonad index but did find variable assimilation efficiency among food types. Because assimilation could be improved without increased gonad index, their findings suggest that the measured indices are maximal for S.franciscanus.McBride et al. (1997) reported gonad indices ranging from 17.3 to 20.1 for experiments with S.franciscanus.The maximal gonad index observed in this study was 16.78 ± 0.85 (mean ± 1 standard error) for the total removal treatments during the fall of 1998. Thus, within nine months of experimental removal, this experiment produced gonad indices that approach the maximal possible for the species. The results of this experiment suggest that removal of urchinsfrombarrens may be a quick, cost effective and environmentally sensitive means of increasing the supply of good quality sea urchins. Within an hour, two divers can readily clear 800 m of sea floor of  124  urchins with a density of 5 urchins / m by crushing (Leighton et al. 1966, personal 2  observations). Alternately, the same population can be removed within 4 hours (personal observation) and sold or penned for enhancement. With little other cost or time investment, the perimeter of the removal area can be revisited in 9 months for urchin harvest. Subsequent revisits would depend on the supply of legal size urchins in surrounding areas and the speed with which urchins migrate towards, and feed in, the created kelp plot. This method eliminates the need for traditional aquaculture expenses related to labour and feed for maintenance of captive urchins and has the added bonus of concentrating quality urchins within a small area. Finally, in the process of field enhancement of urchins, kelp forest is created that provides habitat for a variety of species (Leighton 1971, see also Chapter 5). Interpretation of the results of this study is limited by the ability to control the /  movements of individual urchins and to control for proximity effects of various treatments on one another. For instance, gonad indices I obtained for total removal plots relied on data gathered from urchins which had either reinvaded the study plot or were just on the border of the plot. Such a scenario denies the ability to make corrections based on how long an individual urchin had access to the plot resources. The high gonad growth observed in total removal plots shows the influence of competition and the ability of urchins to recover quickly in the absence of competition. Furthermore, the increases in gonad index for control plots points to one of two conclusions. Either the various treatment plots were affecting each other through inputs of algae over considerable distances, or some other external factor was driving an overall improvement in gonad index during the study. Data obtained for the feeding trial (see Chapter 3) suggests the former hypothesis is correct.  125 I  Treatment plots were separated by a distance of at least 30 m centre to centre. The increasing gonad masses in control plots were probably the result of drift kelp that found its way from urchin removal to control plots. This possibility would explain the inability of the repeated measures ANOVA (Table 4-1) to detect significant treatment effects. However, while this creates a statistical problem, it is a biologically important observation. The creation of even very small and disconnected islands of kelp forests can have far reaching effects for the enhancement of urchins suffering from density-dependent intraspecific competition. The results of the feeding experiment (Chapter 3) supports the idea of localised enhancement in areas as much as 45 m away from kelp forests. Urchins collected for the feeding trial in March and June, 1998 were likely too distant to be affected by any drift kelp made-available by the study. If these urchins were benefited by factors beyond the manipulations in this study, then it would be expected that gonad indices would be similar to the study control plots. The urchins collected for the feeding trial displayed gonad indices below those observed for urchins in control plots during the same time periods. Moreover, the gonad indices for field-collected urchins in the feeding trial were similar to the values observed within the study before manipulation.  CONCLUSIONS AND IMPLICATIONS The results of this experiment support the hypothesis that density-dependent exploitation competition for kelp resources is limiting the fecundity of sea urchins in the study area. Thinning of urchin density by humans results in a quick conversion of barrens to kelp forest (see Chapter 2). The recruitment of kelp provides a valuable and preferred food source for remaining urchins (Leighton 1966, Vadas 1977, Lemire and Himmelman 1996). The provision of food in a non-limiting fashion along with the tendency of S. franciscanus to  form dense feeding groups (Leighton et al. 1966, Low 1975, personal observations) provides managers with an important tool for managing this fishery species. Urchin thinning creates a plot of kelp surrounded by high densities of urchins (up to 12 urchins / m observed) with exceptional fisheries value. At the same time, urchins removed to create such feeding plots can be enhanced using the techniques described in Chapter 3, or with more advanced aquaculture techniques (de Jong-Westman et al. 1995, Lawrence et al. 1997, McBride et al 1998). To implement any form of the recommended management above, the Department of Fisheries and Oceans, Canada (DFO) would have to abandon its current form of management. DFO has many management practices relating to sea urchins, however the two important regulations are a minimum size restriction and the use of quotas. With few exceptions, little is done to segregate harvesters by area. This creates competition among harvesters to simply pick as many of the best urchins as possible. To successfully manage sea urchins, based on the findings of this research, harvests would have to be area constrained with communities and harvesters forming cooperatives for management. Within management areas, cooperatives could draft management plans with rotational schedules and allotments to harvesters based on ecological theory, desired community structure and possible enhancement for other fisheries values (see Chapter 5 for the latter). Without such area-constrained cooperative control, any licensed harvester could cross management zones and reap the benefits of others' work to enhance local urchin stocks (see Chapter 6 for expanded management implications).  REFERENCES Bernard, F. R. 1977. Fishery and reproductive cycle of the red sea urchin, Strongylocentrotusfranciscanus,in British Columbia. Journal of Fisheries Resources Board of Canada 34:604-610. Clayoquot Sound Scientific Panel. 1995. First nations' perspectives relating to forest practices standards in Clayoquot Sound. Appendices V and VI. March, 1995. Victoria, B.C. Davidson, J.K. 1999. Catch Statistics Unit, Pacific Region. Department of Fisheries and Oceans Canada. Personal communication regarding sea urchin catch statistics. Dayton, P. K., M. J. Tegner, P. E. Parnell, and P. B. Edwards. 1992. Temporal and spatial patterns of disturbance and recovery in a kelp forest community. Ecological Monographs 62:421-445. de Jong-Westman, M., B. E. March, and T. H. Carefoot. 1995. The effect of different nutrient formulations in artificial diets on gonad growth in the sea urchin Strongylocentrotus droebachiensis. Canadian Journal of Zoology 73:1495-1502. [DFO] Department of Fisheries and Oceans Canada. Pacific Region. 1999. Pacific region 1999/2000 management plan, red sea urchin. Vancouver, B.C. 30p; Appendices. Duggins, D. O. 1981. Interspecific facilitation in a guild of benthic marine herbivores. Oecologia 48:157-163. Ebert, T. A. 1967. Negative growth and longevity in the Purple Sea Urchin Strongylocentrotus purpuratus (Stimpson). Science 157:557-558. Ebert, T. A. 1968. Growth rates of the sea urchin Strongylocentrotus purpuratus related to food availability and spine abrasion. Ecology 49:1075-1091. Estes, J. A., and J. F. Palmisano. 1974. Sea otters: their role in structuring nearshore communities. Science 185:1058-1060. Farr, A. C. M., and F. L. Bunnell. 1980. The sea otter in British Columbia - a problem or opportunity? Pages 110-128 in R. Stace-Smith, L. Johns and P. Joslin, editors. Threatened and endangered species and habitats in British Columbia and the Yukon. B.C. Ministry of Environment, B.C. Fish and Wildlife Branch, Victoria. Gonor, J. J. 1972. Gonad growth in the sea urchin, Strongylocentrotus purpuratus (Stimpson) (Echinodermata: Echinoidea) and the assumptions of gonad index methods. Journal of Experimental Marine Biology and Ecology 10:89-103. Hicks, C. R. 1983. Fundamental concepts in the design of experiments, 3rd edition. Holt, Rinehart, and Winston, New York. 128  Lawrence, J. M., S. Olave, R. Otaiza, A. L. Lawrence, and E. Bustos. 1997. Enhancement of gonad production in the sea urchin Loxechinus albus in Chile fed extruded feeds. Journal of the World Aqauculture Society 28:91-96. Leighton, D. L. 1960. Studies of kelp-grazing organisms. In, Kelp investigation program, quarterly report, University of California, Institute of Marine Research., IMR Ref. 607, 13-22. Leighton, D. L. 1966. Studies of food preference in algivorous invertebrates of Southern California kelp beds. Pacific Science 20:104-113. Leighton, D. L. 1971. Grazing activities of benthic invertebrates in southern California. In, The biology of giant kelp beds (Macrocystis) in California, edited by W.J. North, Beihfte zur Nova Hedwigia, Vol. 32, pp. 421-453. Leighton, D. L., L. G. Jones, and W. J. North. 1966. Ecological relationships between giant kelp and sea urchins in southern California. Pages 141-153 in E.G. Young and J.L. McLachlan, editors. Proceedings of the fifth international seaweed symposium. Pergamon Press, Oxford. Lemire, M., and J. H. Himmelman. 1996. Relation of food preference to fitness for the green sea urchin, Strongylocentrotus droebachiensis. Marine Biology 127:73-78. Levitan, D.R., M. A. Sewell, and F. S. Chia. 1992. How distribution and abundance influence fertilization success in the sea urchin Strongylocentrotus franciscanus. Ecology 73:248-254. Low, C. J. 1975. The effect of grouping of Strongylocentrotus franciscanus, the giant red sea urchin, on its population biology. Ph.D. dissertation. University of British Columbia, Vancouver.  McBride, S. C , W. D. Pinnix, J. M. Lawrence, A. L. Lawrence, and T. J. Mulligan. 1997. The effect of temperature on production of gonads by the sea urchin Strongylocentrotus franciscanus fed natural and prepared diets. Journal of the World Aqauculture Society 24:357-365. McBride, S. C , J. M. Lawrence, A. L. Lawrence, and T. J. Mulligan. 1998. The effect of protein concentration in prepared feeds on growth, feeding rate, total organic absorption, and gross assimilation efficiency of the sea urchin Strongylocentrotus franciscanus. Journal of Shellfish Research 17:1563-1570. Muse, B. 1998. Management of the British Columbia sea urchin fisheries. Alaska Commercial Fisheries Entry Commission, Juneau, Alaska. CFEC 98-2N. 27p.  North, W. J. 1964. Ecology of the rocky nearshore environment in Southern California and possible influences of discharged wastes. In, Advances in water pollution research. E.A. Pearson ed. Pergamon Press, New York. Pearse, J. S., M. E. Clark, D. L. Leighton, C. T. Mitchell, and W. J. North. 1970. Kelp habitat improvement project, 1969-70, California Institute of Technology, Appendix, 87 pp. Schroeter, S. C. 1978. Experimental studies of competition as a factor affecting the distribution and abundance of purple sea urchins, Strongylocentrotus purpuratus (Stimpson). Ph.D. dissertation. University of California, Santa Barbara. Tegner, M. J., and P. K. Dayton. 1977. Sea urchin recruitment patterns and implications of commercial fishing. Science 196:324-326. Vadas, R. L. 1977. Preferential feeding: an optimization strategy in sea urchins. Ecological Monographs 47:337-371. Watson, J. C. 1993. The effects of sea otter (Enhydra lutris) foraging on shallow rocky communities off northwestern Vancouver Island, British Columbia. Ph.D. dissertation. University of California Santa Cruz, Santa Cruz.  5. Urchins and Fish: The Influence of Urchin Removal on the Abundance of Fish ABSTRACT The shallow (less than 15 m) nearshore fish community near Hot Springs Cove in Clayoquot Sound, British Columbia, is dominated by several species in the perch (Embiotocidae), greenling (Hexagrammidae) and rockfish (Scorpaenidae) families. As part of an experiment designed to determine the possible impacts of conducting a sea urchin fishery in this area, members of these fish families were monitored both before and after experimental manipulations in sea urchin density. Sea urchins were manipulated at four different levels that resulted in the conversion of urchin barrens habitat to variable density kelp forest habitat consisting primarily of Laminaria bongardiana, Pterygophora californica, and Nereocystis luetkeana. Of the seven fish species monitored, pile perch, striped seaperch, kelp perch and black rockfish were most associated with kelp forest habitat. Kelp greenling, lingcod and copper rockfish showed no association with kelp forest habitat. The experimental approach taken indicates that sea urchin removal, and subsequent kelp growth, determined relative fish abundance.  INTRODUCTION It is established that marine vegetation plays an important role in the presence and abundance of some species. In temperate nearshore waters, kelps and seagrasses are of greatest importance (Quast 1968, Phillips 1984, Dean et al. 2000). Habitats with kelp and seagrasses tend towards greater fish density than neighbouring habitats without these features (Briggs and O'Connor 1971, Orth and Heck 1980, Stoner 1983, Larson and DeMartini 1984) and density within sites has been shown to be affected when these features are added or  131  removed (Ebeling and Laur 1985, 1988, Bodkin 1988, Carr 1989). Explanations for vegetation-fish associations often include provision of food resources (Schmitt and Holbrook 1984) and as refugia from predation (Ebeling and Laur 1985). However, a combination of both factors is probably important for many fish species and the relative effects of each may change in the presence of predators and with the age of the fish species concerned (Carr 1989). Factors other than vegetation presence are important to fish distribution. On regional scales, temperature is perhaps one of the most important determinants offishdiversity and abundance (Hubbs 1948, Ojeda and Dearborn 1990, Murawski 1993). Ojeda and Dearborn (1990) showed considerable fluctuation infishdiversity and abundance associated with seasonal water temperature changes on the east coast of the United States. Depth is also important in determiningfishpresence and abundance over both local and regional scales (Quast 1968, Larson 1980). On local scales, recruitment and juvenile survivorship can be limited by the presence of adult conspecifics and predators (Behrents 1987). By increasing the availability of shelter, Behrents (1987) was able to increase the abundance offishon artificial substrates. Auster et al. (1995) also showed the importance of microhabitat presence for the distribution of numerous demersal fish species occupying low-relief, softbottom habitats. Numerous other authors have shown that substratum configuration and type is an important determinant of local fish distributions (Quast 1968, Larson 1980, Larson and DeMartini 1984, Patton et al. 1985, Ojeda and Dearborn 1990, Dean et al. 2000). As well, biological processes such as competition can limit the local abundance and distribution of some species (Larson 1980, Behrents 1987). Some fish change their habitat preferences  based on diel patterns (Auster et al. (1995). Finally, fish mortality can increase due to storm events thereby affecting local and regional abundance offishspecies (Bodkin et al. 1987). Examination of any reason for patterns of abundance can be affected by the methodology (experimental vs. sampling) and scale (temporal and spatial) at which the system is monitored (Bunnell and Huggard 1999). I chose experimental techniques to evaluate links of observed changes in fish abundance with the factor being manipulated. The potential to assign causation comes with the drawback of having less time and resources to expand spatial area covered by the study. Other studies have examined fish-habitat associations across greater spatial scales and with more variation in habitat parameters that may be of importance to the study species (Richards 1987, Dean et al. 2000). Although such studies indicate the importance of habitat parameters, such as kelp, for the presence of fish they lack the ability to assign causation between the sampled factors and fish abundance and may suffer from problems of colinearity with other factors. I also decided to monitor the study during all seasons, as opposed to during a single season of favourable conditions. That decision further reduced accessible sample area, but allowed observation of seasonal changes in fish abundance and how such differences might be related to fluctuations in kelp abundance. I studied the interaction between a potential sea urchin fishery and nearshore fish communities in Clayoquot Sound, British Columbia. The mode of action whereby sea urchins can initiate control overfishpresence and abundance is by their overwhelming control of kelp communities (Ebeling and Laur 1988, and Chapter 2). To evaluate the relationship, the removal of sea urchins at various levels was used to promote kelp growth. This experimental manipulation of kelp habitat allowed the determination of causality among  133  kelp-fish associations and is arguably the "best test of the influence of kelp on fish abundance" (Larson and DeMartini 1984). During the study, I monitored the density of seven fish species in three families (Table 5-1) before and after removal of the sea urchin Strongylocentrotus franciscanus. Evidence is presented that members of the Embiotocidae and Scorpaenidae were closely associated with the resulting kelp structure, while members of the Hexagrammidae were relatively unaffected by the changes in kelp abundance due to the removal of S. franciscanus.  METHODS AND MATERIALS Fish Surveys The study sites and experimental design are the same for this experiment as discussed in Chapter 2. Here, I address the response of fishes to the same levels of urchin removal discussed in previous chapters. To measure fish abundance, I swam three 15-m transects for each site by treatment combination. Transect bearings were the same as those used in the urchin and kelp surveys. To conduct a transect, my dive assistant would affix a 15-m length of twine to the plot centre. I would then swim along the selected bearing while my assistant followed and notified me when we were out of twine. All fish, of the species listed in Table 5-1 were counted that occurred within one meter to either side of the transect line. In the case of young-of-year rockfish, I commonly had to estimate numbers within a school as the fish were moving and the group size was too large to readily count. All transects were conducted approximately 1.0 meter from the bottom and fish were not counted more than one-meter overhead. No attempt was made to vigorously search demersal kelps for fish or to search for canopy-associated fish when canopy was available. Thus, fish abundance in kelp forested study plots are conservatively estimated. All time periods used to sample urchins  134  and kelp were used to sample fish with the exception of fall 1998, when severe weather forced completion of sampling without the fish data.  Data Analysis To present data in a fashion comparable to previous chapters, the three fish families observed were analysed as a repeated measures two-way analysis of variance. The experimental design can be viewed as a split-plot with time as the split-plot factor (as per Kuehl 1994) and the models are presented as such. In the model, time and study sites are treated as random factors and treatment as a fixed effect. Mean fish densities varied greatly between periods. For this reason as well as a noted dependence of sample variances on sample means, data were transformed as logio (x + 0.001), where x is the estimated mean number of fish per 100 m . To evaluate the relationship between kelp and fish abundance, fish frequency data were analysed with Chi-squared goodness of fit tests. All tests were conducted on summer data for years two and three, as these were the periods of greatest fish and kelp abundance. (  Treatment plots were combined because all levels of urchin removal resulted in the recruitment of kelp as compared against control plots (which continued to remain largely free of kelp throughout the study). For the tests, fish families were analysed as groups as well as subdivided for the two species of rockfish and for young-of-year rockfish. Young-of-year rockfish were defined as those fish less than 7-cm long. This distinction was based on a natural break in the size data as well as previous work in the field (Hay et al. 1989). Rockfish were subdivided because they seemed to have the greatest association with the kelp plots, and I wanted to explore the relationship further. The embiotocids also showed promising trends in relation to kelp habitat. They were not subdivided due to smaller sample  135  size and data collection errors that prevent separation of data for Brachyistius frenatus and Embiotoca lateralis.  RESULTS Embiotocids During the course of this study a total of 123 embiotocids were counted, of which 54% were of B. frenatus and E. lateralis combined, and the other 46%) were Rhacochilus vacca. Results of the repeated measures A N O V A show a clear temporal effect for the combined embiotocids (Fiz&=29.54, P<001; Table 5-2). This is not surprising as only 3% of the observations for the group were made in the pre-treatment summer data collection. The first and second post-treatment years accounted for 77% and 20% of observations respectively. Furthermore, all observations except for one were made during summer sampling periods (Table 5-3). The A N O V A failed to detect a significant treatment effect for Embiotocids. I believe this result is due to low power of the experimental design and the highly seasonal pattern of occurrence for Embiotocids. Furthermore, the sampling technique and the patterns of kelp abundance created by the experiment tend to increase the probability of committing a Type II error (see discussion). Because the pattern of embiotocid occurrence tended to follow the pattern of Nereocystis luetkeana recruitment (compare Figure 5-1 and Figure 2-7), Chi-squared goodness of fit analyses were conducted to determine if associations with post-treatment kelp forest exist. For embiotocids, Yate's corrected Chi-squared values were 2.19 and 4.50 with probabilities of 0.14 and 0.03 for years two and three, respectively (Table 5-4). The biological significance of these values is discussed below.  136  Hexagrammids During the study, 196 observations of hexagrammids were made. Only two conspicuous members of this family were encountered frequently and were the only family members noted while counting. Lingcod, Ophiodon elongatus, comprised 7% of the hexagrammid observations, and the remaining 93% of observations were of the kelp greenling, Hexagrammos decagrammus. The hexagrammid group was more consistently encountered than either the embiotocids or the scorpaenids, with a maximum calculated density of 4.26 fish per 100 m in summer 1998 and a minimum of 0.65 fish per 100 m in winter 1999 (see Table 5-3 and Figure 5-2). Analysis of variance results (Table 5-2) indicate a temporal effect for hexagrammid density. Visual inspection of Figure 5-2 and Tukey HSD tests indicate that the maximum and minimum values mentioned above are responsible for the effect (P<0.001). It is possible that the maximal value observed in summer 1998 is r  related to the abundance of juvenile perch and rockfish observed during this time period as these would be suitable prey for lingcod and kelp greenling (Rass 1970, Shaw and Hassler 1989). However, as opposed to the embiotocids and scorpaenids, the effect is not strongly related to the implementation of treatments. The same can be said for the observed significant treatment by site interaction (Error I) in Table 5-2. No patterns were found indicating biological significance for this interaction. Moreover, there are no clear trends within the interaction that would suggest that either component (treatment or site) should be considered biologically significant. For consistency, I did subject the summer data for years two and three to a Chi-square goodness of fit as for the embiotocids above (Table 5-4). There was no observed association between hexagrammids and kelp forest habitat for either year using this technique (Yates corrected Chi-square = 1.86 & 0.18; P= 0.17 & 0.67 for years two and three respectively).  Scorpaenids The scorpaenids followed trends similar to those observed for the embiotocids. Of the 838 observed members of the family, 0.6% were observed in the summer of 1997 while 40%) and 59% were observed in the summers of 1998 and 1999, respectively. Black rockfish, (Sebastes melanops) and copper rockfish (Sebastes caurinus) were the two primary rockfish species observed in the area and the only species enumerated. Of these 13% were black rockfish beyond their first year and 2% were copper rockfish older than one year. The remaining 85% of observations were of combined young-of-year rockfish. Results of the repeated measures ANOVA support these observations and reveal a significant temporal effect for scorpaenid densities (F^s-l 34.64, PO.001). The fact that fish density increased greatly in the summers of 1998 (31.30 fish per 100 m ) and 1999 (45.46 fish per 100 m ) but 2  2  not during the first summer of 1997 (0.46 fish per 100 m ) indicate that implementation of the treatments is the likely cause of the temporal effect (see Table 5-3 and Figure 5-3). As with the embiotocids, the experimental design and analysis decreases the power to detect treatment effects as many time periods were devoid of or had very low densities of scorpaenids (Table 5-3). In addition, the sampling methodology may actually increase the probability of committing a Type II error (discussed below) further affecting my ability to determine significant treatment effects. Thus, the Chi-square technique was again used to assess associations between members of the scorpaenidae and kelp habitat. Chi-square analyses for rockfish have been broken into three groups for 1998 and four groups for 1999. In 1998,1 counted separately copper, black and combined young-ofyear (yoy) rockfish. In 1999,1 further subdivided the yoy rockfish into copper and black yoy rockfish. Black rockfish showed a low association with kelp in 1998 (Yates % = 2.08, P = 0.15)"and a very high association in 1999 (Yates % = 22.72, P < 0.001). The difference is 138  likely due to the larger sample size in 1999 when 87 black rockfish were observed, as opposed to only 16 in 1998. Copper rockfish showed no association with kelp in either 1998 or 1999. However, the small sample size prevents accurate measurement of such an association for copper rockfish (see Table 5-4). Combined yoy rockfish showed a strong association with kelp habitats in 1998 (Yates % = 35.94, P < 0.001), but a weaker association 2  in 1999 (Yates x = 1-85, P = 0.17). Subdivision of the 1999 data show that yoy black 2  rockfish are more highly associated with kelp habitat than yoy copper rockfish (Yates x = 2  3.19 and 0.60, P = 0.07 and 0.44, respectively; Table 5-4).  139  Table 5-1. Classification of fish enumerated during this experiment by family name, common name, and genus and species.  Family  Common Name  Scientific Name  Embiotocidae  Kelp Perch  Brachyistius frenatus  Embiotocidae  Pile Perch  Rhacochilus vacca  Embiotocidae  Striped Seaperch  Embiotoca lateralis  Hexagrammidae  Kelp Greenling  Hexagrammos decagrammus  Hexagrammidae  Lingcod  Ophiodon elongatus  Scorpaenidae  Black Rockfish  Sebastes melanops  Scorpaenidae  Copper Rockfish  Sebastes caurinus  Table 5-2. Split-plot analysis of variance offish density (+ 0.001 logm transformed) for the three families of fish encountered during the study. Treatment effects tested using the pseudo f-test. Error d.f. for treatment effects is 4.42, 4.45 and 5.50 for the Embiotocidae, Hexagrammidae and Scorpaenidae respectively (refer to Hicks 1982). DF  MS  Error MS  F  P  Site  2  3.18  2.04  1 . 5 6  0.23  T r e a t m e n t  2  6.60  3.71  1 . 7 8  0.29  Error 1  4  3.15  2.04  1 . 5 4  0.22  T i m e  7  60.37  2.04  29.54  0.00  Site by T i m e  1 4  7.59  T r e a t m e n t by T i m e  1 4  2.60  Error 2  2 8  2.04  —  —  —  Sampling Error  1 8 9  1 . 8 3  —  —  —  Site  2  10.94  5.39  2.03  0.15  T r e a t m e n t  2  14.00  26.64  0.53  0.69  Error 1  4  24.89  5.39  4.61  0.01  T i m e  7  20.12  5.39  3.73  0.01  Site by T i m e  1 4  8 . 0 7  T r e a t m e n t by T i m e  1 4  7.15  Error 2  2 8  5.39  —  —  —  Sampling Error  1 8 9  6.91  —  —  —  Site  2  1 . 7 0  1 . 3 6  1 . 2 5  0.30  T r e a t m e n t  2  2.57  2.27  1 . 1 3  0.42  Error 1  4  1 . 4 7  1 . 3 6  1 . 0 8  0.38  T i m e  7  1 8 2 . 6 6  1 . 3 6  1 3 4 . 6 4  0.00  Site by T i m e  1 4  1 . 3 4  T r e a t m e n t by T i m e  1 4  2.16  Error 2.  2 8  1 . 3 6  —  —  —  Sampling Error  1 8 9  2.36  —  —  —  Effect Embiotocidae  —  2.04  —  1 . 2 7  —  0.28  Hexagrammidae  —  5.39  —  1 . 3 3  —  0.25  Scorpaenidae  —  1 . 3 6  —  1 . 5 9  —  0.14  Al  CU  Ml O  .5 s= S  oo  CU  NO  O  "8 «  +1  CN CN CN  ii  ea cs — t«  NO  NO  ro  d  d  ON  cN d  m d  m d  m  t-«  oo r» CN  ON  CN >n d  o m d  NO  I—1  ro  CN  CN  oo CN d  CN CN d  CN CN d  +1  NO  +1  d  +1  +1  in CN  ON o +1  NO t-; NO ro  +1  1 —1 d  CN d  oo CN d  in  NO  ON  +1  NO  d  +1  m d  ON  o d  00  ON  NO  +1  o d  ro  +1  00  r-; CO  +1  ON  o  ©  +1  NO ON  ON  1 1 1  1 1 1  1 1 1  NO ON  NO ON  o  00  o d  OJ  a *  ZS  ON OS OS  cu u  '3  r-  cu V a w  00  a oo  .  +1  i j  • • •  1 1 1  ON  +1  ON  o d  —-*  d  +1  o  +1  ON  o d  ON • • •  o d  •« ^ a  "3  o .2 cs O cs  5 .a « s (W  JM  a  S  +1  E  (H  o  -s ° 2  O O >> >> ja o  i i  i  ro  o CN  NO  ON  ,—^  ,—^  ,—^  i  | 1 1  >n  +1  >n  NO  NO  d  d  ro  ro  d  d  NO  NO  •  CM  cs .S s cu ~ o .  5 "B 0»  +1  )M  M  » 2 » rt  I£ CS  S  o = ,o cu cu  IS  oo  CU  o  o\  Q,  si  +1  o oo  NO  00  •<t  ro  in  +1  t~  oo .S a O0  §1  +1  i i •  ii •  1 1 1  l_  ON  CO rM  E Q •O —  1 1  +1  i i  • • •  1  1 1  NO  +1  CN  NO  ro d  +1  +1  r~  ro  —<  ro  CN  —;  ON  CN ro  ?, m NO d  ro d  +1  NO >n d  ON o ?l ON o d  ?, 00  ?, OO  ON  ON o  +1  ?l ON  O O  +1  d  in d  +1  d  '  O  +1  ON  ro  ON  1  <n oo  o • • •  ON  ro  CN in oo  ON O d  ©  ON  o d  o  +1  ON  in CN  o in CN  ON  ON  NO  ro  +1  +1  ?l ON  +1  ON  o d  i (A  es <S — i_ cu U +1  ON  ir* .  ON  *i  u CU  UO  uo in d  2*  oo  +1  CN  Tf d  i- o CU  NO  cu i—i >> -a .-a o A, cu cn  o d  +1  a  0  ON  O ©  o d  +1  ON  • • •  p  ro d NO  d  ro © in  d  +1  ON  rt  d  NO  oo  ON  ^  1 8. c  OA  ro ©  r-ro d  CN d  ro d  rro d  rro d  ro  NO  +1  ii cs a  2B  +1  • • •  S es cu >> cs cu  +1  +1  CN  o CN d  '—'  d  +1  +1 00  rro d  CN d  ON  ON  d  d  ON  ON '—'  — . I  d  d  d  p  +1  p  +1  ON  --^ d +1  ON  ON  IM  a I -  CJ  a CJ  es o w U *** S  a  « S  J3  E * .3 c«  O  in E  a  cu i_ cu  3  00  1  a  g  H cu u  a oo  I>  '&  oo  <3  a  cs  u 00 CS  cu  cu cs  *3 u  o ft  IS  -a o CJ  00  .3  '5  cu es  IS o  u  CJ CU  a  IM  o u  «5  O  CD  a a o U  00  it  o cd  3  ca  CD  00  s  oo  142  Figure 5-1. Calculated density of fish per 100 m for combined members of the Embiotocidae for each time period sampled. Bars are ± one standard error of the mean (sample size is 27 for Kelp and 9 for Control). A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 2  Hexagrammidae Density Through Time  Figure 5-2. Calculated density offish per 100 m for combined members of the Hexagrammidae for each time period sampled. Bars are ± one standard error of the mean (sample size is 27 for Kelp and 9 for Control). A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 2  S c o r p a e n i d Density Through T i m e 90  r  80 • 70 • CN  E o o •— CO cj=,  60 50 40 30  CO  c0 Q  20 10 0  -o- Kelp Sum 97  Fall 97  Win 98  Spr 98  Sum 98  Win 99  Spr 99  Sum 99  Control  Time Figure 5-3. Calculated density offish per 100 m for combined members of the Scorpaenidae for each time period sampled. Bars are ± one standard error of the mean (sample size is 27 for Kelp and 9 for Control). A list of sampling dates corresponding to each sampling season is given in the Appendix (Appendix; Table 1-3). 2  145  Table 5-4. Chi-squared goodness of fit between numbers offish observed in kelp containing vs. non-kelp containing study plots for the summers of 1998 (year 2) and 1999 (year 3). Kelp plots were all plots where some level of urchin removal was conducted and non-kelp plots were the three control plots.  Kelp  No Kelp  Observed Expected Observed Expected  x  xVates) P  2  P  (Yates)  Year 2  Embiotocidae  78  71.25  17  23.75  2.56  2.19  0.11  0.14  Hexagrammidae  30  34.50  16  11.50  2.35  1.86  0.13  0.17  304  253.50  34  84.50 40.24  39.45  0.00  0.00  15  12.00  1  4.00  3.00  2.08  0.08  0.15  7  6.00  1  2.00  0.67  0.17  0.41  0.68  282  235.50  32  78.50  36.72  35.94  0.00  0.00  Embiotocidae  23  18.00  1  6.00  5.56  4.50  0.02  0.03  Hexagrammidae  21  22.50  9  7.50  0.40  0.18  0.53  0.67  400  360.25  91  122.75  12.60  12.23  0.00  0.00  85  65.25  2  21.75  23.91  22.72  0.00  0.00  5  5.25  2  1.75  0.05  0.05  0.82  0.83  YOY Rockfish  310  297.75  87.  99.25  2.01  1.85  0.16  0.17  YOY Black  279  264.00  73  88.00  3.41  3.19  0.06  0.07  31  33.75  14  11.25  0.89  0.60  0.35  0.44  Scorpaenidae Black Rockfish Copper Rockfish YOY Rockfish  Year 3  Scorpaenidae Black Rockfish Copper Rockfish  YOY Copper  DISCUSSION  Embiotocids The levels of urchin removal conducted in this experiment affected the observed members of the perch family. A temporal trend was evident in which perch densities were highest in the summers following the removal of sea urchins, while they were virtually absent before implementation of the treatments. These results are not surprising for perch as a group are known to associate with kelp forest habitats (Ebeling and Laur 1985, Fritzsche and Hassler 1989). The one deviation in the results that places any doubt on the association of the group with kelp habitat is the statistically insignificant association observed for kelp habitats and perch in the summer 1998 data. In the field, I observed an obvious association between kelp t  and all observed perch during this period. The study was not designed explicitly to determine this association. The sampling methodology and the spatial patterns of available kelp forest habitat may enhance the probability of committing a Type II error. The primary factor is the selection of a 15-m long transect to study fish abundance. Given that urchins were removed using a 15-m lead line along the sea floor, the use of a 15-m transect while swimming (not following bottom contour) for fish meant that the fish transects extended beyond the treatment area. In plot areas with the greatest relief, this effect was most exaggerated. The effect is even further compounded by the urchins' ability to re-invade and remove kelp in treatment areas between sampling periods. The net effect is a greater probability of having lower abundance of kelp approximately 10-15 m beyond plot centre than between 0-10 m for urchin removal plots, with the opposite being true for control plots. Thus, if association occurs betweenfishand kelp, the sampling design would tend to miss the 147  association because some fish encountered in kelp near the end of transects would be included in counts for control plots. Generally, such plots were shown to have no kelp based on the sampling methodologies of Chapter 2. Moreover, the presence of kelp itself presents difficulty. The increased structure provided by the kelp plants works to conceal fish, perhaps the reason the fish are there in the first place. Thus, concealment by kelp works to suppress the number of fish observed in kelp plots. Given my experience, data and the increased likelihood of committing a Type II error, I support the hypothesis that embiotocids are attracted to the kelp provided in the absence of sea urchins. Other researchers have suggested that kelp habitat is only of importance to young perch (Ebeling and Laur 1985). Ebeling and Laur (1985) included five species in their study and justified grouping them for analysis. Only E. lateralis was shared with this experiment. They convincingly showed that adults had no dependence on kelp habitat due to 'greater latitude in foraging space'. I propose the greater association of all size classes with kelp habitat (the data showed no trends with regards to size) were the result of enhanced food resources within the kelp habitat. I observed high densities of various molluscs and arthropods in the kelp plots following removal of urchins. Both of these prey groups are important to pile perch and striped seaperch (see Fritzsche and Hassler 1989 for review). Outside of the kelp patches, these food items were greatly reduced by the overwhelming competitive dominance of sea urchins. The attraction to kelp seems to be so complete that not even the striped seaperch remain when kelp cover is absent, but see Stouder(1987).  Hexagrammids Hexagrammids enumerated for this experiment did not show any clear trends with regards to removal of sea urchins and the appearance of kelp habitats. Temporal trends associated with the group were generally the effect of a single increase in fish density during one season. In the summer of 1998, kelp greenling were found at the highest density encountered during the experiment. Because this data point is not part of an apparent trend or repeated cycle, it is difficult to attribute the significance of the observed temporal effect to anything other than random variation. It is interesting that the maximum observed density was during a summer sampling period. Embiotocids and scorpaenids also are observed in greatest numbers during summer sampling periods. While it is impossible to say that the hexagrammids are associated with kelp habitats, it could be that the increases in the other groups provide an increase in suitable prey species (Rass 1970, Hallacher 1977, Shaw and Hassler 1989). This feature could attract more greenling and lingcod to the study area during the summer months. Recent research in fish-habitat associations in the coastal eastern Pacific has indicated an association between kelp greenling and kelp habitats (Dean et al. 2000). However, Dean et al. (2000) did not sample areas that adequately controlled for similar habitats without kelp but compared areas with variable kelp and seagrass assemblages across sites with differing relief and exposure. While it cannot be stated to what degree kelp greenling may select one area over another based on physical site characteristics, there is no specific relationship with kelp abundance. I suggest that any associations reported by Dean et al. (2000) suffer from problems of colinearity where both the fish and the kelp are responding to similar physical site properties. I suspect the same is true for lingcod; however, the low sample size for this species requires caution in making any specific statements. 149  It is important to point out that the data could be impacted by the immobility of the two species in this group. Both kelp greenling and lingcod tend to lie on the bottom and move little when approached. This behaviour could prevent the observation of fish in areas with high demersal kelp abundance. Given the fact that a consistent number of fish were sighted throughout the study, this is not a likely problem. Furthermore, most of the observations within this group were of mature individuals (beyond first year) whose size made them easily observed.  Scorpaenids Scorpaenid trends with kelp are complex because the trends vary across species and size classes. Overall, scorpaenids responded similarly to the embiotocids with increased observations in post-treatment, summer sampling periods when N. luetkeana canopy cover was greatest. By subdividing the observations by species and size class, I found that small and young-of-year black rockfish are highly associated with kelp forest. Large black rockfish and all copper rockfish tended not to associate with kelp forest (these comparisons sufferfromsmall sample size and lack of power). Moreover, my daylight observations may not completely represent associations with kelp. Leaman (1976) noted greater numbers of large black rockfish in kelp beds nocturnally. There is one observation that does not fit well with the above pattern. In the summer of 1998 following urchin removal, black rockfish (excluding young of year) were not observed to associate with kelp plots (Table 5-4). However, this point suffers from low sample size (n=16) as well as the problems associated with committing a Type II error as discussed for the embiotocids. If the observed trend were maintained with a doubling of  sample size, a significant relationship (at alpha = 0.05) would be observed with kelp for the 1998 black rockfish data. Copper rockfish in all categories as well as larger black rockfish were not observed to associate with kelp. Large rockfish were in fact largely absent from the study plots. However, during exploration of areas beyond the study sites, I noted that the larger members of this group tend to associate with high rocky relief structure in more exposed habitats. The lack of association with kelp for copper rockfish also suffers from small sample size in all sampled years (Table 5-4). However, this trend is likely real because within the sample there are few trends tending towards association. In fact, there were more observations than expected in control sites completely lacking kelp for young-of-year copper rockfish in 1999 and for non-young of year in 1998. The provision of kelp habitat for young-of-year black rockfish is beneficial not only to the young of year. Young-of-year rockfish form the basis for a "trophic freeway" whereby young fish are a trophic link between the plankton and their larger carnivorous adults (Hallacher 1977). Further, young rockfish are an often-important prey item for many other carnivorous fish (Hallacher 1977) and thus any factor that enhances the recruitment of rockfish should tend to enhance the productivity of their predators. These results generally support the findings of other researchers who have noted associations between juvenile and young-of-year black rockfish with kelp habitats (Bodkin 1988) and no association for copper rockfish with kelp habitat (Richards 1987). Bodkin (1988) grouped young-of-year rockfish making determination of kelp associations independently of other species impossible. Furthermore, Richards (1987) did not utilise an experimental approach and so assigning causation to her observations is more equivocal.  151  The fact that these morphologically similar species diverge greatly in habitat association reveals a need for caution in studies that utilise taxonomic guild groupings instead of individual species data.  CONCLUSIONS AND IMPLICATIONS Obtaining accurate natural history data on subtidal marine species presents unique obstacles for ecologists. The constraints of SCUBA and the cost of doing research in an environment in which humans are physiologically constrained presents difficulty. Compounded with often-dangerous conditions, the result is an inadequate amount of focused time spent studying a given organism in its natural surroundings. For this reason, no work on a subtidal organism should ever be considered definitive. Instead, we need to pool information from numerous studies to ensure we adequately understand a given species or community. As well, there is the need to understand the distribution and associations for species on a variety of scales (Bunnell and Huggard 1999). For instance, surface trawls can not offer much needed information on microhabitat selection, just as scuba based research cannot adequately address distributional patterns and fluctuations due to changes in climate and or current patterns. The data in this chapter provide a better understanding of the ecological processes by which sea urchins can affect an entire community as well as understanding the microhabitat ft  requirements of some common coastal fish of the northeastern Pacific. By experimentally manipulating the density of S. franciscanus, I was able to alter the primary communitystructuring agent (kelp) and therefore alter the abundance of some ecologically and economically important fish species. However, the abundance of fish is not an adequate measure of the long-term viability of a population. The small isolated kelp patches created in  this study may offer poor quality habitat (reducing reproductive success and / or increasing predation) for the fish species studied (sensu Pulliam 1988). The attraction of fish to the study plots indicates the importance of the habitat modifications and that at some spatial scale such modifications will provide for sources offishpopulations. Thus, any management plan involving S. franciscanus should consider the possible impacts on fish species. These data suggest that in many cases the removal of sea urchins will increase the local density of some species of fish (see Chapter 6 for more detailed implications).  REFERENCES Auster, P. J., R. J. Malatesta, and S. C. LaRosa. 1995. Patterns of microhabitat utilization by mobile megafauna on the southern New England (USA) continental shelf and slope. Marine Ecology Progress Series 127:77-85. Behrents, K. C. 1987. The influence of shelter availability on recruitment and early juvenile survivorship of Lythrypnus dalli Gilbert (Pisces: Gobiidae). Journal of Experimental Marine Biology and Ecology 107:45-59. Bodkin, J. L. 1988. Effects of kelp forest removal on associatedfishassemblages in central California. Journal of Experimental Marine Biology and Ecology 117:227-238. Bodkin, J. L., G. R. VanBlaricom, and R. J. Jameson. 1987. Mortalities of kelp-forest fishes associated with large oceanic waves off central California, 1982-1983. Environmental Biology of Fishes 18:73-76. Briggs, P. T., and J. S. O'Connor. 1971. Comparison of shore-zone fishes over naturally vegetated and sand-filled bottoms in Great South Bay. New York Fish and Game Journal 18:15-41. Bunnell, F. L., and D. J. Huggard. 1999. Biodiversity across spatial and temporal scales: problems and opportunities. Forest Ecology and Management 115:113-126. Carr, M. H. 1989. Effects of macroalgal assemblages on the recruitment of temperate zone reef fishes. Journal of Experimental Marine Biology and Ecology 126:59-76. Dean, T. A., L. Haldorson, D. R. Laur, S. C. Jewett, and A. Blanchard. 2000. The distribution of nearshore fishes in kelp and eelgrass communities in Prince William Sound, Alaska: associations with vegetation and physical habitat characteristics. Environmental Biology of Fishes 57:271-287.  153  Ebeling, A. W., and D. R. Laur. 1985. The influence of plant cover on surfperch abundance at an offshore temperate reef. Environmental Biology of Fishes 12:169-175. Ebeling, A. W. 1988. Fish populations in kelp forest without sea otters: effects of severe storm damage and destructive sea urchin grazing. Pages 169-189 in G.R. VanBlaricom and J.A. Estes, editors. The community ecology of sea otters. SpringerVerlag, Berlin. Fritzsche, R. A., and T. J. Hassler. 1989. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (Pacific Northwest)~pile perch, Striped Seaperch, and Rubberlip Seaperch. Biological Report 82(11.103) U.S. Fish and Wildlife Service. Hallacher, L. E. 1977. Patterns of space and food use by inshore rockfishes (Scorpaenidae: Sebastes) of Carmel Bay, California. Ph.D. dissertation. University of California, Berkeley. Hay, D. E., M. C. Healey, L. J. Richards, and J. B. Marliave. 1989. Distribution, abundance, and habitat of prey fishes in the Strait of Georgia. Pages 37-49 in K. Vermeer and R.W. Butler (eds.). The ecology and status of marine and shoreline birds in the Strait of Georgia, British Columbia. Canadian Wildlife Service Special Publication, Ottawa. Hicks, C. R. 1983. Fundamental concepts in the design of experiments. Holt, Rinehart, and Winston, New York. Hubbs, C. L. 1948. Changes in the fish fauna of western North America corrrelated with changes in ocean temperature. Journal of Marine Research 7:459-481. Kuehl, R. O. 1994. Statistical principles of research design and analysis. Duxbury Press, Belmont, CA. Larson, R. J. 1980. Competition, habitat selection, and the bathymetric segregation of two rockfish (Sebastes) species. Ecological Monographs 50:221-239. Larson, R. J., and E. E. DeMartini. 1984. Abundance and vertical distribution of fishes in a cobble-bottom kelp forest off San Onofre, California. Fishery Bulletin 82:37-53. Leaman, B. M. 1976. The association between the black rockfish, Sebastes melanops (Girard), and the beds of the giant kelp, Macrocystis integrifolia (Bory), in Barkley Sound British Columbia. M.Sc. thesis. University of British Columbia. Murawski, S. A. 1993. Climate change and marine fish distributions: forecasting from historical analogy. Transactions of the American Fisheries Society 122:647-658.  154  Ojeda, F. P., and J. H. Dearborn. 1990. Diversity, abundance, and spatial distribution of fishes and crustaceans in the rocky subtidal zone of the Gulf of Maine. Fishery Bulletin 88:403-410. Orth, R. J., and K. L. Heck. 1980. Structural components of eelgrass (Zostera marina) meadows in the lower Chesapeake Bay-fishes. Estuaries 3:278-288. Patton, M. L., R. S. Grove, and R. F. Harman. 1985. What do natural reefs tell us about designing artificial reefs in Southern California. Bulletin of Marine Science 37:279298. Phillips, R. C. 1984. The ecology of eelgrass meadows in the Pacific Northwest: a community profile. U.S. Fish and Wildlife Service. FWS/OBS-82/24. 85 pp.. Pulliam, H.R. 1988. Sources, sinks, and population regulation. The American Naturalist 132:652-661 Quast, J. C. 1968. Fish fauna of the rocky inshore zone. Pages 35-55 in W. J. North and C. L. Hubbs, editors. Utilization of kelp-bed resources in Southern California. California Department of Fish and Game, Sacramento. Rass, T. S., editor. 1970. Greenlings: taxonomy, biology, interoceanic transplantation. Keter Press, Jerusalem. Richards, L. J. 1987. Copper rockfish (Sebastes caurinus) and quillback rockfish (Sebastes maliger) habitat in the Strait of Georgia, British Columbia. Canadian Journal of Zoology 65:3188-3191. Schmitt, R. J., and S. J. Holbrook. 1984. Ontogeny of prey selection by black surfperch Embiotoca jacksoni (Pisces: Embiotocidae): the roles of fish morphology, foraging behavior, and patch selection. Marine Ecology Progress Series 18:225-239. Shaw, W. N., and T. J. Hassler. 1989. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (Pacific Northwest)~lingcod. Biological Report 82(11.119) U.S. Fish and Wildlife Service. Stoner, A. W. 1983. Distribution of fishes in seagrass meadows: role of macrophyte biomass and species composition. Fishery Bulletin 81:837-846. Stouder, D. J. 1987. Effects of a severe-weather disturbance on foraging patterns within a California surfperch guild. Journal of Experimental Marine Biology and Ecology 114:73-84.  6 . Summary Studies were conducted at three sites within the traditional territory of the Hesquiat First Nation in Clayoquot Sound on the west coast of Vancouver Island, British Columbia. Each of the sites varied in wave exposure, currents, substratum, and initial densities of S. franciscanus. However, sea urchin grazing heavily influenced all sites. The densities of urchins, kelp, and fish were monitored and samples of urchins were collected to obtain measures of gonad index. The system was monitored for nine seasons and the density of S. franciscanus manipulated to four levels during the last seven of those seasons. The results of this investigation can help guide the management of sea urchin resources in the northeastern Pacific Ocean. This work also contributes to the general knowledge of kelp forest and sea urchin ecology. The research has expanded on previous studies by both manipulating urchins at biologically relevant scales and through herbivore removals beyond the all or nothing approach common in past urchin-algae studies (Table 1-1).  The Hesquiat People and their program, Management for a Living Hesquiat Harbour, helped to frame and guide this research. One of the main interests of the Hesquiat was to understand the following question: how does the removal of sea urchins by a commercial fishery or sea otters influence the coastal resources of the traditional ecological territory of the Hesquiat First Nation? The experimental approach taken involved setting and maintaining sea urchins at different densities in study areas within Hesquiat territory. This research has practical relevance to the Hesquiat. The Hesquiat now better understand the potential results of removing sea urchins for commercial harvest and have seen what much of their coastal environments may look like if sea otters repopulate the Hesquiat traditional  156  territory. Moreover, this research has better prepared the Hesquiat to engage federal management agencies on resource management issues affecting their cultural and economic heritage.  MAJOR FINDINGS The removal of sea urchins resulted in a rapid increase in the abundance of laminarialeann algal species (kelps). Study plots where all urchins were removed developed a dense understory and canopy of kelps. Intermediate levels of removal resulted in a mosaic of smaller urchin-dominated and kelp-dominated patches. Control plots maintained the urchin-dominated barrens state throughout the study period. Increases in sea urchin fecundity (measured as gonad index) were also found after urchin densities were thinned. All study plots showed an increase in gonad index over time, but this increase was statistically greater for urchins in removal plots compared to those in control plots. A lowering of intraspecific competition and greater availability of kelp explained the increased gonad index of urchins remaining after thinning the population when urchins were reduced in density. Further support of food limitation in this population was obtained by containing and feeding urchins from a nearby barren area. Fed urchins increased their gonad index over a relatively short time (less than three months), illustrating the possibility of enhancing urchins taken from barrens areas for the uni market. Reductions in urchin density were shown to affect the abundance of fish in the study areas. Of the seven fish species monitored, pile perch, striped seaperch, kelp perch and black rockfish were more abundant in the kelp forest habitat created by urchin removal as opposed to pre-urchin removal densities. Kelp greenling, lingcod and copper rockfish did not show association with kelp forest habitats in either pre- versus post-removal comparisons or in  157  comparisons among removal and control plots. The experimental approach taken indicates that the relationships noted for perch and black rockfish were determined by urchin removal. The fact that some of these species are commercially fished and are important food items for the Nuu-Chah-Nulth (Clayoquot Scientific Panel 1995) indicate that these (and likely other) species need to be considered as part of sea urchin management in the study area. The findings suggest that there is an inverse relationship between the density of sea urchins and the occurrence of algal communities and supports the findings of other studies (Leighton et al. 1966, Paine and Vadas 1969, Steneck 1993, Estes and Duggins 1995, Andrew and Underwood 1993, Leinaas and Christie 1996). The exclusion of kelp forests in areas of high urchin density has implications far beyond those investigated in this study. When considered in conjunction with the results of other research into the importance of kelps, and kelp forests, this research illustrates that the management of S. franciscanus can have important ecological and economic side effects. Kelp forests provide increased habitable area for numerous species (see review in Harrold and Pearse 1987). In turn, some of these species are important links in kelp forest food webs (Hallacher 1977, Schmitt and Holbrook 1984, Fritzsche and Hassler 1989). Kelps also have economic value as food and a source of chemical compounds (Woodward 1965). The high productivity of kelps is also important to detrital and filter feeding organisms (Mann 1977, Duggins et al. 1989). Finally, increased kelp abundance may improve the herring roe on kelp industry (personal observation). Taking an experimental approach as opposed to a larger scale non-experimental sampling program allowed more definitive discussion of relationships that were observed between urchin density, roe production, kelp forests, and fish abundance. These data offer  useful information to management of not only sea urchins but for many vertebrate fisheries as well.  MANAGEMENT IMPLICATIONS Strongylocentrotus franciscanus are currently harvested in British Columbia. Managers and harvesters strive to maximise urchin harvest quantity and quality while ensuring a sustainable harvest level (Campbell et al. 1999). This work suggests that sea urchin management based on ecological information can increase the quality and abundance of urchin harvests while managing and sustaining biodiversity. This study has shown that harvesting sea urchins is likely to have important implications for kelp forest communities and associated fish species. It is apparent from this study that urchin manipulation can be used as a tool to control the density and localised abundance of both urchins and kelp. In turn, this tool can be used to improve the quality of numerous urchin beds, which are not now commercially fished due to poor urchin roe quality. The provision of kelp forest habitat is beneficial to some fish species (Chapter 5) and kelps are known to be important providers of food and habitat for numerous organisms in varying trophic pathways (refer to Chapter 1 and see review in Harrold and Pearse 1987). The low urchin-density threshold between the barrens state and kelp forest (Figure 2-9) indicates that wholesale culling to increase kelp abundance is an inappropriate management procedure on its own. Removal of urchins to one / m would leave too few urchins remaining to support the current harvesting strategy which requires high-density aggregations of urchins. However, the creation of mosaic patterns of kelp- and urchindominated patches after urchin removal could be an important management tool. This pattern has implications for a fishery conducted intensively at small spatial scales seeking to  159  produce quality product. Localised removals, below one urchin / m , create mosaics of kelp 2  surrounded by high-density aggregations of urchins. This strategy increases the available area of grazing fronts. Urchins in grazing fronts are commonly sought by harvesters due to their high density and the nearby presence of kelp. First Nations have long recognised kelp as an indicator of urchin quality (Clayoquot Sound Scientific Panel 1995) and numerous scientific studies have shown increased gonad production when kelp is non-limiting (Minor and Scheibling 1997, Kato and Schroeter 1985, Vadas 1977; see also Chapter 3). Increasing the density and area of urchins in grazing fronts may improve recruitment and survival of juvenile urchins. Localised kelp groups increase drift available to adult urchins. In turn, this may lessen competition between adult and juvenile urchins because the latter depend largely on diatomaceous films for food and compete with adults for this food item when macroalgae are limited (discussed by Cameron and Schroeter 1980). Furthermore, the high density of urchins in grazing fronts may provide increased protection of juvenile urchins from Pycnopodia, a predatory starfish (Duggins 1981), which is likely the predator of greatest impact in northern latitudes where sea otters are absent (Duggins 1981, 1983). Moreover, the high densities of grazing fronts may provide increased recruitment because of increased area of spine canopy (Low 1975, Tegner and Dayton 1977, Breen et al. 1985). Finally, sea urchin fertilisation is a function of group size and degree of aggregation (Levitan et al. 1992). Urchins in high-density aggregations may provide greater contribution to the planktonic larval supply than urchins in less dense aggregations (Allee effect; Allee 1931). However, Ebert (1998) calculated that substantial decreases in fertilisation (99.9%) have little impact on population growth. He suggests that increases in post-juvenile survival of 3 % will balance this decrease in fertilisation. I suggest that although lowered fertilisation  results in only small losses in population growth, the declines will compound through time. Moreover, the inverse density dependence of juvenile mortality means that density is still an important consideration to the management of sea urchins. Although increased rates of juvenile survival are likely as density increases (urchin aggregations provide more protection than individuals taken collectively) this hypothesis has not been addressed in the literature. The current use of fisheries quota and size restrictions is aimed at ensuring that harvest does not outpace recruitment and the ability of populations to make up for losses from harvesting. Although maximum size limits have been experimented with in the past and considered for permanent use in this fishery, there is currently no upper size limit. The lack of a maximum size limit is perhaps most important now that the Department of Fisheries and Oceans, Canada has decreased the lower size limitfrom100 mm to 90 mm (L. Chan pers. comm.). The concern is that large individuals will not remain to protect juveniles under spine canopies (Tegner and Dayton 1977, Breen et al. 1985, Rowley 1989; but see Botsford et al. 1994). Although the current fishery tends to limit the take of larger urchins because of higher demand on small urchins (D. Bureau pers. comm.), there is no assurance that a given harvester would not locally remove numerous large individuals. The lack of a rotational schedule means that, in theory, beds can be fished year after year. Such repeated disturbance may negatively affect juvenile urchins that suddenly lose the protection of an adult spine canopy. Perhaps more importantly, it could lead ultimately to the loss of larger individuals because there is no temporal protection to ensure members of preharvest size classes can obtain size refugia from harvest pressure. This could ultimately lead to the collapse of the fishery unless bed data are collected annually and quotas adjusted based on size class distributions. Such measures are likely beyond the resources currently available  to the management of this fishery. Spatial and temporal harvest rotation may lessen the demands for data needed to prevent the above scenario. Moreover, Pfister and Bradbury (1996) have predicted that the rotational fishery model followed in Washington, USA, will provide lower but more sustainable long-term yields compared to yearly harvesting. The results of the feeding experiment (Chapter 3) as well as other work on the enhancement of sea urchins (Klinger et al. 1986, de Jong-Westman et al. 1995, Lawrence et al. 1997, McBride et al. 1997), indicates gonadal plasticity based upon food availability. These data suggest considerable ability for roe enhancement in this species. If harvesters and managers want to increase the value of the fishery, then roe enhancement could be an economically viable solution. The approach could be in tandem with the preceding recommendation to clear localised areas of sea urchins to provide increased areas of urchin fronts adjacent to kelp forests. Urchins removed from relatively poor quality habitats to create the localised kelp forests can be enhanced before being shipped to market. As well, urchin roe can be enhanced before removal from the sea floor by the addition of farmed kelp (Druehl pers. comm.) or by the addition of drift kelp cut from existing canopies in urchin free areas. The latter methodologies are likely to result in less significant enhancement due to inability to target specific urchins (populations and size classes) and loss of kelp material out of the system. Within a management context, enhancement techniques can be used to increase the value of urchin resources. By increasing the value, managers can lower harvest levels if needed while maintaining profits for harvesters. If managers wish to regulate sea urchin harvests based on the ecology of this species, then perhaps the greatest issue is the spatial scale of management. Given recruitment that is low and sporadic over even small spatial scales (Bernard and Miller 1973, Low 1975, Breen  162  et al. 1976, 1978, Ebert 1983, Sloan et al. 1987, but compare to Tegner and Dayton 1981), and the dynamic relationship between sea urchins and kelp forests (Chapter 2, Harrold and Pearse 1987), managers need to consider management plans that are highly adaptive at relatively small spatial units. The spatial units may be areas such as the sub-areas currently managed by the DFO. Clayoquot Sound is one such current management unit (area 24). Within each area, intensive area-based management could be conducted at spatial scales as fine as the urchin bed. Some beds or regions could be enhanced, some fished based on the ecology outlined above, and others left for local community consumption or to repopulate. Although such management may not be feasible for the DFO (Campbell et al. 1999), it is feasible from a community management perspective. For instance, many Japanese coastal fisheries are managed at the local community level through fisheries co-operatives. In Japan, cooperative control of coastal areas has resulted in making Japan a world leader in fisheries production despite its relatively limited area (Mottet 1980). Cooperatives provide a forum whereby harvesters and aquaculturists can work in a system with mutual support instead of competition. Membership in a cooperative means access to fishing rights as well as profit sharing from activities that occur in areas under cooperative control. Moreover, cooperatives often provide extended services such as credit unions, mutual aid insurance, product marketing, and research, which act to stabilise the fishing community's economy. Finally, community-based stewardship of local resources means people are highly involved in the resources and attempt to maximise long-term productivity. All this occurs with reduced government regulatory involvement because of the democratic nature of the cooperatives (Mottet 1980).  Perhaps the greatest argument for cooperative control of this and other fisheries is based on the social framework within which this work was conducted. Many First Nations are currently involved in treaty negotiations over land and resource rights with the Provincial and Federal governments. If coastal First Nations gain total or partial control over nearshore fisheries, then current regulatory methods imposed by the DFO (a federal agency) will likely be insufficient to manage some fisheries. Management will need to be conducted at more localised levels to address the concerns of local First Nations. Cooperative management by local communities with oversight and some regulatory controlfromthe Federal government may be a viable solution. The variable abundance and recruitment of sea urchins across small and large temporal scales combined with the ecological relationships between sea urchins and kelp forest communities also seems to support cooperative management at the local community level. I have two final comments on sea urchin removal that are in complete opposition to the idea of a sustainable sea urchin fishery. The first is most pertinent to the Hesquiat and involves sea otters. The second relates to the potential 'damage' being done by sea urchins throughout BC. My personal qualitative observations of sea otters during this study indicate that sea otters will likely be present along the Hesquiat coastline in the near future (see Watson 1993 for studies of sea otter range expansion on Vancouver Island). In 1997, sea otters were rarely spotted at Matlahaw Point along the northern portion of the Hesquiat coast. By the summer of 1999, sea otters commonly were seen in this area. Dives made between Matlahaw point and Estevan point (see map, Appendix, Figure 1-14) revealed the presence of few urchins and luxuriant growth of Macrocystis integrifolia, Nereocystis luetkeana, and Pterygophora californica. As well, I observed numerous schools of black rockfish. During  164  the last two field seasons (May and August, 1999), I observed an occasional single sea otter near the study area and found some evidence of sea otter predation on red sea urchins (tests with holes punched in them). If sea otters are indeed expanding into Hesquiat territory, then the Hesquiat need to decide what to do with the sea urchin resource. Their options are simple: immediately harvest numerous urchins before otters get to them, manage sea otters to provide for future urchin harvests, or do nothing. The first option is currently illegal, and would need to be developed with the DFO. The second option would require culling the otter population, is currently illegal, and would likely generate public outcry. Doing nothing may provide benefits to other fisheries through increased production within kelp forests as mentioned above. Finally, unregulated harvesting, leading to an ultimate collapse of the sea urchin fishery in BC, may have ecological support. Although not studied here, the loss of kelp due to urchin grazing means lost primary and secondary productivity to many coastal waters. This loss has ramifications for planktonic, detrital and filter-feeding communities. The organisms at these levels are important links in coastal and pelagic food webs. In this study, I have shown that some fish species are benefited by the presence of kelp forest habitats. The loss of primary production along with the structure of kelp forests may have coast-wide implications in other fisheries.  REFERENCES Allee, W. C. 1931. Animal aggregations. University of Chicago Press, Chicago. Andrew, N. L. 1993. Spatial heterogeneity, sea urchin grazing, and habitat structure on reefs in temperate Australia. Ecology 74:292-302. Bernard, F. R., and D. C. Miller. 1973. Preliminary investigation on the red sea urchin resources of British Columbia (Strongylocentrotus franciscanus (Agassiz)). Fisheries Research Board of Canada Technical Report 400:37p. 165  Botsford, L. W., B. D. Smith, and J. F. Quinn. 1994. Bimodality in size distributions: The red sea urchin Strongylocentrotus franciscanus as an example. Ecological Applications 4:42-50. Breen, P. A., D. C. Miller, and B. E. Adkins. 1976. An examination of harvested sea urchin populations in the Tofino area. Fisheries Resources Board of Canada Manuscript Reports 1401:1-23. Breen, P. A., B. E. Adkins, and D. C. Miller. 1978. Recovery rate in three exploited sea urchin populations from 1972 to 1977. Fisheries and Marine Service Manuscript Report (Canada) 1446:1-27. Breen, P. A., W. Carolsfeld, and K. L. Yamanaka. 1985. Social behaviour of juvenile red sea urchins, Strongylocentrotus franciscanus (Agassiz). Journal of Experimental Marine Biology and Ecology 92:45-61. Bureau, D. 2001. Stock Assessment Unit, Pacific Region. Department of Fisheries and Oceans Canada. Personal communication regarding urchin fishing schedules and market preferences based on sea urchin size. Cameron, R. A., and S. C. Schroeter. 1980. Sea urchin recruitment: effect of substrate selection on juvenile distribution. Marine Ecology Progress Series 2:243-247. Campbell, A., J. Boutillier, and J. Rogers. 1999. Discussion on a precautionary approach for management of the red sea urchin fishery in British Columbia. Department of Fisheries and Oceans Canada. Canadian Stock Assessment Secretariat Research Document 99/094. Chan, L. 2001. Hi-To Fisheries. Vancouver, British Columbia. Personal communication regarding sea urchin quality, wholesale pricing, processing, and shipping. Clayoquot Sound Scientific Panel. 1995. First nations' perspectives relating to forest practices standards in Clayoquot Sound. Appendices V and,VI. March, 1995. Victoria, B.C. de Jong-Westman, M., B. E. March, and T. H. Carefoot. 1995. The effect of different nutrient formulations in artificial diets on gonad growth in the sea urchin Strongylocentrotus droebachiensis. Canadian Journal of Zoology 73:1495-1502. Druehl, L.D. 2001. Professor. Simon Fraser University, Burnaby, British Columbia. Personal communication regarding feasibility of feeding urchins in situ with farmed kelp. Duggins, D. O. 1981. Interspecific facilitation in a guild of benthic marine herbivores. Oecologia 48:157-163.  166  Duggins, D. O. 1983. Starfish predation and the creation of mosaic patterns in a kelpdominated community. Ecology 64:1610-1619. Duggins, D. O., C. A. Simenstad, and J. A. Estes. 1989. Magnification of secondary production by kelp detritus in coastal marine ecosystems. Science 245:170-173. Ebert, T. A. 1983. Recruitment in echinoderms. Pages 169-203 in M. Jangoux and J. M. Lawrence, editors. Echinoderm studies 1. A.A. Balkema, Rotterdam. Ebert, T. A. 1998. An analysis of the importance of Allee effects in management of the red sea urchin Strongylocentrotus franciscanus. Pages 619-627 in R. Mooi and M. Telford, editors. Echinoderms: proceedings of the ninth international echinoderm conference. A.A. Balkema, Rotterdam. Estes, J. A., and D. O. Duggins. 1995. Sea otters and kelp forest in Alaska: generality and variation in a community ecological paradigm. Ecological Monographs 65:75-100. Fritzsche, R. A., and T. J. Hassler. 1989. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (Pacific Northwest)~pile perch, striped seaperch, and rubberlip seaperch. Biological Report 82(11.103) U.S. Fish and Wildlife Service. Hallacher, L. E. 1977. Patterns of space and food use by inshore rockfishes (Scorpaenidae: Sebastes) of Carmel Bay, California. Ph.D. dissertation. University of California, Berkeley. Harrold, C , and J. S. Pearse. 1987. The ecological role of echinoderms in kelp forests. Pages 137-233 in M. Jangoux and J. M. Lawrence, editors. Echinoderm studies 2. A.A. Balkema, Rotterdam. Kato, S., and S. C. Schroeter. 1985. Biology of the red sea urchin, Stongylocentrotus franciscanus, and its fishery in California. Marine Fisheries Review 47:1-20. Klinger, T. S., H. L. Hsieh, R. A. Pangallo, C. P. Chen, and J. M. Lawrence. 1986. The effect of temperature on feeding, digestion, and absorption by Lytechinus variegatus (Lamarck) (Echinodermata: Echinoidea). Physiological Zoology 59:332-336. Lawrence, J. M., S. Olave, R. Otaiza, A. L. Lawrence, and E. Bustos. 1997. Enhancement of gonad production in the sea urchin Loxechinus albus in Chile fed extruded feeds. Journal of the World Aqauculture Society 28:91-96. Leighton, D. L., L. G. Jones, and W. J. North. 1966. Ecological relationships between giant kelp and sea urchins in Southern California. Pages 141-153 in E.G. Young and J.L. McLachlan, editors. Proceedings of the fifth international seaweed symposium. Pergamon Press, Oxford.  Leinaas, H. P., and H. Christie. 1996. Effects of removing sea urchins (Strongylocentrotus droebachiensis): stability of the barren state and succession of kelp forest recovery in the east Atlantic. Oecologia 105:524-536. Levitan, D. R., M. A. Sewell, and F. S. Chia. 1992. How distribution and abundance influence fertilization success in the sea urchin Strongylocentrotus franciscanus. Ecology 73:248-254. Low, C. J. 1975. The effect of grouping of Strongylocentrotus franciscanus, the giant red sea urchin, on its population biology. Ph.D. dissertation. University of British Columbia, Vancouver. Mann, K. H. 1977. Destruction of kelp-beds by sea-urchins: a cyclical phenomenon or irreversible degradation? Helgolander Wissenschaftliche Meeresuntersuchungen 30:455-467. McBride, S. C , W. D. Pinnix, J. M. Lawrence, A, L. Lawrence, and T. M. Mulligan. 1997. The effect of temperature on production of gonads by the sea urchin Strongylocentrotusfranciscanusfed natural and prepared diets. Journal of the World Aqauculture Society 24:357-365. Minor, M. A., and R. E. Scheibling. 1997. Effects of food ration and feeding regime on growth and reproduction of the sea urchin Strongylocentrotus droebachiensis. Marine Biology 129:159-167. Mottet, M. G. 1980. Factors leading to the success of Japanese aquaculture, with an emphasis on Northern Japan. Technical report no. 52, State of Washington Department of Fisheries. Paine, R. T., and R. L. Vadas. 1969. The effects of grazing by sea urchins, Strongylocentrotus spp., on benthic algal populations. Limnology and Oceanography 14:710-719. Pfister, C.A., and A. Bradbury. 1996. Harvesting red sea urchins: recent effects and future predictions. Ecological Applications 6:298-310. Rowley, R. J. 1989. Settlement and recruitment of sea urchins (Strongylocentrotus spp.) in a sea-urchin barren ground and a kelp bed: are populations regulated by settlement or post-settlement processes? Marine Biology 100:485-494. Schmitt, R. J., and S. J. Holbrook. 1984. Ontogeny of prey selection by black surfperch Embiotoca jacksoni (Pisces: Embiotocidae): the roles of fish morphology, foraging behavior, and patch selection. Marine Ecology Progress Series 18:225-239.  168  Sloan, N. A., C. P. Lauridsen, and R. M. Harbo. 1987. Recruitment characteristics of the commercially harvested red sea urchin Strongylocentrotus franciscanus in southern British Columbia, Canada. Fisheries Research 5:55-69. Steneck, R. S. 1993. Is herbivore loss more damaging to reefs than hurricanes? Case studies from two Caribbean reef systems (1978 - 1988). Pages 220-226 in R. N. Ginsburg, editor. Proceedings of the colloquium on global aspects of coral reefs: health, hazards and history. Rosenstiel School of Marine and Atmospheric Science, Miami. Tegner, M. J., and P. K. Dayton. 1977. Sea urchin recruitment patterns and implications of commercial fishing. Science 196:324-326. Tegner, M. J. 1981. Population structure, recruitment and mortality of two sea urchins (Stongylocentrotusfranciscanusand S. purpuratus) in a kelp forest. Marine Ecology Progress Series 5:255-268. Vadas, R. L. 1977. Preferential feeding: an optimization strategy in sea urchins. Ecological Monographs 47:337-371. Watson, J. C. 1993. The effects of sea otter (Enhydra lutris) foraging on shallow rocky communities off northwestern Vancouver Island, British Columbia. Ph.D. dissertation. University of California Santa Cruz, Santa Cruz. Woodward, F. N. 1966. The seaweed industry of the future. Pages 55-69 in E. G. Young and J. L. McLaughlan, editors. Proceedings of thefifthinternational seaweed symposium. Pergamon Press, Oxford.  I.  Appendix  \  170  Figure 1-1. Map of Vancouver Island, British Columbia showing the distribution of the Nuu-Chah-Nulth group of First Nations within the dashed line.  Figure 1-2. Study site locations near Hot Springs Cove, Vancouver Island, Canada. Latitude and longitude for Barney Rocks is 49° 20' 44" N, 126° 16' 41" W. Inset shows placement of Hot Springs Cove (HSC with arrow) on Vancouver Island.  Figure 1-3. Diagram of experimental unit sampling design. All sampling components in the diagram were conducted on each of the three sample transects within a study plot. The design is systematic with random starts utilised for the side of the line to begin kelp sampling as well as the initial bearing for transects. At each line urchins were sampled within belt transects that were subdivided into three 4-m sections on the left and right side of each line. Kelp was sampled within 0.1-m quadrats alternating to the left and right of the transect line at each meter point. Treatments were carried out within a 15-m diameter treatment zone with sampling occurring within a 3-m buffer. Drawing not to scale. z  173  Table 1-1. Definitions used for substrate classification. Substrate Type Reef  Definition Continuous rocky substrate greater than 3-m across and imbedded into sand or rock such that true measurement of its size would be impossible.  Boulder  Between 25-cm and 3-m at greatest cross section and imbedded or resting on bottom.  Rock  Less than 25-cm across at greatest cross section.  Shell  Shells (or pieces of shell greater than 2-cm) of marine invertebrates such as large mussels, abalone, urchins.  Sand  Broken shell pieces, fine and coarsely broken rock less than 2-cm at greatest cross section  174  Sample Urchin Removal Calculation November 1997 Study site = 3 Desired treatment = 1 urchin / m  2  Sampled area within plot Number of urchins in sampled area Observed urchin density Ratio treatment density / observed density I* observation in sorted size data corresponding to needed ratio 0.29 = I /175 Removal size to achieve treatment  50.4 m 175 3.47 / m 0.29 2  2  51 8.2 cm  Pre-treatment Sorted size data for study Site 3; treatment plot = 1 urchin / m  2  2.2 3.1 3.1 3.2 3.5 3.9 4.9 5.0 5.1 6.1 6.4 6.5 6.7 6.7 6.7 6.8 6.8 7.0 7.0 7.0 7.1 7.1 7.2 7.2 7.2  7.3 7.3 7.4 7.5 7.5 7.5 7.5 7.6 7.6 7.6 7.7 7.7 7.8 7.8 7.8 7.8 7.9 8.0 8.1 8.1 8.1 8.1 8.1 8.2 8.2  11  8.2 8.2 8.3 8.3 8.3 8.4 8.4 8.4 8.4 8.5 8.6 8.6 8.7 8.7 8.7 8.7 8.7 8.8 8.8 8.8 8.9 8.9 8.9 8.9  8.9 9.0 9.0 9.0 9.0 9.0 9.1 9.1 9.2 9.2 9.2 9.2 9.3 9.3 9.3 9.4 9.4 9.4 9.5 9.5 9.5 9.5 9.5 9.6 9.6  9.6 9.6 9.6 9.7 9.7 9.7 9.7 9.7 9.8 9.8 9.8 9.8 9.9 9.9 9.9 9.9 9.9 10.0 10.0 10.0 10.0 10.1 10.1 10.1 10.1  10.1 10.1 10.1 10.1 10.2 10.2 10.2 10.3 10.3 10.3 10.4 10.4 10.4 10.5 10.5 10.5 10.6 10.6 10.6 10.7 10.7 10.7 10.7 10.7 10.7  10.7 10.7 10.8 10.8 10.9 10.9 11.4 11.4 11.4 11.4 11.5 11.5 11.5 11.6 11.6 11.8 11.9 12.1 12.2 12.2 12.5 12.5 12.7 13.2 13.4  Figure 1-4. Sample calculation of sea urchin removal size necessary to achieve a given treatment level. The highlighted size (8.2 cm) is the minimum size limit for removal that would result in 1 urchin / m remaining after removal. 2  CD T  -  CN  Q  0> 0) 0) T—  « 3  10 ED •tf  < » a.  CO  s  o  B"S  T-  CN T—  CO  %*  0)  o>  cu u  a T~  0. —  Q.  o  41  (0  z  0) 0) 0)  CC!  3  w  cu  O  1. C4  cu  S  »J  .2 ~  T-  3  e TJ  C  cu  Wi  CO  •—  L.  „  CO CN  S S  a  CO  o r(0  0) 0) co ra •cf LL CN  GO  CO  CO L.  cn o  CD CN  a)  "5>  E  oo  0) 0)  "  b CD  CO  T-  co •tf  oo  0) c T  -  a  CO  o CO CN  c75  CO  0) 0)  •cf  Wi  CO  0)  o  •  i—  T-  c  CD T—  CN 1  s  1— 0_> p 00 0)  li  00 ra LL * li.  CO |B101  IOJIUOQ  o o o o o o 2 ° ° o ° o O o O C O t f C N O c O C O t f C N MOT  suo!iBAj9sqo p jeqainN  Uimp3|/M  b  "So cu  *  = s«  A  S ISD O ©  s *  II IS i  b CU  CU  u  g  a  5  C  B  ca </) cu ^  cfl  M  12 =1 u  £  o  CN  « « cu e  0) 0)  co E  .SS cu  CO  i *  D  o o 00 CO  *- s  .2  .—  o  a> E  o  CD  o  o 00  A  C/>  B  o CN  r-  E  00 E T(0  "GO  CO  cu  .tS  03  co  CO  g  H g  o  GO  S  M  s§  T—  £  cs  U  Cfl  cu  « i cu B l- •S M  JS  fZ  U  3  176  w S  2.2  I  oM ,-  E es W  c o cu cu S o>  _o 5  a> —. k, es es  S  s  32 kc es  CU  s S « S 2 2 •* * 5 o  00  E CO  B  »  CD O  g  cu j a  .•2 O  "55  E  a> a> E co  5  O  II •R •o  cs  §  6*  E  TO  b  -•—• 00  cu  o  es  CD  *-*  "GO  -  cu  J—  Q  CA  6  cu . . .  S *  11 es  CM  2  •<•*  o  "> w  CS  CA  _.  es  tS  2  •S  CO  g  II  « *-  •2  . 2 cj T3 cu cn " X  ss  1 9 9 0 0 0 CM O 00 CD |BJ01  0 CM  0 ° , 9 0 0 0 0 0 CM O CO CO T CM IOJJUOO  -, CM  cu  O o o o o O CO CD CM MOT  suoi;BAJ9sqo P jaqwriN  B  uin|p3|/\| cu  E = * i ID <u  .es B  177  C O T—  0> CN 0) T* 0)  4  Q  co E 3 •<r w o C O .—  S  C N 0) i: cn o> ^ 0) CO ^ oo t_ •tf  a  CO  o CO  Wi -  co E 3  •<f CO  o CO 1—  "GO CD  CN 00 0) T— 0) T— CO l_  a w  o CO  Wi  o T-  IBJ01  IOJJUOO  MOT  suoiJBAjesqo jo jaqiunN  O CO CM 0)  S  g  M  g  2 .2 S I cn 8 t/i *-  — 2 •2 S _a g S DO o o CJ .5  <U  cn  •-  s* -o  .5 A  3  es . cn  CS  e s CUcn c u o -a U  e +• cn DO fen  cn ^  aj y *>  r-  CM 0) 0) I— CO E T—  ^ CM  S  CU  cS 2  CO  CO  es  £ §  CM N0) 0) T— CO CO LL  O (0 CM 0)  i- «J  IS  CO  CO  —  >- cs cs  O CM C 0) T— 0) T— CO C  o  CN  a  CO  c75  |[ja  u  I- CS  <" -2  CO C M 0) T~ 0) 1— T  E ro Q  CU  .ti o  CD  a>  w -s  cs P  o  X  CM  a o *> »  g  CN oo T— 0) CO CO LL  <n  ^  co c  CO  o  •s e  «s  T—  -•—*  •a t  CM 0) o> 0)  o  oo E ro  ..a ja  •cr CO  CU  .H -  B V 0>  1£ cu  c  CM  i  ainipaLM  vi  .es  .go «  b« s a  178  d  o  CO CO  d  T  ~  (none)  LO  d  (N/A)  (N/A)  CT)  d  s  LO  CO  d  1^  d  CD  d  (9.1)  (none)  (N/A)  CO  00  CM  d  d  CO  (none)  (10.6)  (N/A)  (N/A)  (none)  (N/A)  CM  CT) o  CT)  CT)  d  d  (N/A)  d  (none)  (none)  (N/A)  0 0  T-  LO  CM  T-  CM  d  co  d  d  d  d  d  CD  d  (N/A;  (7.3)  (none)  s  d  (N/A)  I-  o  o  1^  CD CM  E co  (N/A;  LO csi  o  o  T-  CM  rsi  T _  CN  CT) o CN  Is  CM  csi  CM  5  •ct  CO  (4.8)  r--  CT>  (8.7)  CD  (9.6)  .a j» g  T—  d  T-  s  "EL  •B a| l a .2 s .2 a £  LO  I--  (12.1  .S ^ cu .5 *a B  d  (N/A;  S |  Ico  d  o  (none)  Is  d  cu  CO CO  (none)  LO  CM'  (N/A)  (N/A)  c  £ =  CO  CM  (none)  s  LO  d  (8.0)  I-;  CT) 00  E cu  CT)  CO  d  (none)  o o  >o  (N/A)  CO  x—  co  (N/A)  CT) CL  o d  (N/A)  00  LO CO  (N/A;  E  (N/A)  ** o  CD  d  CO  (N/A)  « acu  d  CO  d  (N/A)  B  cn © • ^* >• -e •a O *S  M  o  (N/A]  B  d  :V/N)  O  •<*  (8.8)  E co  cu  (V/N)  cu  CT)  (8.0)  CT) 00  (V/N)  TO  T—  d  (N/A)  (N/A)  d  s  I §  (N/A)  (N/A)  (N/A)  d  (none)  O  cu cs CS 3 cs . cs £ cu o\ > ~ ^  i--  d  cu  •"^ cu JS lu •*  Vi  (none)  (N/A)  (none)  h~ c\i  o  s  • « « a" i- >  LO  (none)  ^  cu  3  — cs  CM  (none)  (9.5)  d  (none)  1-  O  (none)  CO  CM  (none)  CD  C  cu i- cu s  s  IO  d  CT)  S .-a  <u cn - J"  _ JH »> "O cu cs  51  (N/A)  oo  -F cu  a. £  S  00  d  H  •a cs .2 »> C >33 cu CU  LO  d  (N/A)  8 -o£ »• §  00  O  (N/A)  O B  CM  (N/A)  cn |  5 8  J3  O  (N/A)  £ S  cs  o d  C0  (N/A)  <u «  B  cu CU '£ cu ~  cu  CD  (V/N)  CT) CT)  2 cu en £  u  o d  O  d  S3  °a  cu  LO  d  (none)  I-  CO  (none)  o  d  O  (none)  o  o d  (none)  oq  o  O  (V/N)  Q_ CO  00  (9.4)  co  o  CT)  sI 8I a>  (N/A)  o  o d  (8.6)  c3  o d  (none)  _:  >> o .2  o  (N/A)  E co  ^ cn cn  (N/A)  01  •S **  CT) CT) o  T—  CO CO  co CM  LO CO  Is  CO  CD  CO  LO  CM  CM  5  oo  00  LO  CD  CM  CM  73 M ~u « >  M  4)  O  3  A  E  ~  E  7^  E  PC xi ^  T3  c o O  TJ CD  c o O  T3  2 a S 2 S cu 5 cu  •« S * S a •a 8 •2 I * a  o O  cu £  •s f °  -  J3  cu  C7\ ON  H  CU  "cs  s  B  2  B  ca- cu cu  o  CD  CM  o I-  o  CM  co  CM  CM  5 o  ro  •4—«  O t—  CD  CO  CL  O CS  a©  ^ xS - «  s s CS  B  £  o  3  J5  >  Cl) CO  CM  CM  CO  co  co  co  3  (Zl  179  Site 1  N  Figure 1-6. Plot layout for Site 1 indicating treatment allocation, approximate depth at plot centre and proportion of availability for the substrate categories. Plots measure 30 m in diameter. Layout not to scale. For substrate category definitions see Table 1-1.  Site 2  Figure 1-7. Plot layout for Site 2 indicating treatment allocation, approximate depth at plot centre and proportion of availability for the substrate categories. Plots measure 30 m in diameter. Layout not to scale. For substrate category definitions see Table 1-1.  Site 3  Figure 1-8. Plot layout for Site 2 indicating treatment allocation, approximate depth at plot centre and proportion of availability for the substrate categories. Plots measure 30 m in diameter. Layout not to scale. For substrate category definitions see Table 1-1.  182  Costaria costata Log Weight = 0.69 + 0.015 * Length Regression: R = 0.83; p < 0.001  0.6 I 0  • 20  • 40  • 60  • 80  • 100  :  1  120  Total Length (cm)  Figure 1-9. Wet weight as a function of total length for Costaria costata. Weights were logio transformed to linearise the exponential relationship between total length and weight. Length measured to nearest centimetre by use of a ruler and weight measured to nearest tenth of a gram by use of a spring scale. Solid line represents the least-squares regression and dotted lines are the 95% confidence intervals for the regression (n = 21). Specimens collected in August and September 1999.  183  Laminaria bongardiana Log Weight = -2.14 + 2.52 * Log Length Regression: R = 0.94; p < 0.001 2.5 ,  -0.2  .  .  .  0.2  0.6  1.0  1.4  1.8  Log Stipe Length (cm)  Figure I-10. Wet weight as a function of stipe length for Laminaria bongardiana. Lengths and weights were log transformed to linearize the relationship between stipe length and weight. Length measured to nearest 0.5-centimeter by use of a ruler and weight measured to nearest tenth of a gram by use of a spring scale. Solid line represents the least-squares regression and dotted lines are the 95% confidence intervals for the regression (n = 326). Specimens collected in August and September 1999. 10  Pterygophora californica Log Weight = -2.300 + 2.7113 * Log Length Regression: R = 0.87; p < 0.001  0.4  0.6  0.8  1.0  1.2  1.4  1.6  1.8  2.0  Log Stipe Length (cm)  Figure 1-11. Wet weight as a function of stipe length for Pterygophora californica. Lengths and weights were logio transformed to linearize the relationship between stipe length and weight. Length measured (stipe base to meristem band) to nearest 0.5-centimeter by use of a ruler and weight estimated to nearest tenth of a gram by use of a spring scale. Solid line represents the least-squares regression and dotted lines are the 95% confidence intervals for the regression (n = 118). Specimens collected in August and September 1999.  Nereocystis leutkeana Log Weight = -1.48 + 1.45 * Log Length Regression: R = 0.82; p < 0.001  Log Stipe Length (cm)  Figure 1-12. Wet weight as a function of stipe length for Nereocystis luetkeana. Lengths and weights were logio transformed to linearize the relationship between stipe length and weight. Length measured to nearest 1.0-centimeter by use of a ruler and weight measured to nearest 10-grams by use of a spring scale. Solid line represents the least-squares regression and dotted lines are the 95% confidence intervals for the regression (n = 85). Specimens collected in August and September 1999.  Macrocystis integrifolia Weight = 8.78+ 1.21 * Length Regression: R = 0.88; p < 0.001 1600  r  -200 t—^ -100  .  .  .  •  •  •  •  •  1  100  300  500  700  900  1100  Individual Frond Length (cm)  Figure 1-13. Wet weight as a function of frond length for Macrocystis integrifolia. Length measured to nearest 1.0-centimeter by use of a ruler (to stipe terminus) and weight measured to nearest 10-grams by use of a spring scale. Solid line represents the least-squares regression and dotted lines are the 95% confidence intervals for the regression (n = 39). Specimens collected in August and September 1999.  Table 1-3. Sampling seasons and dates samples were collected during the study.  Sampling Season  Season Code  Actual Sample Dates  Summer 1997  Sum 97  27 July 1997  8 August 1997  Fall 1997  Fall 97  21 October 1997  -  2 December 1997  Winter 1998  Win 98  11 February 1998  -  20 March 1998  Spring 1998  Spr 98  31 May 1998  9 June 1998  Summer 1998  Sum 98  23 August 1998  2 September 1998  Fall 1999  Fall 99  27 November 1998  -  15 December 1998  Winter 1999  Win 99  26 February 1999  -  10 March 1999  Spring 1999  Spr 99  22 May 1999  31 May 1999  Summer 1999  Sum 99  24 Aug 1999  9 September 1999  188  s cu  s* * o H  _S  ccj JB cn  1  oo  co co  «< ^  . _  i  a  3  cu  'S  '—.  CN  CN  +1 in  vo CO vd " " +1 +1 ^ 1 ro ro +  ^  l  CO  +1 Ov  oo o  • vo So O +1 +1  cd 'C  K  •2 x g cu o J3  oo cu — co" O oo  a  cu u ^ cd fi  «  +1  +1  S  co'  °°. ©  oo  +1 co ©  00 +1 +1 ON vo 2 ,  -  © 6 £© m  in  vo  CN CN  'Cl-  CN  io  +1  ON ON 00  +1 +1 VO ON vo oo w CO © VO CN  *>  _S  o cj  S  £2 ca cn cu X = cu  oo -S  VO  -S <n  .2 <= S o 22 ->  a  bo cu  R  I  u  53 03  2  •  © B 73 u. 5 .2 O cn u . 1 , 0 ) i . +i a  CN  © +1 <n CN  CN CN  m  CO  co CN +1  +1 CN CO +1 ©  vd © t-  Ov  CN +1 +1 ON +1 ON +1 CN ON vd VO  CO  ON in CN © +1 +1 OO <n o ©  ©  +1 —  vq CO © CN  ©  VO"  vo  ©  oo © vo vo  ©  +1 +1 00  © m  r+1  •n  +1 •<t  +1 —  co" ° ° . vo  ©§ +1 CO +1 © vq  +1  © oo  w  CN  CO t-^  ©  CN^  CN  *->. CO t-~ <n  5\ o  Ov +1 +1 +1 ON +1 ON CO 0\ CO -—' +i  VO © +1  3086.6  II  >> B x> w  in  © +1 +1  2376.5  •5  © ©  V O  /—N  r-~ vo  0? <1 0 ? • ON —' © +1 r -  CN /—s r-. co CN ^—s vq oo © 00 c~ 00 CN  CN +1 © ON  1  co iri  -a©'  vo  2 *CO  s  CO CN ON CO ^ o +1 +1 +, +1 +1 vq — I co "H ro  ©  © 2 © s  CO  •2 E  ^  -n" c §  w  CN CO  §  cjv +1 vo 00  CO r—N in CO © in + 1 ^ + 1 ^ .. +1 +1 +1 +1 r © CO Tj 00 +1 ON +1 * — i +1 © © 00 CO © vd CO © >n °°. co </•; —• •^-^ co vo oo w ON Ov CN VO OO in —i uo CN CN  ^  w  co  o >o  W  o  1359.  cn  CN — +1 +1 UO  in ° • CO  1041.  c  • o  1910.  cu X CU cu  cs  1907.  2  a *  il>  4417.  co  2016.  cu cu - r  v/3 «  <~.  ° - CN CO •<t m +1 +1 CO VO © '  —l  ^ +1 ^ —  ©  © cs  o  ^  m  <N +1 +| 00 co  oo vd  CN  +1  +1 CO  +1  ^  £ • =i © ^.  ™  +  ©  in  1  CN  +  L  co  +1  >n + L  CO  00 in  CN  2 aa  CN  ,— © CN ,—v CN CO OO 00 CN © © CN ON +1 +1 +1 +1 +1 © co CN 00 00 © ' ^—' CN © in ON ON  CO ©'  +1 +1 CN co  CQ —. H « » B •» • es 0 , 5 3  r-^ © CN  ^  » -s »-  8§ cn •— B cn S  2 53 XI cu W u  a a ft  0 cu X ffl 5 cu ft  B  o H  6 g  •a  O  o  8 O  o  a  o H  o  T3 CJ  o -J  I  o U  3  o H  T3 CJ  o H-1  o  O  J» A p  1 £* «  ON  ft  H  H  s  —  CN  189  Matlahaw Point  Figure 1-14. General location of dives to assess the impacts of range expansion of sea otters in summer 1999. Inset shows location of the Hesquiat Peninsula (HP) in relation to Vancouver Island, BC.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0090721/manifest

Comment

Related Items