Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Seagrass epifaunal communities of Barkley Sound : epifaunal diversity varies across small spatial and… Whippo, Ross Douglas Byron 2013

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

Item Metadata

Download

Media
24-ubc_2014_spring_whippo_ross.pdf [ 1.11MB ]
Metadata
JSON: 24-1.0166841.json
JSON-LD: 24-1.0166841-ld.json
RDF/XML (Pretty): 24-1.0166841-rdf.xml
RDF/JSON: 24-1.0166841-rdf.json
Turtle: 24-1.0166841-turtle.txt
N-Triples: 24-1.0166841-rdf-ntriples.txt
Original Record: 24-1.0166841-source.json
Full Text
24-1.0166841-fulltext.txt
Citation
24-1.0166841.ris

Full Text

Seagrass Epifaunal Communities ofBarkley SoundEpifaunal diversity varies across small spatial andtemporal scalesbyRoss Douglas Byron WhippoBachelor of Science, The University of Washington, 2011Associate of Arts, Seattle Central Community College, 2009A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinThe Faculty of Graduate Studies(Zoology)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)December 2013? Ross Douglas Byron Whippo 2013AbstractThe assembly and persistence of ecological communities is a phenomenonthat occurs across large spatial and temporal scales. However, the relativeeffects of regional versus local processes on community structure are not wellunderstood in marine ecosystems. In order to understand how scale can alterprocesses that drive variation in community assembly it is necessary to de-termine patterns of diversity across multiple scales. Here, I used invertebrateepifaunal communities in the foundation species Zostera marina to test 1)whether this marine community exhibits meadow-scale variability throughtime, and 2) whether we can identify patterns of connectivity and diver-sity within and among meadows in the same region. I found that seagrassepifaunal communities are variable in terms of their rarefied richness, alphaand beta diversity, and evenness among meadows. In addition, differencesin these metrics were detected over the course of a summer season.iiPrefaceThis work was funded by a research scholarship from the Bamfield MarineSciences Centre, generous donations to the SciFund Challenge by PacificRegion Security Corporation and others, and an NSERC Discovery Grantawarded to Mary I. O?Connor. Approval for portions of this work to becarried out in Huu-ay-aht First Nations land was granted by the Huu-ay-ahtNation: permit # HFN-102-12. Animal collections and care was approvedby the Department of Fisheries and Oceans Canada: permit # XR 98 2012& XR 43 2013; the UBC Animal Care Committee: certificate # 4990-11;and the Bamfield Marine Sciences Centre Animal Care Committee: permit# RS-12-12 & RS-13-04. Collection of seagrass was approved by AgricultureCanada: permit # MP13-LOA1.iiiTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiiAcknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . xiDedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Seagrass Meadow Communities . . . . . . . . . . . . . . . . . 21.2 Structure & Objectives . . . . . . . . . . . . . . . . . . . . . 32 Quantifying the spatial scale of epifaunal diversity acrossseagrass meadows in Barkley Sound, BC . . . . . . . . . . . 42.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 42.2 Materials & Methods . . . . . . . . . . . . . . . . . . . . . . 82.2.1 Study System & Field Sites . . . . . . . . . . . . . . 82.2.2 Epifaunal Sampling . . . . . . . . . . . . . . . . . . . 102.2.3 Quantifying Abiotic Conditions, Seagrass Density, &Productivity . . . . . . . . . . . . . . . . . . . . . . . 112.2.4 Statistical Analyses . . . . . . . . . . . . . . . . . . . 132.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16ivTable of Contents2.3.1 Comparison of Biotic & Abiotic Factors Among Sites 162.3.2 Quantifying Epifaunal Abundance/Diversity In & Be-tween Seagrass Meadows . . . . . . . . . . . . . . . . 202.3.3 Quantifying & Comparing Epifaunal Communities AmongMeadows . . . . . . . . . . . . . . . . . . . . . . . . . 232.3.4 Quantifying Relationships Between Structural Param-eters & Epifaunal Communities . . . . . . . . . . . . 372.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393 General Conclusions . . . . . . . . . . . . . . . . . . . . . . . . 453.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453.2 Recommendations For Future Research . . . . . . . . . . . . 453.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47AppendicesA First Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . 54B Second Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . 68B.1 Diversity & Evenness . . . . . . . . . . . . . . . . . . . . . . 68B.2 Community Composition . . . . . . . . . . . . . . . . . . . . 74vList of Tables2.1 ANODE-Chi of mean temperatures across all sites from Mayto August . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162.2 ANODE-Chi of mean salinities across all sites from May toAugust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162.3 ANOVA of seagrass density in May . . . . . . . . . . . . . . . 182.4 ANOVA of seagrass density in August . . . . . . . . . . . . . 182.5 ANOVA of LAI in May . . . . . . . . . . . . . . . . . . . . . 192.6 ANOVA of LAI in August . . . . . . . . . . . . . . . . . . . . 202.7 ANOVA of species richness comparing ASU inside (in) versusoutside (out) seagrass meadows . . . . . . . . . . . . . . . . . 212.8 Presence (X) and absence (-) of species identified in ASUsinside (IN) and outside (OUT) seagrass meadows, July 2013. 222.9 ANODE-Chi of organismal abundance by site and time. . . . 232.10 List of organisms enumerated grouped by morpho-species. . . 272.11 ANODE-Chi of species richness by site and time. . . . . . . . 312.12 ANOVA of rarefied species richness by site and time. . . . . . 312.13 ANODE-Chi of species diversity measured as Simpson?s Index. 322.14 ANODE-Chi of species evenness by site and time. . . . . . . . 332.15 PERMANOVA of community composition. . . . . . . . . . . 362.16 Akaike Information Criterion for density models. . . . . . . . 372.17 Akaike Information Criterion for LAI models. . . . . . . . . . 38A.1 Similarity percentage of aggregated species composition forall sites and times. . . . . . . . . . . . . . . . . . . . . . . . . 55B.1 ANODE-Chi of raw species richness by site and time. . . . . 68viList of TablesB.2 ANOVA of rarefied species richness by site and time. . . . . . 68B.3 ANODE-Chi of raw species evenness by site and time. . . . . 68B.4 Similarity percentage analysis of raw species data with pair-wise comparisons between sites and times. . . . . . . . . . . . 75B.5 PERMANOVA of raw species composition data. . . . . . . . 92viiList of Figures2.1 Map of Trevor Channel in Barkley Sound, BC. Primary sam-pling sites are indicated by red circles; Dodger Channel (DC),Wizard Island (WI), Robber?s Passage (RP), Numukamis Bay(NB), Crickitt Bay (CB). Additional artificial seagrass de-ployment sites are indicated with green stars; Eagle Bay (EB),Bulldozer Beach (BB), Clifton Point (CL). . . . . . . . . . . . 92.2 Sampling regime at each meadow. Inset depicts horizontalsection of seagrass sampled in each quadrat. . . . . . . . . . 112.3 Average water temperature and salinity from May to August,2012. Temperatures and salinities were different among siteswith pairwise differences between sites (Tests for General Lin-ear Hypotheses, p < 0.05). . . . . . . . . . . . . . . . . . . . . 172.4 Shoot density in May (A) and August (B) in 0.28m2 quadratsfor each sampling site. Bars represent standard error. . . . . 192.5 Leaf area index in May (A) and August (B) in 0.28m2 quadratsfor each sampling site. Bars represent standard error. . . . . 202.6 Mean richness of ASUs placed inside (in) and outside (out)seagrass meadows in July 2013. . . . . . . . . . . . . . . . . . 212.7 Log abundance of all organisms for each site in May and Au-gust, 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242.8 Rank abundance curves of most abundant epifauna for eachsite in May, 2012. . . . . . . . . . . . . . . . . . . . . . . . . . 252.9 Rank abundance curves of most abundant epifauna for eachsite in August, 2012. . . . . . . . . . . . . . . . . . . . . . . . 26viiiList of Figures2.10 Mean raw richness (A) and rarefied richness (B) across allsites from May and August 2012. Pairwise comparisons weremade within each time among sites (Raw richness: Testsfor General Linear Hypotheses, p < 0.05, Rarefied richness:Tukey, p < 0.05). . . . . . . . . . . . . . . . . . . . . . . . . . 302.11 Mean diversity per 0.28 m2 plot across all sites in May andAugust 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . 322.12 Mean community evenness per 0.28 m2 plot of all sites in Mayand August. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332.13 Beta diversity as Bray-Curtis distance for each site in May (A)and August (B). Red points represent a multidimensional cen-troid for each site, blue vectors are distance of sampled 0.28m2 plots within site to the centroid. Axes are dimensionless. . 342.14 Mean distance to centroid, Bray-Curtis beta diversity of allsites in May (A) and August (B). Differences in variationamong sites within each time detected with post-hoc Tukey(p < 0.05). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352.15 nMDS of epifaunal communities across all sites in May andAugust. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362.16 Organismal abundance as a function of shoot density in 0.28m2 plots. 372.17 Organismal abundance as a function of leaf area in 0.28m2 plots. 38B.1 Mean richness of raw species data (A) and rarefied richnessof all raw species (B). All statistical values shown are an in-teraction between site and time. Pairwise comparisons weremade within each time among sites (Tukey and General Lin-ear Hypotheses, p < 0.05) . . . . . . . . . . . . . . . . . . . . 69B.2 Simpson?s Index of raw species data diversity across all sitesin May and August 2012. All pairwise comparisons are withintime period among sites (General Linear Hypotheses, p < 0.05). 70ixList of FiguresB.3 Average community evenness, Simpson?s Evenness measure ofall sites in May and August with raw species data. All pair-wise comparisons are within time period among sites (GeneralLinear Hypotheses, p < 0.05). . . . . . . . . . . . . . . . . . . 71B.4 Beta diversity as Bray-Curtis distance for each site and timewith raw species data. . . . . . . . . . . . . . . . . . . . . . . 72B.5 Mean distance to centroid, Bray-Curtis beta diversity of allsites in May (A) and August (B). Differences in variationamong sites within each time detected with post-hoc Tukey(p < 0.05). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73B.6 nMDS of raw species data epifaunal communities across allsites in May and August. . . . . . . . . . . . . . . . . . . . . 93xAcknowledgementsFirst and foremost I would like to thank my advisor Mary I. O?Connor forguidance in developing my research program, endless patience, intellectualsupport, and impromtu thesis meetings that got me this far. I also thankmy committee, Chris Harley and Greg Crutsinger for valuable feedback. Forassistance in designing a sampling protocol, Nate Sanders provided manyhelpful suggestions and resources. For help in the lab and the field I owea great debt of gratitude to: Carolyn Prentice, Matthew Siegle, SamanthaJames, Nicole Knight, John Cristiani, Frances Ratcliffe, Siobhan Gray, KatAnderson, Suz Anthony, Danielle de Jonge, and the Robles Lab. For reviewof drafts of this manuscript and other related writings, Joey Bernhardt andNatalie Caulk. For apples, chocolate, bad jokes, and moral support I thankNatalie Stafl. For being an amazing lab manager I thank Winnie Cheung.For sage advice on manuscript preparation Ange?lica Gonza?lez. For assis-tance with various statistical analyses: Andrew MacDonald, Kyle Demes,Matthew Barbour, and Bill Harrower. For keeping me on track with alldeadlines and paperwork, Alice Liou. For local knowledge of seagrass mead-ows in Barkley Sound and various field techniques, Ramona DeGraff. Andfinally, for endless assistance with all aspects of field work, lab facilities, andadministrative support, I thank the amazing staff of the Bamfield MarineSciences Centre.xiDedicationFor my grandfather, Harold Eugene Whippo.The fishermen know that the sea is dangerous and the stormterrible, but they have never found these dangers sufficient reasonfor remaining ashore.-Vincent Van GoghxiiChapter 1IntroductionHow biological communities assemble and persist remain fundamental ques-tions of ecology. These phenomena are driven by local processes such aspredation, competition and facilitation (e.g., by foundation species), as wellas regional processes of colonization, extinction and dispersal. Togetherthese processes interact to influence the diversity and composition of localcommunities, but their relative importance can vary among ecosystems andabiotic conditions (Leibold et al., 2004). The relative importance of localand regional processes determines the predominant spatial and temporalscales of community structuring, and therefore is critical to understandingthe resultant communities we observe.Marine communities of foundation species and mobile marine animals areexpected to exhibit a balance of local and regional control on species richnessand diversity. For example, in a global analysis Witman et al. (2004) foundthat while up to 76% of local diversity in subtidal invertebrate communitiescould be explained by the regional species pool overall, much of the variationthat was observed at some sites was due to competitive dominance andpredation. In addition, Bruno (2002) observed an interaction between localfacilitation and regional colonization in determining diversity of cobble beachcommunities. These patterns have also been seen in coral reef habitatswhere strong regional enrichment of the local species pool interacts withspatial heterogeneity and species coexistence mechanisms to produce therealized community (Karlson et al., 2004). The relative role of local andregional control of animal communities in seagrass habitats, however, hasnot been addressed directly in the field. Understanding the spatial scales atwhich critical processes operate within seagrass-associated communities isvital, especially in light of the global decline of seagrass species (Short and11.1. Seagrass Meadow CommunitiesWyllie-Echeverria, 1996).1.1 Seagrass Meadow CommunitiesSeagrass meadows are highly productive ecosystems that support a broad di-versity of taxa including invertebrates, fish, marine mammals, and birds (Hem-minga and Duarte, 2000). The invertebrate communities consist of primarilysub-sediment infauna, mobile benthic organisms, and mobile and sessile epi-fauna. The epifaunal component of seagrass meadows include many taxa,but most commonly Crustacea and Gastropoda, and to a lesser extent variousmembers of Echinodermata and Polychaeta, among others (Jaschinski andSommer, 2008; Vonk et al., 2010). Epifaunal communities in seagrass arethe main link between meadow primary production, and higher trophic lev-els (Best and Stachowicz, 2012). This transfer is largely the result of grazingon epiphtyic algae which opportunistically grow on the seagrass blades (Cooket al., 2011). The control of these algae is affected by top-down forcing ofthe mesograzing epifauna, though bottom-up control is also seen to someextent (Douglass et al., 2007).Seagrass net community production including primary and secondaryproduction is between 20 and 101 Tg C per year, globally, and nutrientcycling, including trophic transfer by epifauna, is a critical step in thecycling and sequestration of carbon (Hemminga and Duarte, 2000; Duffy,2006). The ecosystem functions that epifauna facilitate in seagrass meadowssupport community processes that deliver vital ecosystem services to hu-mans such as sediment stabilization and wave attenuation (Costanza et al.,1997; Orth et al., 2006). Seagrass meadows also provide shelter and feedinggrounds for economically valuable taxa including rockfish and perch, inverte-brates such as dungeness crab, and burrowing infauna including cockels andlongneck clams (Baldwin and Lovvorn, 1994; Best and Stachowicz, 2012).Globally, seagrass is in decline due to multiple stressors including nu-trient inputs (Orth et al., 2006), physical disturbance (Short and Wyllie-Echeverria, 1996), and invasive species (Malyshev and Quijo?n, 2011). Un-derstanding the variables that collectively determine seagrass community21.2. Structure & Objectivesassembly and persistence is necessary to make informed management deci-sions (Cullen-Unsworth and Unsworth, 2013). While a wealth of ecologicalinformation has been drawn from seagrass ecology studies in the Mediter-ranean, the east coast of North America, and Australia, relatively few studieshave focused on the community ecology of seagrass meadows in the PacificNorthwest. The majority of meadows in the Pacific Northwest are com-prised of the seagrass Zostera marina, a temperate seagrass that occurs asboth as large, contiguous meadows, and as naturally fragmented patchesin coastal regions. By understanding the scale at which biotic and abioticforces structure diversity within and among communities, we will be betterable to successfully manage these systems locally.1.2 Structure & ObjectivesThe goal of this thesis is to describe the composition and diversity of epi-faunal mesograzers in Pacific Northwest seagrass meadows, and to identifypatterns of diversity within and among meadows. This thesis consists ofthree chapters: an introduction, a data chapter, and a conclusions chapter.In Chapter 2 I quantified the community composition, richness, and diver-sity of epifaunal mesograzers, and tested the dispersal abilities of epifauna toanswer the following questions: 1) Do seagrass epifauna exhibit patch-scalevariability across multiple scales? 2) Can we identify patterns of diversitywithin and among meadows in the same region? Chapter 3 concludes thisthesis by placing my findings in the context of seagrass ecosystem functionand contributes to our knowledge of marine community structure. Finally,I suggest future work and additional questions that this body of work couldpotentially inform, as well as implications for managment strategies bestsuited to conserve seagrass ecosystems.3Chapter 2Quantifying the spatial scaleof epifaunal diversity acrossseagrass meadows in BarkleySound, BC2.1 IntroductionThe assembly and persistence of ecological communities have historicallybeen viewed as acting on a local scale (biotic interactions, fine-scale spatialstructuring, etc.) or a regional scale (dispersal, landscape gradients, etc.),but the interaction of processes across these scales is now seen as funda-mental to the resulting community composition (Leibold et al., 2004). Tounderstand patterns of local diversity, we must know regional patterns ofdiversity (Caley and Schluter, 1997). Regional processes such as dispersaland abiotic filtering across a gradient, and local processes such as facilitationcan interact to produce a realized community that could not be predictedby considering these scales separately (Witman et al., 2004). Therefore, it isimportant to consider how measures of community diversity can vary fromthe scale of a single patch, to multiple patches within a region, which maybe unique in habitat structure and cross landscape gradients.By considering within and between patch variation we can better deter-mine the scale at which local and regional community structuring processesare acting. The effect of these processes on realized communities can re-sult in very different community outcomes depending on timing and spatial42.1. Introductionextent of the structuring force.The role of scale in ecology has been tested primarily in terrestrial, freshwater, and pelagic habitats, with fewer studies describing the impact ofprocesses across scales on near-shore marine communities (Cottenie, 2005;Logue et al., 2011). However, the life history and dispersal dynamics ofmany marine macroinvertebrates can fundamentally differ from those offreshwater and pelagic organisms, suggesting different relationships betweenlocal and regional processes compared to other systems. For example, manynear-shore marine organisms exhibit a mix of passive and active dispersalstrategies. The combination of these strategies in a community would sug-gest different community assembly predictions than if considering only thepassive dispersal observed in planktonic organisms, or the active dispersalobserved in some aquatic insects.In addition, temporal variability of invertebrate assemblages in marinesystems has been previously underestimated (Balata et al., 2006). As such,the timing of dispersal may play a key role in community assembly. Forinstance, a planktonic larval stage may allow early colonization of a habitatby particular organisms, but later dispersal by adults of another species mayallow colonization of habitat patches if the adults are competitively superiorto the dominant larval taxa. Here, priority effects may be stronger or weakerin structuring a community depending on the phenology and life-history ofthe dispersing species.The presence of abiotic gradients in marine communities may also explainvariation in community assembly across scales. Salinity and temperaturehave been shown to be strong structuring forces in the marine environmentthat can result in species sorting along a gradient (Barnes and Ellwood, 2012;Fortes et al., 2014; Leibold et al., 2004). Additionally, diversity in marinehabitats such as the rocky intertidal has been observed to be greater withinlocations than between locations (de Juan and Hewitt, 2011). But the role ofregional gradients can alter predictions of community diversity and richnessdepending on whether those gradients are the primary structuring force, orinteract with local conditions to make a habitat more or less suitable forcolonizing organisms (Sanders et al., 2007b; Witman et al., 2008).52.1. IntroductionSeagrass meadows are spatially structured, globally distributed near-shore habitat. They are highly productive ecosystems that support a broaddiversity of taxa, and are subject to various abiotic gradients and regimesacross spatial and temporal scales (Bostro?m et al., 2006a; Heck Jr. et al.,2003). Much of the ecological knowledge tested in and applied to seagrassanimal communities has been drawn from a narrow set of climate regimesand regions: up to 75% of ecological seagrass studies have been restrictedto between 30? and 40? latitude in Australia, the Mediterranean, and theAtlantic regions, representing only a small portion of global seagrass distri-bution (Bostro?m et al., 2006a; Duarte, 1999). Temperate seagrass meadowsare subject to greater seasonal temperature fluctuations, tidal regimes, andstrong seasonal drivers, such as limited sunlight and prolonged precipita-tion events, that could change the relative contribution of local and regionalprocesses to structuring animal communities in seagrass meadows.Seagrass is valuable habitat for many marine invertebrates and hostsa large number of epifaunal species that constitute a large proportion ofsecondary production (Hemminga and Duarte, 2000). These epifaunal com-munities constitute the main trophic link between primary production andhigher trophic levels in temperate seagrass (Hemminga and Duarte, 2000).They consist of macroinvertebrates including many arthropods and mol-luscs that have varying dispersal capabilities. Many of the epifauna knownto commonly inhabit temperate seagrass meadows have low active dispersalcapability in larval and juvenille stages. For example, the zebra sea hare(Phyllaplysia taylori), many amphipod species including caprellids (Am-phipoda:Caprellidae) and some gammarids (Amphipoda:Gammaridea), andthe seagrass isopod (Idotea resecata) have crawl-away larvae and broodedoffspring respectively that do not undergo a planktonic stage. Passive disper-sal via rafting on macrophyte detritus and phoresy (non-parasitic dispersalby an animal vector) are the most likely dispersal modes for many of thesetaxa, though the relative strength of each mode is not known.Variation in community response to the abiotic environment (i.e. salin-ity, temperature, etc.) within and among seagrass meadows even in a singleregion can be very large (Bostro?m et al., 2006a). This suggests that current62.1. Introductionstructure of the community might be based on abiotic filtering, and the out-come of community dynamics (persistence/coexistence or local extinctionof species) modulated by local abiotic conditions within meadows. Some ofthese differences could be due to variable spatial complexity and permeabil-ity of the inter-meadow matrix. This could act as a community structuringprocess among meadows, resulting in effectively isolated communities in eachpatch. Alternatively, regional seagrass meadows could operate as a meta-community, and the differences observed among communities could reflectasynchronous dynamics, stochastic extictions/colonizations, or interactionsbetween local and regional processes. Components of the abiotic environ-ment including temperature/salinity and meadow density and leaf area canalso have direct effects on invertebrates, though these relationships can betaxon-specific (Richmond and Woodin, 1999; Bostro?m and Bonsdorff, 2000;Sirota and Hovel, 2006). Indirect effects on epifauna driven by abiotic ef-fects on primary production can also contribute to complexity in communityassembly processes (Frankovich and Zieman, 2005; Jaschinski and Sommer,2008).Despite evidence of spatial structuring, abiotic variation among mead-ows, and moderate dispersal by epifauna, no study has explicitly tested thespatial scale of community structure in seagrass systems. Therefore, to un-derstand the processes and scale that determine community assembly, wemust quantify patterns of diversity within and among seagrass meadows tobetter formulate empirical tests that can reveal underlying processes struc-turing these communities. Here, I quantified the epifaunal diversity andcomposition of five seagrass meadows in Barkley Sound, British Columbia,to determine if seagrass epifaunal communities exhibit patch-scale variabil-ity across multiple scales. Additionally, I compared communities throughspace and time to identify patterns of diversity within and among meadowsin the same region. If passive dispersers are not seen at all sites, or relativeabundance of each species is not the same across sites, we might expect thateither metapopulation dynamics are structuring the community, or post-colonization processes may be determining community assembly. If epifau-nal seagrass communities are simply unstructured metapopulations of many72.2. Materials & Methodstaxa, they should be randomly distributed across all localities. However, ifwe see partitioning of communities among meadows, then post-colonizationprocesses are likely (Bostro?m and Bonsdorff, 2000). To address these ques-tions I tested the following hypotheses:1. Epifaunal diversity and composition driven by abiotic factors (tem-perature, salinity) and meadow characteristics (meadow area, density)differs among seagrass meadows, and within seagrass meadows throughtime.2. Variation in community richness and diversity will be greater amongmeadows than within meadows.3. Post-settlement processes are more important than dispersal in struc-turing epifaunal communities.2.2 Materials & Methods2.2.1 Study System & Field SitesBarkley Sound, British Columbia, is a well-mixed marine ecosystem thatsupports many seagrass meadows with associated epifaunal communities.In the case of such a well mixed system it would be expected that epi-faunal communities within a small region (10?s of kilometers) would bevery similar to one another due to frequent colonization events. There-fore, I conducted field surveys at five sites along Trevor Channel in BarkleySound separated by a mean distance of 5.18 km, with a maximum distanceof 17.74 km between furthest sites (Figure 2.1). Barkley Sound experi-ences mixed semi-diurnal tides and ranges in temperature from 7 ?C to 20?C with an average of 12 ?C, and salinity ranges from 10 ppt to 30 pptwith an average of 23 ppt annually (Pawlowicz, 2005). The sites are ori-ented from Southwest to Northeast from the Pacific Ocean to Alberni Inletas follows: Dodger Channel (DC, 48?50.023?N, 125?11.720?W), Wizard Is-land (WI, 48?51.495?N, 125?09.520?W), Robber?s Passage (RP, 48?53.648?N,125?07.104?W), Numukamis Bay (NB, 48?54.391?N, 125?00.740?W), and Crickitt82.2. Materials & MethodsBay (CB, 48?56.457?N, 125?01.117?W). Trevor Channel is a very deep rockyfjord (+200m at its deepest point) and seagrass meadows are constrainedto a fringing zone of fairly shallow photosynthetic depth limits of less than10m, and bounded by rare suitable substrates, effectively isolating habitatpatches from one another. The sites consisted of highly exposed locales,subject to wind waves and swell (WI and NB) and sites sheltered from windand waves (DC, RP, CB).Figure 2.1: Map of Trevor Channel in Barkley Sound, BC. Primary samplingsites are indicated by red circles; Dodger Channel (DC), Wizard Island (WI),Robber?s Passage (RP), Numukamis Bay (NB), Crickitt Bay (CB). Addi-tional artificial seagrass deployment sites are indicated with green stars;Eagle Bay (EB), Bulldozer Beach (BB), Clifton Point (CL).92.2. Materials & Methods2.2.2 Epifaunal SamplingTo facilitate community comparisons of plots within and among meadows,I sampled standard plots in a structured spatial regime in May and Augustof 2012. Each epifaunal community was sampled twice at Wizard Island(20 May 2012, 5 August 2012), Robber?s Passage (22 May 2012, 7 August2012), Numukamis Bay (24 May 2012, 9 August 2012), Crickitt Bay (26May 2012, 11 August 2012), and Dodger Channel (28 May 2012, 3 August2012). I sampled subtidally using SCUBA in a 4 x 4 m design for a totalof 16 0.28 m2 plots per site, each plot spaced 1 m from its nearest neighborafter Sanders et al. (2007a) (Figure 2.2). This method allows for comparisonof diversity among plots and meadows while standardizing for total areasampled. Sampling at each site took place within contiguous meadows notcloser than 2 m to the edge to minimize the risk of capturing edge effectsin diversity and composition of epifauna. All sampled areas were at least 1m below LLWLT, and did not vary in depth by more than a meter at anygiven site. I cut away seagrass within each sampled quadrat at the sediment-water interface and placed it into a 250?m mesh bag sufficient to capture alladult epifauna. Seagrass was then transported back to the lab in seawaterand processed within 24 hours of collection. Samples were preserved in 70%EtOH and identified to species when possible, or grouped by morpho-speciesto finest taxonomic resolution with light microscopy.102.2. Materials & MethodsR.#Whippo#Figure 2.2: Sampling regime at each meadow. Inset depicts horizontal sec-tion of seagrass sampled in each quadrat.2.2.3 Quantifying Abiotic Conditions, Seagrass Density, &ProductivityTemperature/Salinity MonitoringTo characterize water conditions, all sites were monitored for temperatureand salinity by boat using a hand held temperature/salinity probe (YSIInc., OH USA). Measurements were taken at various times of day and atdifferent points in the tidal cycle. Three stations were established at eachsite representing approximately the center, and furthest edges of the largestcontiguous meadow. At each station temperature and salinity were recordedat the surface, 2 m below the surface, and directly above the bottom. If thebottom was at 2m, only surface and 2m depth were recorded. If the bottomwas shallower than 2m, bottom depth was recorded and surface and bottommeasurements were taken.Meadow Structural ParametersMeadow area was calculated by visual survey, paired with the CommunityMapping Network database (CMN, 2013) and estimations made in GoogleEarth v 7.1.1.1888.112.2. Materials & MethodsMeadow structural parameters were measured concurrently with epifau-nal sampling. A 0.28m2 quadrat was placed outside each corner of a 4 x 4m grid demarcated for community sampling. All shoots within the quadratwere counted, and one haphazardly selected shoot was removed for leaf areafor a maximum of four shoots per site/time.Number of blades per shoot were determined for each seagrass shoot andthe longest blade was measured for length (from top of sheath to tip of blade)and width (at the midpoint). To calculate leaf area the width and length ofthe longest blade were multiplied by the number of blades for each shoot (af-ter Borg et al., 2010). It should be noted that this resulted in a probableoverestimation of total leaf area, however, consistent measuring procedureallows for comparison of relative leaf area across sites and sampling times.Artificial Seagrass & DispersalTo estimate the potential for presettlement processes to influence the com-munity composition, I quantified the presence of seagrass associated epi-fauna outside of seagrass meadows. Three artificial seagrass units (ASUs)were deployed at six sites in Barkley Sound including Dodger Channel, Rob-ber?s Passage, Crickitt Bay, Eagle Bay (48?50.024?N, 125?08.788?W), Bull-dozer Beach (48?52.134?N, 125?09.874?W), and Clifton Point (48?55.105?N,125?03.697?W) (Figure 2.1). Units were constructed of 30cm lengths of PVCanchoring material with two 25 cm segments of frayed polypropylene ropeaffixed at one end. Units were deployed in groups of three at each site,approximately 1 m apart and anchored to the sediment parallel to shore atapproximately 0.5m above lower low water large tide (LLWLT). ASUs werefirst primed in flowthrough seawater tables for seven days to accumulatenatural algal epiphytes and provide more suitable substrate for epifaunalorganisms. Units were left in the field for two weeks in July 2013. Oncecollected all ASUs were gently rinsed and attached organisms were removedand filtered through a 500?m sieve to capture target species. All specimenswere preserved in 70% EtOH and identified to lowest taxonomic categorypossible.122.2. Materials & Methods2.2.4 Statistical AnalysesTreatment of DataThe variance of all response variables in the following datasets were tested fornormality using the Shapiro-Wilk test and visualized with quantile-quantileplots. Where variance among samples was homogenous, linear models werecreated and analysis of variance (ANOVA) tests were run and pairwise dif-ferences between groups detected with post-hoc Tukey tests. Where vari-ance among samples was different, generalized linear models (GLM) weremade using a Poisson Distribution and subjected to analysis of devianceChi-squared tests (ANODE-Chi). General linear hypotheses (GLH) wereused to detect pairwise differences in the GLM. Where data was not nor-mally distributed and bounded between 0 and 1, but not strictly binomial(diversity and evenness measures), values were multiplied by 100 and anoffset term was added to the GLM.Comparison of Biotic & Abiotic Factors Among SitesAll differences in abiotic and structural parameters among sites were testedwith ANOVA, post-hoc Tukey tests, ANODE-Chi, and GLH.Quantifying Epifaunal Abundance/Diversity In & BetweenSeagrass MeadowsDifferences in richness between ASUs inside and outside seagrass were testedwith ANOVA. Presence/absence of organisms was used to determine thedispersal capability of collected taxa.Quantifying & Comparing Epifaunal CommunitiesMorpho-species Grouping The variable appearance of some epifaunalgroups based on age and sex, as well as the damaging effects of preserva-tion, can artificially inflate measures of richness and diversity. In order toaccount for this effect, similar taxonomic groups were condensed into broadermorpho-species categories. These more conservative groupings allow more132.2. Materials & Methodsaccurate assignment to taxonomic categories. For the purpose of all futheranalyses, only results for these data are shown. For results of the raw datasetsee Appendix A.Abundance Abundance of epifauana consisted of mean numerical countsof organisms across all size classes within each sample. Log abundance ofcounts was used for visual comparisons where counts spanned more thanone order of magnitude. Rank abundance curves were generated for eachsite and time.Richness Epifaunal richness was calculated for each sample collected astotal number of species encountered per sample. Given the same regionalspecies pool, a greater abundance of organisms in a sample could correlatewith a greater likelihood of detecting rare species, and therefore, a higherestimate of local richness. To control for this effect of abundance, I alsocalculated rarified richness (Buddle et al., 2005). Differences between mea-sures of raw richness were analysed using ANODE-Chi and GLH. Rarefiedrichness was analysed with ANOVA and post-hoc Tukey.Diversity & Evenness Diversity within meadows was calculated for eachsite and sampling time with Simpson?s Index:Dsimp = 1?n?i=1p2i (2.1)This measure accounts for heterogeneity (richness and evenness) in thecommunity and is robust to the effects of highly abundant species (Magur-ran, 2004). All values are bounded between 0 and 1, with greater diversityindicated by higher values. Evenness was calculated using Simpson?s Even-ness measure:E1/D =1/DS(2.2)This measure is derived from Simpson?s Diversity Index and is recom-mended for use when Simpson?s Index is already used (Magurran, 2004).142.2. Materials & MethodsBeta diversity within and among meadows was calculated using the Bray-Curtis dissimilarity measure:dBCDij =n?k=1|xik ? xjk|n?k=1(xik + xjk)(2.3)The Bray-Curtis dissimilarity measure is a quantitative method for esti-mating beta diversity as variation among samples that accounts for compo-sition of a community as well as the relative abundance of organisms (Clarkand Warwick, 2001; Anderson et al., 2011). Values are bounded between 0and 1, 0 representing complete similarity, 1 representing complete dissimi-larity.Community Composition Variation within and among communities at-tributed to species abundance and composition was tested with permuta-tional analysis of variance (PERMANOVA) using a Modified Gower trans-formation and visualized in a non-metric multidimensional scaling plot (nMDS).Contributions of species to dissimilarity among sites were calculated with4th root transformed data in similarity percentage (SIMPER) analysis us-ing pairwise comparisions of the Bray-Curtis dissimilarity measure (Equa-tion 2.3) which were then compared with rank abundance curves to deter-mine if trends were driven by common or rare species.Quantifying Relationships Between Structural Parameters &Epifaunal CommunitiesI used linear and mixed-effects to test for linear correlations between esti-mates of epifanual community abundance and diversity, and seagrass meadowbiotic and abiotic properties. Best model fits were determined by the AkaikeInformation Criterion (AIC).152.3. ResultsStatistical PackagesAll analyses were run in R version 3.0.0 (R Core Team, 2013) using thevegan package (Oksanen et al., 2013) for community analyses, and visualizedin the Lattice package (Sarkar, 2008). Additional multivariate statisticalapproaches were run and visualized with Primer-E (Clark and Gorley, 2006).2.3 Results2.3.1 Comparison of Biotic & Abiotic Factors Among SitesThe mean summer water temperature of sites between May 1, 2012 and Au-gust 11, 2012 (DC: n = 63, WI: n = 76, RP: n = 59, NB: n = 57, CB: n = 55)increased closer to the Alberni watershed in the northeast (Table 2.1), anda concurrent decline in mean salinity was also seen (Table 2.2, Figure 2.3).Mean temperatures ranged from 12.5 to 14.5 ?C, mean salinities were seenfrom 26 ppt to 17 ppt.Table 2.1: ANODE-Chi of mean temperatures across all sites from May toAugustDf Deviance Resid. Df Resid. Dev Pr(>Chi)NULL 313.00 160.282site 4 14.89 309.00 145.394 0.0000Table 2.2: ANODE-Chi of mean salinities across all sites from May to AugustDf Deviance Resid. Df Resid. Dev Pr(>Chi)NULL 313.00 160.282site 4 14.89 309.00 145.394 0.0000162.3. ResultsDC WI RP NB CBaverage sea temp (C)111315AAB BB BDC WI RP NB CBaverage sea salinity (ppt)14182226 CB BA AFigure 2.3: Average water temperature and salinity from May to August,2012. Temperatures and salinities were different among sites with pairwisedifferences between sites (Tests for General Linear Hypotheses, p < 0.05).Meadow area over the course of the summer was approximately 22,980m2 (DC), 2,645 m2 (WI), 7,213 m2 (RP), 26,960 m2 (NB), and 4,950 m2 (CB).Mean shoot density was higher at DC that CB in May (Table 2.3). ByAugust no differences in density among meadows were seen (Table 2.4), how-172.3. Resultsever, a trend for higher densities at DC and lower densities at CB remained(Figure 2.4).Table 2.3: ANOVA of seagrass density in MayDf Sum Sq Mean Sq F value Pr(>F)site 4 143.63 35.91 3.908 0.0295Residuals 12 110.25 9.19Table 2.4: ANOVA of seagrass density in AugustDf Sum Sq Mean Sq F value Pr(>F)site 4 63.36 15.84 1.926 0.1620Residuals 14 115.17 8.23182.3. Resultsshoot density, cm2  per 0.28m2  plot051015 AB AB ABABADC WI RP NB CBshoot density, cm2  per 0.28m2  plot051015 BFigure 2.4: Shoot density in May (A) and August (B) in 0.28m2 quadratsfor each sampling site. Bars represent standard error.Leaf area index was different among sites in May and August (Tables 2.5,2.6). LAI was higher at DC than NB in May, but was higher than CB inAugust (Figure 2.5).Table 2.5: ANOVA of LAI in MayDf Sum Sq Mean Sq F value Pr(>F)site 4 110292881.02 27573220.25 3.710 0.0345Residuals 12 89176588.60 7431382.38192.3. ResultsTable 2.6: ANOVA of LAI in AugustDf Sum Sq Mean Sq F value Pr(>F)site 4 36446542.56 9111635.64 3.256 0.0438Residuals 14 39172607.46 2798043.39leaf area, cm2  per 0.28m2  plot02000600010000ABABABAABDC WI RP NB CBleaf area, cm2  per 0.28m2  plot02000600010000BBAB AB ABAFigure 2.5: Leaf area index in May (A) and August (B) in 0.28m2 quadratsfor each sampling site. Bars represent standard error.2.3.2 Quantifying Epifaunal Abundance/Diversity In &Between Seagrass MeadowsDue to loss in the field and logistical difficulties, not all three ASUs persite were used in analysis (DC: n = 2, CB: n = 1, RP: n = 2, CL: n = 2,BB: n = 3, EB: n = 2). No differences in richness were detected between202.3. ResultsASUs placed within seagrass meadows compared to those outside seagrassmeadows (Table 2.7). However, a trend for more species collected by ASUswithin seagrass was observed (Figure 2.6). ASUs captured both overlappingand non-overlapping epifauna, 12 species which were shared, 9 found only inASUs inside seagrass, and 5 found only in ASUs outside seagrass (Table 2.8).Table 2.7: ANOVA of species richness comparing ASU inside (in) versusoutside (out) seagrass meadowsDf Sum Sq Mean Sq F value Pr(>F)treatment 1 34.29 34.29 4.412 0.0620Residuals 10 77.71 7.77in out4681012treatment# of speciesFigure 2.6: Mean richness of ASUs placed inside (in) and outside (out)seagrass meadows in July 2013.212.3. ResultsTable 2.8: Presence (X) and absence (-) of species identified in ASUs inside(IN) and outside (OUT) seagrass meadows, July 2013.IN OUTSite DC RP CB EB BB CLCaprella sp. X X X X X XAmphipoda X X X X X XNereidae X X X X X XCopepoda X X X X - XTanaidacea X X X - - XGastropoda X X - X X XUnidentified Caridea X X - - - -Haminoea X - X - - -Unidentified Microarthropod X - X - - -Leptostraca X - - X - -Ostracoda X - - - - -Bittium spp. X - - - - -Idotea resecata - X X X X XMytilus sp. - X X X X XParacerceis sp. - X X - - XNematoda - X X - - -Unidentified Bivalvia - X - X - -Tubeworm - X - - - -Unknown Decapoda - X - - - -Acari - - X - - -Chironomidae - - X X - XNemertea - - - X - -Clinocardium nuttallii - - - X - -Calliostoma sp. - - - X - -Pisaster ochraceus - - - - - XPycnogonida - - - - - X222.3. Results2.3.3 Quantifying & Comparing Epifaunal CommunitiesAmong MeadowsAbundanceAbundance of organisms differed among sites and times (Table 2.9). Meanorganismal abundance in May 2012 was relatively low and differed amongsites, except for WI and CB (Figure 2.7). Abundances in August were largerat all sites, with differences observed among all sites. Strength of changesin abundance through time were dependent on site identity. Abundancedecreased from highest abundance at DC, to lowest abundances at NB andCB. Dominant taxa contributing to organismal abundances differed amongsites through time (Figures 2.8 and 2.9). Caprella sp. were most abundantat DC and RP in May, while Amphipoda, Idotea resecata, and Pycnogonidawere most abundant at WI, NB, and CB respectively. In August, the numer-ically dominant taxa were Phyllaplysia taylori at DC and RP, and Mytilusspp. at WI, NB, and CB.Table 2.9: ANODE-Chi of organismal abundance by site and time.Df Deviance Resid. Df Resid. Dev Pr(>Chi)NULL 159.00 80579.939site 4 22924.24 155.00 57655.695 0.0000time 1 35316.47 154.00 22339.227 0.0000site:time 4 1614.74 150.00 20724.491 0.0000232.3. ResultsSiteslog abundance123DC WI RP NB CBD A C B AMayDC WI RP NB CBEDCB AAugustFigure 2.7: Log abundance of all organisms for each site in May and August,2012.242.3.Results0 5 10 20 300100200300400500600700DCspecies rankabundanceCaprella.sp.Amphipoda0 5 10 20 30WIspecies rankabundanceAmphipodaNereidae0 5 10 20 30RPspecies rankabundanceCaprella.sp.NereidaeAmphipoda0 5 10 20 30NBspecies rankabundance Idotea.resecataCopepoda0 5 10 20 30CBspecies rankabundancePycnogonidaIdotea.resecataCaprella.sp.Figure 2.8: Rank abundance curves of most abundant epifauna for each site in May, 2012.252.3.Results0 5 10 20 3002000400060008000DCspecies rankabundancePhyllaplysia.tayloriCaprella.sp.AmphipodaNereidae0 5 10 20 30WIspecies rankabundanceMytilus.spp.Caprella.sp.Tubeworm0 5 10 20 30RPspecies rankabundancePhyllaplysia.tayloriCaprella.sp.Mytilus.spp.NereidaeTubeworm0 5 10 20 30NBspecies rankabundanceMytilus.spp.NematodaIdotea.resecata0 5 10 20 30CBspecies rankabundanceMytilus.spp.TanaidaceaFigure 2.9: Rank abundance curves of most abundant epifauna for each site in August, 2012.262.3. ResultsRichnessA total of 54 taxa belonging to 7 phyla were identified across all meadowsfrom May to August 2012 (Table 2.10). The most common phyla repre-sented included Arthropoda, Mollusca, and Polychaeta, containing 85% of allindividual organisms identified.Table 2.10: List of organisms enumerated grouped bymorpho-species.Species & Morpho-speciesAmphipodaAoroides columbidaePhotis brevipesEogammarus confervicolusCorophium sp.Amphithoe sp.Unidentified Amph 1Unidentified Amph 2Cirripedia:Sessilia Cirripedia:SessiliaBittium sp. Bittium sp.Ophiuroidea OphiuroideaCaprella sp. Caprella sp.Cirolana sp. Cirolana sp.Unidentified Bivalvia Unidentified BivalviaClinocardium nuttallii Clinocardium nuttalliiCopepoda CopepodaUnidentified Decapoda Unidentified DecapodaEuphausiacea EuphausiaceaHaminoea HaminoeaPaguridae PaguridaeIdotea resecata Idotea resecataLeptostraca LeptostracaContinued on next page272.3. ResultsTable 2.10 ? continued from previous pageMorpho-species & SpeciesLottia pelta Lottia peltaAcari AcariMytilus spp. Mytilus spp.NematodaNematode ANematode BNemertea NemerteaPisaster ochraceus Pisaster ochraceusOlivella sp. Olivella sp.Ostracoda OstracodaUnidentified Microarthropod Unidentified MicroarthropodParacerceis sp. Paracerceis sp.Phyllaplysia taylori Phyllaplysia tayloriNereidaeNereid ANereid BNereid CNereid DNereid ENereid FNereid GNereid HNereid IPugettia spp. Pugettia spp.Pycnogonida PycnogonidaChlamys spp. Chlamys spp.ShrimpEualus spp.PandalidaeUnidentified CarideaGastropoda Margarites spp.Lacuna spp.Continued on next page282.3. ResultsTable 2.10 ? continued from previous pageMorpho-species & SpeciesSolaster sp. Solaster sp.Strongylocentrotus spp. Strongylocentrotus spp.Tanaidacea TanaidaceaTubeworm Tubeworm292.3. Results# of Species051015MayABBABABAugustSitesRarefied Richness246810DC WI RP NB CBABA AADC WI RP NB CBABCABBCABFigure 2.10: Mean raw richness (A) and rarefied richness (B) across all sitesfrom May and August 2012. Pairwise comparisons were made within eachtime among sites (Raw richness: Tests for General Linear Hypotheses, p <0.05, Rarefied richness: Tukey, p < 0.05).Mean richness values per within-meadow sample ranged from 4.63-6.31 /0.28 m2 in May, to 6.00-11.38 / 0.28 m2 in August (Figure 2.10). Minimumand maximum richness values for within-meadow samples ranged from 0-3 /0.28 m2 and 7-10 / 0.28 m2 in May, to 0-7 and 8-16 in August, respectively302.3. Results(Table 2.11). Rarefaction standardized richness estimates to the lowestobserved abundance using a cut-off of 10 individuals minimum per sam-ple for each time period. Rarefaction changed the relative mean richnessestimates among meadows, reflecting patterns in richness and abundanceacross the five meadows sampled (Figure 2.10). Rarefied richness variedsignificantly among meadows, but no differences were detected through time(Table 2.12).Table 2.11: ANODE-Chi of species richness by site and time.Df Deviance Resid. Df Resid. Dev Pr(>Chi)NULL 159.00 271.823site 4 24.16 155.00 247.661 0.0001time 1 55.59 154.00 192.075 0.0000site:time 4 12.82 150.00 179.260 0.0122Table 2.12: ANOVA of rarefied species richness by site and time.Df Sum Sq Mean Sq F value Pr(>F)site 4 62.92 15.73 22.882 0.0000time 1 0.00 0.00 0.007 0.9336site:time 4 19.77 4.94 7.191 0.0000Residuals 137 94.18 0.69Diversity & EvennessDiversity Within & Among Meadows Diversity varied among sites aswell as within sites through time (Figure 2.11). Highest mean diversity wasobserved at WI in May, lowest at DC, RP, and NB, with CB intermediatebetween sites. In August, diversity at the most spatially distant sites wasdifferent. CB was 30% higher than at DC, with all other sites falling inter-mediate between the two, indicating that the relative diversity of meadowschanged across space through time. However, site diversity did not changepredictably through time, indicated by an interaction between site and timein analysis (Table 2.13).312.3. ResultsSitesSimpson's Index0.20.40.60.81.0DC WI RP NB CBABCAABBMayDC WI RP NB CBAB ABABCAugustFigure 2.11: Mean diversity per 0.28 m2 plot across all sites in May andAugust 2012.Table 2.13: ANODE-Chi of species diversity measured as Simpson?s Index.Df Deviance Resid. Df Resid. Dev Pr(>Chi)NULL 153.00 630.200site 4 93.75 149.00 536.449 0.0000time 1 16.94 148.00 519.506 0.0000site:time 4 27.51 144.00 491.994 0.0000Evenness Evenness varied among sites through time (Figure 2.12). InMay WI and CB had higher evenness measures than DC, RP, or NB. How-ever, in August CB and NB had higher values than DC, WI, and RP.322.3. ResultsSitesSimpson's Evenness0.20.40.60.8DC WI RP NB CBABCAC BMayDC WI RP NB CBA A ABBAugustFigure 2.12: Mean community evenness per 0.28 m2 plot of all sites in Mayand August.Table 2.14: ANODE-Chi of species evenness by site and time.Df Deviance Resid. Df Resid. Dev Pr(>Chi)NULL 153.00 1101.807site 4 129.99 149.00 971.816 0.0000time 1 227.49 148.00 744.331 0.0000site:time 4 110.25 144.00 634.082 0.0000Beta Diversity Beta diversity varied among sites within and betweentime periods (Figure 2.13). DC and RP tended to be more similar in Mayand August, while WI, NB, and CB were distinct in May, becoming moresimliar in August.332.3. Results?0.4 0.0 0.2 0.4?0.50.00.51.0ADCRPWINBCBPCoA 2?0.4 0.0 0.2 0.4BDCRPWICBNBFigure 2.13: Beta diversity as Bray-Curtis distance for each site in May (A)and August (B). Red points represent a multidimensional centroid for eachsite, blue vectors are distance of sampled 0.28 m2 plots within site to thecentroid. Axes are dimensionless.Multivariate distance of each sample to a multidimensional centroid gen-erated for each site indicated variation in mean beta diversity within eachsite and time. In May, beta diversity as variation in community compositionand abundance was significantly higher at WI than at DC, RP, or CB (Fig-ure 2.14). In August, variation increased across sites from DC to CB, withthe greatest differences observed between the most spatially distant sites,DC and CB.342.3. ResultsDC WI RP NB CB0.00.20.40.60.8Distance to CentroidAABAABADC WI RP NB CBBAABABABB     SitesFigure 2.14: Mean distance to centroid, Bray-Curtis beta diversity of all sitesin May (A) and August (B). Differences in variation among sites within eachtime detected with post-hoc Tukey (p < 0.05).Community CompositionCommunity composition varied among sites through time (Figure 2.13).Composition was unique within sites, and changed differently through timedepending on the site observed (Table 2.15). SIMPER analysis paired withspecies rank abundance curves (see section 2.3.3) indicated that often themost abundant organisms were responsible for driving differences in compo-sition among sites and times (Table A.1). Caprella sp., Amphipoda, Mytilusspp., Nereidae, and Phyllaplysia taylori were included in the top five con-tributors to variation in epifaunal composition for at least half of all pairwisesite comparisons.352.3. ResultsTable 2.15: PERMANOVA of community composition.Source df SS MS Pseudo-F P(perm) Unique perms P(MC)site 4 41.817 10.454 2.2608 0.0219 7239 0.0138time 1 19.299 19.299 42.687 0.0001 9940 0.0001site x time 4 18.497 4.6242 10.228 0.0001 9867 0.0001Res 150 67.815 0.4521Total 159 147.43Figure 2.15: nMDS of epifaunal communities across all sites in May andAugust.362.3. ResultsTable 2.16: Akaike Information Criterion for density models.Model AIC ?AIC ?AIClog abundance average density 20.06 0 0.7770log abundance average density, random = time 22.56 2.50 0.22302.3.4 Quantifying Relationships Between StructuralParameters & Epifaunal CommunitiesTotal abundance of organisms increased with increasing mean seagrass shootdensity per site and time (Figure 2.16). Best model fit was linear regression(Table 2.16.4 6 8 10 123.03.54.0Mean Shoot DensityLog Organismal Abundancep < 0.05r^2 = 0.23Figure 2.16: Organismal abundance as a function of shoot density in0.28m2 plots.372.3. ResultsTable 2.17: Akaike Information Criterion for LAI models.Model AIC ?AIC ?AIClog abundnace average LAI 18.95 0 0.9996log abundnace average LAI, random = time 35.43 16.48 0.0003log abundnace average LAI, random = average LAI | time 36.95 18.00 0.0001Total organismal abundance did not significantly increase with increas-ing leaf area index (Figure 2.17). Best model fit was linear regression (Ta-ble 2.17).2000 3000 4000 5000 6000 7000 80003.03.54.0Mean Leaf AreaLog Organismal Abundance:p = 0.08r^2 = 0.31Figure 2.17: Organismal abundance as a function of leaf area in 0.28m2 plots.382.4. DiscussionMean leaf area and density did not correlate with any other communitymeasures including rarefied richness, diversity, or evenness.2.4 DiscussionEpifaunal communities in seagrass meadows of Barkley Sound exhibit meadow-scale variability over small spatial and temporal scales. The patterns ob-served in these meadows are consistent with the highly variable seagrasscommunity patterns reviewed by Bostro?m et al. (2006a). In my study, bothintra- and inter-meadow variability in abundance, diversity, evenness, andcomposition were seen in May and August, however, inter-meadow vari-ability was more pronounced in August. Shifts in the relative size of inter-meadow variability, which was driven partly by recruitment events, indicatesseasonal fluctuations in epifaunal communities across meadows. These re-sults have identified the scale of epifaunal community similarity relative toseagrass meadows in the region, and suggest that local underlying processesare important for structuring these communities.The variability in community patterns observed suggests processes struc-turing epifaunal communities may differ within sites over the course of asummer. Seasonality of epifaunal taxa has been observed in other Z. marinameadows, though target species have been limited in scope and are likelyonly a subset of total community diversity (Thom et al., 1995). In con-trast, studies of seagrass systems in other regions have found little evidencefor seasonal variability in distribution of multiple dominant taxa through-out the year (Gambi et al., 1992). In the meadows I observed, variation inthe epifaunal community differed through time depending on site. In May,intra-meadow variability in rarefied richness and diversity was greater thaninter-meadow variability, except at WI which had greater richness than theother sites. In August, greater inter-meadow variation in diversity and rich-ness became more apparent as the most distant sites (DC and CB) beganto diverge, with the other sites intermediate between the two. In addition,abundance of organisms increased at all sites from May to August, thoughsome were greater by an order of magnitude. This variation in abundance392.4. Discussionis reflected in the composition of epifaunal communities, which were uniqueacross all sites through time.Many of the differences observed in these epifaunal communities are aresult of recruitment events seen at all sites by the end of the summer. Thesea hare Phyllaplysia taylori and caprellid Caprella sp. were seen in muchhigher abundances relative to other fauna at DC and RP, while WI, NB,and CB were dominated by the mussel Mytilus spp.. All of these individ-uals were likely the result of reproductive events isolated within meadows(for P. taylori, attached egg masses with crawl-away juveniles), or occur-ring within and among meadows (for Mytilus spp., broadcast spawners withpelagic larval stage). In the absence of other structuring forces, these differ-ential dispersal strategies would be expected to create patchy distributionsof sea hares, with more homogenous abundances of mussels. In fact, thiswas largely observed in both epifaunal community analyses and with ASUs.However, the near exclusion of Mytilus sp. from DC meadows and ASUsindicated that other local processes are determining colonization at that site.The ability of epifaunal organisms to disperse across the study area didnot appear to be limited for the most common taxa. All major contributorsto meadow composition in 2012, with the notable exceptions of P. tayloriand pycnogonids, were observed in ASUs placed both inside, and outsideseagrass meadows in 2013. Interestingly, the observed pycnogonid was asingle specimen from the CL site on an ASU placed outside seagrass. Thiscontrasts with the relatively larger population of pycnogonids seen withinseagrass at the CB site both in May and August of 2012, which is the closestsite to CL where ASUs were placed in 2013. Even though pycnogonids werepresent in the regional species pool, their absence, or reduced abundance atCB in 2013 suggests that inter-annual seasonal variation occurrs within sites.The complete absence of P. taylori from all ASU samples in 2013, is likelyeither an artifact of the ASU itself, or an indicator that patterns of diversityin Barkley Sound vary inter-annually. This variation is also seen in thesurprising inclusion of chironomid larvae within ASU samples. Chironomidlarvae are typically fresh-water young of midge flies and allies and weredetected at three ASU sampling sites, and were not detected at all in 2012.402.4. DiscussionAn inter-annual shift in taxa within and among seagrass meadows agreeswith observations of Balata et al. (2006) that indicate temporal variabilityin marine invertebrate assemblages is much greater than previously thought.This presents many challenges for the management of seagrass systems andassociated communities, which may be experiencing structuring forces, notonly across spatial scales, but across temporal scales that could be operatingover years or decades.The relative strength of post-settlement processes and connectivity instructuring epifaunal communities likely differs depending on location ofmeadow in a larger regional context, as well as life-history of particularepifaunal organisms. While some organisms were observed across a broadspatial range, others were not. Amphipods were seen to have the one ofthe highest colonization successes, as has been observed in other seagrasssystems (Gustafsson and Salo, 2012). Surprisingly, some taxa with highlydispersive pelagic larval stages were not seen at all sites, notably the mus-sel Mytilus spp., one of the most abundant community members in August2012. Mytilus spp. was completely absent from DC in May 2012, and ASUsplaced within DC in 2013. It was present in relatively low numbers in Au-gust 2013. The exclusion of mussels from DC could be a combination ofpredatory and competitive processes. In addition, their successful coloniza-tion and proliferation, even at sites which experience relatively lower meansalinity compared to oceanic norms, suggest that abiotic tolerance to localsalinity conditions could be favoring mussels, which have been shown to bothtolerate, and acclimate to low salinity conditions (Riisg?ard et al., 2013).Variations in diversity across space and time may be due to the strength-ening regional temperature and salinity gradients observed in the region.Gradients can be the strongest structuring force for some animal commu-nity parameters (Sanders et al., 2007b). In addition, salinity gradients areknown to structure near-shore marine communities, particularly in estuarineenvironments (Barnes and Ellwood, 2012). However, a pattern of increas-ing richness with declining average salinity exhibits an inverse relationshipcompared to previous observations in other Z. marina invertebrate commu-nities (Yamada et al., 2007). This may be in part a result of the regionally412.4. Discussionlow salinity of Barkley Sound compared to other studies which tended tosee breaks in species richness at 30ppt (Barnes and Ellwood, 2012). An in-teraction between abiotic and biotic forces such as abiotic filtering coupledwith competition, might be allowing small numbers of many taxa to colonizethe less favorable, low-salinity condition, while at the same time preventinglocal dominance of any one taxa. The general decline of organismal abun-dance across the observed salinity gradient tends to support this conclusion.Another possibility is that the least saline sites were highly diverse due tostochastic processes, and that diversity was maintained though an alterationof local trophic structure or productivity. Duffy (2006) observed that the di-versity of landscapes can determine ecosystem structure and processes, thismay then in turn maintain historic patterns of diversity.Relative exposure to local physical forcing may also explain some varia-tion among epifaunal communities in Barkley Sound. In the winter, seagrassmeadows are subject to heavy wind, waves, and other physical disturbancecaused by seasonal storms, but these mechanical processes can be very sitespecific depending placement of meadows within the landscape. Physicalforcing by wind and waves has been shown to maintain high diversity inseagrass meadows (Bostro?m and Bonsdorff, 2000). Variation in species as-semblages have also been explained by factors such as fetch, shore angle,and grain size of substrate (Bostro?m et al., 2006b). This would suggest thatsites with greater exposure to mechanical disturbance should exhibit higherdiversity. Indeed, Wizard Island, which is in the middle of Satellite Passagebetween Trevor Channel and Imperial Eagle Channel, and the most exposedof all the sites, had the highest measured diversity at the beginning of thesummer after majority of the seasonal storms had passed. As the winterstorms subside, diversity among meadows begins to diverge as the role oflocal predation and abiotic gradient begin to more strongly structure theepifaunal communities.Variation in nutrient inputs and seagrass epiphytes, which exert a bottom-up regulatory force, also structure epifaunal communities (Douglass et al.,2007). Inputs of nutrients can be seen both regionally as run-off from frag-mented or developed terrestrial habitat, or locally as point-source inputs422.4. Discussionfrom concentrated human, or animal activity. Nutrient enhancement likelyhas different impacts on secondary production and trophic dynamics withina system depending on the strength and spatial scale of the input. Thiscan fundamentally alter the community structure of seagrass meadows bydirectly increasing epiphyte biomass, which is then grazed by the epifaunalcommunity (Douglass et al., 2007; Bryars et al., 2011; Frankovich and Zie-man, 2005). A relative increase in nutrients could decrease diversity withinmeadows by allowing a rapid population increase of dominant grazing taxa,which are then able to supress less competitive community members. Thisagrees with findings by Duffy and Harvilicz (2001) that suggest inter-specificcompetition is stronger than intra-specific competition for some common epi-faunal taxa. Increased epiphyte biomass can also alter community dynamicsby increasing structural complexity (Bologna and Heck, 1999; Gartner et al.,2013). Structural complexity of seagrass habitat can provide indirect mod-ulation of trophic interactions by offering protection from predators andaltering predator behavior (Cardoso et al., 2007).In addition, seasonal migration of predatory fish such as shiner perch Cy-matogaster aggregata could alter abundance and richness of epifaunal taxavia preferential predation on competitively dominant species (Morrow, 1980;Caine, 1989). This could release rare species from competition and allowthem to proliferate, increasing local diversity. The role of fish predatorscannot be underestimated in this system as the role of top-down controlin seagrass is well documented (Moksnes et al., 2008; Baden et al., 2012;Lewis and Anderson, 2012). These effects could change depending on thetiming and frequency of these fish returns to nearshore habitat. The shinerperch Cymatogaster aggregata is a common predator of both Idotea andCaprella (Caine, 1989) and was observed frequently in all meadows in thisstudy. Populations of caprellids, which are abundant in winter, can begreatly reduced by predatory fish such as shiner perch (Caine, 1991). Thestructural effects of seagrass itself on predator behavior may also impact epi-faunal communities. Seagrass ecosystem structure can be modified directlyand indirectly by grazing fish behaviour which is modulated by habitat het-erogeneity (Macia? and Robinson, 2005).432.4. DiscussionThe loss of seagrass meadow area worldwide suggests that a workingknowledge of the scales at which different community processes function isvital to effectively preserving these habitats (Short and Wyllie-Echeverria,1996; Leibold et al., 2004). Here, I presented data that highlight intra- andinter-meadow variability in epifaunal diversity in several seagrass meadowsacross a summer season. These meadows demonstrate patterns of seasonal-ity as well as the potential for inter-annual variation in community structure.Biotic and abiotic forces acting across scales contribute to variation observedin patterns of epifaunal diversity and abundance. The relative strength ofthese various forces remains unknown, and future work identifying primarydrivers of community assembly and persistence are needed to better under-stand the dynamics of epifaunal communities in seagrass.44Chapter 3General Conclusions3.1 SummaryEpifaunal communities of Pacific Northwest seagrass meadows differ acrossrelatively small spatial and temporal scales. In Barkley Sound, various abi-otic drivers such as temperature and salinity may be contributing to ob-served patterns, though no definitive relationships between environmentalparameters and community composition have been determined.3.2 Recommendations For Future ResearchWhile broad scale patterns of diversity have been detected in the currentwork, a more detailed analysis of abiotic factors that control organismal colo-nization and persistence is recommended to more clearly elucidate drivers ofcommunity assembly. Effects of sub-lethal temperatures and salinity whichare seen in Barkley Sound throughout the summer may be determining com-munity composition among meadows. In addition, information regardingtrophic structure within these meadows, particularly those involving com-mon fish predators might allow us to make more specific predictions of epi-faunal diversity and composition based on seasonality and prey preferencesof the fish.One approach which accounts for processes at various spatial and tem-poral scales is the use of the ?metacommunity? concept (Declerck et al.,2011; Frederic Guichard & Steenweg, 2008; Leibold et al., 2004). A re-cent review of seagrass communities on the landscape level only brieflytouches on metacommunity models (Bostrom et al. 2006). The importanceof local versus regional diversity and environmental factors in structuring453.3. Conclusioncommunities has been addressed for near-shore ecosystems including sea-grass (Sanvicente-Anorve, Sanchez-Ramirez, Ocana-Luna, Flores-Coto, &Ordonez-Lopez, 2011; Witman, Etter, & Smith, 2004; de Juan & Hewitt,2011), but the connections to metacommunity theory have remained im-plicit, without invoking predictions provided by the major paradigms of thetheory.3.3 ConclusionOverall, I have found that seagrass meadow epifaunal communities of the Pa-cific Northwest are structured across local and regional scales. While muchwork remains to be done on the ecology of seagrass-associated communities,this study represents the first attempt to link diversity and composition ofseagrass invertebrate communities in this region. Futher testing of abioticand biotic parameters are likely to uncover primary drivers of communityassembly, and associated function provided by multiple guilds of inverte-brates. The testing of metacommunity principles in seagrass may proveespecially fruitful as they integrate processes across scales, and could be arobust predictive framework in these systems. This is especially importantfor the conservation of seagrass habitat which is currently under threat.46BibliographyAnderson, M. J., Crist, T. O., Chase, J. M., Vellend, M., Inouye, B. D.,Freestone, A. L., Sanders, N. J., Cornell, H. V., Comita, L. S., Davies,K. F., Harrison, S. P., Kraft, N. J., Stegen, J. C., and Swenson, N. G.(2011). Navigating the multiple meanings of beta diversity: A roadmapfor the practicing ecologist. Ecology Letters, 14(1):19?281.Baden, S., Emanuelsson, A., Pihl, L., Svensson, C.-J., and A?berg, P. (2012).Shift in seagrass food web structure over decades is linked to overfishing.Marine Ecology Progress Series, 451:61?73.Balata, D., Acunto, S., and Cinelli, F. (2006). Spatio-temporal variabilityand vertical distribution of a low rocky subtidal assemblage in the north-west Mediterranean. Estuarine, Coastal and Shelf Science, 67(4):553?561.Baldwin, J. and Lovvorn, J. (1994). Habitats and tidal accessibility ofthe marine foods of dabbling duck and brant in Boundary Bay, BritishColumbia. Marine Biology, 120(4):627?638.Barnes, R. and Ellwood, M. (2012). Spatial variation in the macrobenthicassemblages of intertidal seagrass along the long axis of an estuary. Estu-arine, Coastal and Shelf Science, 112:173?182.Best, R. J. and Stachowicz, J. J. (2012). Trophic cascades in seagrass mead-ows depend on mesograzer variation in feeding rates, predation suscepti-bility, and abundance. Marine Ecology Progress Series, 456:29?42.Bologna, P. A. and Heck, K. L. (1999). Macrofaunal association with sea-grass epiphytes: Relative importance of trophic structure characteristics.Journal of Experimental Marine Biology and Ecology, 242:21?39.47BibliographyBorg, J. A., Rowden, A. A., Attrill, M. J., Schembri, P. J., and Jones, M. B.(2010). Spatial variation in the composition of motile macro invertebrateassemblages associated with two types of the seagrass Posidonia oceanica.Marine Ecology Progress Series, 406:91?104.Bostro?m, C. and Bonsdorff, E. (2000). Zoobenthic community establishmentand habitat complexity ? The importance of seagrass shoot-density, mor-phology and physical disturbance for faunal recruitment. Marine EcologyProgress Series, 205:123?138.Bostro?m, C., Jackson, E. L., and Simenstad, C. A. (2006a). Seagrass land-scapes and their effects on associated fauna: A review. Estuarine, Coastaland Shelf Science, 68:383?403.Bostro?m, C., O?Brien, K., and Roos, C. (2006b). Environmental variablesexplaining structural and functional diversity of seagrass macrofauna inan archipelago landscape. Journal of Experimental Marine Biology andEcology, 335:52?73.Bruno, J. F. (2002). Causes of landscape-scale rarity in cobble beach plantcommunities. Ecology, 83(8):2304?2314.Bryars, S., Collings, G., and Miller, D. (2011). Nutrient exposure causesepiphytic changes and coincident declines in two temperate Australianseagrasses. Marine Ecology Progress Series, 441:89?103.Buddle, C. M., Beguin, J., Bolduc, E., Mercado, A., Sackett, T. E., Selby,R. D., Varady-Szabo, H., and Zeran, R. M. (2005). The importance anduse of taxon sampling curves for comparative biodiversity research withforest arthropod assemblages. Canadian Entomologist, 137(1):120?127.Caine, E. A. (1989). Caprellid amphipod behavior and predatory strikes byfish. Journal of Experimental Marine Biology and Ecology, 126:173?180.Caine, E. A. (1991). Caprellid amphipods: Fast food for the reproductivelyactive. Journal of Experimental Marine Biology and Ecology, 148:27?33.48BibliographyCaley, M. J. and Schluter, D. (1997). The relationship between local andregional diversity. Ecology, 78(1):70?80.Cardoso, P., Raffaelli, D., and Pardal, M. (2007). Seagrass beds and inter-tidal invertebrates: An experimental test of the role of habitat structure.Hydrobiologia, 575:221?230.Clark, K. and Gorley, R. (2006). PRIMER v6: User Manual/Tutorial.PRIMER-E.Clark, K. and Warwick, R. (2001). Change in marine communities: An ap-proach to statistical analysis and interpretation. Primer-E Ltd, PlymouthMarine Laboratory, UK.CMN (2013). Eelgrass bed atlas @ONLINE, retrieved fromhttp://www.cmnbc.ca/atlas gallery/eelgrass?bed?mapping on 15 Nov2013.Cook, K., Vanderklift, M. A., and Poore, A. G. (2011). Strong effects of her-bivorous amphipods on epiphyte biomass in a temperate seagrass meadow.Marine Ecology Progress Series, 442:263?269.Costanza, R., d?Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B.,Limburg, K., Naeem, S., O?Neill, R. V., Paruelo, J., Raskin, R. G., Sutton,P., and van den Belt, M. (1997). The value of the world?s ecosystemservices and natural capital. Nature, 387:253?260.Cottenie, K. (2005). Integrating environmental and spatial processes inecological community dynamics. Ecology Letters, 8:1175?1182.Cullen-Unsworth, L. and Unsworth, R. (2013). Seagrass meadows, ecosys-tem services, and sustainability. Environment, 55(3):14?28.de Juan, S. and Hewitt, J. (2011). Relative importance of local biotic and en-vironmental factors versus regional factors in driving macrobenthic speciesrichness in intertidal areas. Marine Ecology Progress Series, 423:117?129.49BibliographyDouglass, J. G., Duffy, J. E., Spivak, A. C., and Richardson, J. P. (2007).Nutrient versus consumer control of community structure in a ChesapeakeBay eelgrass habitat. Marine Ecology Progress Series, 348:71?83.Duarte, C. M. (1999). Seagrass ecology at the turn of the millennium:Challenges for the new century. Aquatic Botany, 65:7?20.Duffy, J. E. (2006). Biodiversity and the functioning of seagrass ecosystems.Marine Ecology Progress Series, 311:233?250.Duffy, J. E. and Harvilicz, A. M. (2001). Species-specific impact of graz-ing amphipods in an eelgrass-bed community. Marine Ecology ProgressSeries, 223:201?211.Fortes, W. L., Alameida-Silva, P. H., Prestrelo, L., and Monteiro-Neto,C. (2014). Patterns of fish and crustacean community structure in acoastal lagoon system, Rio de Janeiro, Brazil. Marine Biology Research,10(2):111?122.Frankovich, T. and Zieman, J. (2005). A temporal investigation of grazer dy-namics, nutrients, seagrass leaf productivity, and epiphyte standing stock.Estuaries, 28(1):41?52.Gambi, M. C., Lorenti, M., Russo, G. F., Scipione, M. B., and Zupo, V.(1992). Depth and seasonal distribution of some groups of vagile faunain the Posidonia oceanica leaf stratum: Structural and trophic analyses.Marine Ecology, 13(1):17?39.Gartner, A., Tuya, F., Lavery, P., and McMahon, K. (2013). Habitatpreferences of macro invertebrate fauna among seagrasses with varyingstructural forms. Journal of Experimental Marine Biology and Ecology,439:143?151.Gustafsson, C. and Salo, T. (2012). The effect of patch isolation on epifaunalcolonization in two different seagrass ecosystems. 159:1497?1507.50BibliographyHeck Jr., K., Hays, G., and Orth, R. (2003). Critical evaluation of thenursery role hypothesis for seagrass meadows. Marine Ecology ProgressSeries, 253:123?136.Hemminga, M. and Duarte, C. (2000). Seagrass ecology. Cambridge Univer-sity Press.Jaschinski, S. and Sommer, U. (2008). Functional diversity of mesograzersin an eelgrass-epiphyte system. Marine Biology, 154:475?482.Karlson, R. H., Cornell, H. V., and Hughes, T. P. (2004). Coral communitiesare regionally enriched along an oceanic biodiversity gradient. Nature,429:867?870.Leibold, M., Holyoak, M., and Mouquet, N. (2004). The metacommunityconcept: A framework for multi-scale community ecology. Ecology, 7:601?613.Lewis, L. S. and Anderson, T. W. (2012). Top-down control of epifauna byfishes enhances seagrass production. Ecology, 93(12):2746?2757.Logue, J. B., Mouquet, N., Peter, H., Hillebrand, H., and Group, T. M. W.(2011). Empirical approaches to metacommunities: A review and com-parison with theory. Trends in Ecology & Evolution, 26(9):482?491.Macia?, S. and Robinson, M. P. (2005). Effects of habitat heterogeneity inseagrass beds on grazing patterns of parrotfishes. Marine Ecology ProgressSeries, 303:113?121.Magurran, A. E. (2004). Measuring Biological Diversity. Blackwell Publish-ing.Malyshev, A. and Quijo?n, P. A. (2011). Disruption of essential habitat bya coastal invader: New evidence of the effects of green crabs on eelgrassbeds. Journal of Marine Science, 68(9):1852?1856.Moksnes, P.-O., Gullstro?m, M., Tryman, K., and Baden, S. (2008). Trophiccascades in a temperate seagrass community. Oikos, 117:763?777.51BibliographyMorrow, J. E. (1980). The freshwater fishes of Alaska. Alaska NorthwestPublishing Company, Anchorage, Alaska.Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R.,O?Hara, R. B., Simpson, G. L., Solymos, P., Stevens, M. H. H., and Wag-ner, H. (2013). vegan: Community Ecology Package. R package version2.0-7.Orth, R. J., Carruthers, T. J., Dennison, W. C., Duarte, C. M., Fourqurean,J. W., Heck Jr., K. L., Hughs, A. R., Kendrick, G. A., Kenworthy, W. J.,Olyarnik, S., Short, F. T., Waycott, M., and Williams, S. L. (2006). Aglobal crisis for seagrass ecosystems. Bioscience, 56(12):987?996.Pawlowicz, R. (2005). Barkley sound time series @ONLINE, retrieved fromhttp://www.eos.ubc.ca/ rich/bsts/ on 15 Nov, 2013.R Core Team (2013). R: A Language and Environment for Statistical Com-puting. R Foundation for Statistical Computing, Vienna, Austria.Richmond, C. and Woodin, S. (1999). Effect of salinity reduction on oxygenconsumption by larval estuarine invertebrates. Marine Biology, 134:259?267.Riisg?ard, H., Lu?skow, F., Pleissner, D., Lundgreen, K., and Lo?pez, M.(2013). Effect of salinity on filtration rates of mussels Mytilus edulis withspecial emphasis on dwarfed mussels from the low-saline Central BalticSea. Helgoland Marine Research, 67(3):591?598.Sanders, N. J., Gotelli, N. J., Wittman, S. E., Ratchford, J. S., Ellison,A. M., and Jules, E. S. (2007a). Assembly rules of ground-foraging ant as-semblages are contingent on disturbance, habitat and spatial scale. Jour-nal of Biogeography, 34(9):1632?1641.Sanders, N. J., Lessard, J.-P., Fitzpatrick, M. C., and Dunn, R. R. (2007b).Temperature, but not productivity or geometry, predicts elevational di-versity gradients in ants across spatial grains. Global Ecology and Bio-geography, 16:640?649.52Sarkar, D. (2008). Lattice: Multivariate Data Visualization with R. Springer,New York. ISBN 978-0-387-75968-5.Short, F. T. and Wyllie-Echeverria, S. (1996). Natural and human-induceddisturbance of seagrasses. Environmental Conservation, 23(1):17?27.Sirota, L. and Hovel, K. A. (2006). Simulated eelgrass Zostera marina struc-tural complexity: Effects of shoot length, shoot density, and surface areaon the epifaunal community of San Diego Bay, California, USA. MarineEcology Progress Series, 326:115?131.Thom, R., Miller, B., and Kennedy, M. (1995). Temporal patterns of grazersand vegetation in a temperate seagrass system. Aquatic Botany, 50:201?205.Vonk, J. A., Christianen, M. J., and Stapel, J. (2010). Abundance, edgeeffects, and seasonality of fauna in mixed-species seagrass meadows insouthwest Sulawesi, Indonesia. Marine Biology Research, 6:282?291.Witman, J. D., Cusson, M., Archambault, P., Pershing, A. J., andMieszkowska, N. (2008). The relation between productivity and speciesdiversity in temperate-arctic marine ecosystems. Ecology, 89(11):S66?S80.Witman, J. D., Etter, R. J., Smith, F., and Paine, R. T. (2004). Therelationship between regional and local species diversity in marine benthiccommunities: A global perspective. Proceedings of the National Academyof Sciences or the United States of America, 101(44):15664?15669.Yamada, K., Hori, M., Tanaka, Y., Hasegawa, N., and Nakaoka, M. (2007).Temporal and spatial macrofaunal community changes along a salinitygradient in seagrass meadows of Akkeshi-ko estuary and Akkeshi Bay,northern Japan. Hydrobiologia, 592:345?358.53Appendix AFirst AppendixSIMPER table of all pairwise comparisons between sites and times for ag-gregated species data.54AppendixA.FirstAppendixTable A.1: Similarity percentage of aggregated species com-position for all sites and times.Aggregated SIMPERGroups DC & WIAverage dissimilarity = 54.84Group DC Group WISpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 3.45 1.72 8.79 1.30 16.02 16.02Phyllaplysia taylori 2.82 0.54 7.53 1.75 13.73 29.75Mytilus spp. 0.37 2.16 4.95 1.13 9.03 38.78Amphipoda 2.25 1.76 3.48 0.85 6.35 45.12Nereidae 1.23 1.77 3.38 0.94 6.16 51.29Gastropoda 0.48 0.97 3.07 1.03 5.60 56.88Tubeworm 0.00 1.20 3.00 0.96 5.46 62.35Nematoda 0.06 0.44 2.81 0.69 5.13 67.48Unidentified Bivalvia 0.03 0.81 2.64 0.90 4.81 72.29Idotea resecata 1.20 1.12 2.44 0.77 4.44 76.73Continued on next page55AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERCirolana sp. 0.00 0.52 2.31 0.85 4.21 80.93Tanaidacea 0.31 0.30 2.16 0.74 3.93 84.86Copepoda 0.00 0.21 1.38 0.47 2.52 87.38Pycnogonida 0.06 0.43 1.14 0.66 2.09 89.47Lottia pelta 0.11 0.22 0.75 0.54 1.36 90.83Groups DC & RPAverage dissimilarity = 43.75Group DC Group RPSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Phyllaplysia taylori 2.82 1.92 6.62 0.82 15.14 15.14Amphipoda 2.25 1.20 6.11 0.96 13.97 29.11Caprella sp. 3.45 2.65 5.76 0.68 13.17 42.28Idotea resecata 1.20 0.53 4.45 1.18 10.16 52.44Nereidae 1.23 1.56 4.21 0.99 9.62 62.06Tubeworm 0.00 0.90 3.56 1.02 8.13 70.19Gastropoda 0.48 0.27 2.40 0.79 5.49 75.68Continued on next page56AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERMytilus spp. 0.37 1.01 2.27 0.74 5.19 80.87Tanaidacea 0.31 0.44 2.15 0.80 4.91 85.78Nematoda 0.06 0.55 1.38 0.66 3.15 88.93Pugettia spp. 0.22 0.03 0.89 0.47 2.03 90.96Groups WI & RPAverage dissimilarity = 59.01Group WI Group RPSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 1.72 2.65 8.96 1.24 15.19 15.19Phyllaplysia taylori 0.54 1.92 5.56 1.52 9.43 24.62Tubeworm 1.20 0.90 4.50 1.08 7.62 32.24Amphipoda 1.76 1.20 4.38 0.85 7.43 39.67Mytilus spp. 2.16 1.01 4.22 0.78 7.15 46.82Nereidae 1.77 1.56 4.00 0.84 6.78 53.60Nematoda 0.44 0.55 3.99 0.97 6.77 60.37Idotea resecata 1.12 0.53 3.95 1.01 6.69 67.06Continued on next page57AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERGastropoda 0.97 0.27 3.31 1.01 5.60 72.66Unidentified Bivalvia 0.81 0.22 2.82 0.88 4.78 77.44Tanaidacea 0.30 0.44 2.44 0.76 4.14 81.58Cirolana sp. 0.52 0.00 2.43 0.82 4.11 85.70Copepoda 0.21 0.00 1.45 0.43 2.46 88.15Pycnogonida 0.43 0.06 1.29 0.63 2.18 90.33Groups DC & NBAverage dissimilarity = 62.68Group DC Group NBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 3.45 0.79 12.61 2.40 20.12 20.12Phyllaplysia taylori 2.82 0.00 11.45 1.62 18.27 38.38Mytilus spp. 0.37 1.74 6.57 1.84 10.48 48.87Amphipoda 2.25 1.27 5.64 1.13 9.00 57.86Idotea resecata 1.20 1.85 5.07 0.97 8.09 65.95Nereidae 1.23 0.82 3.52 0.98 5.62 71.57Continued on next page58AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERAcari 0.00 0.53 3.20 0.72 5.11 76.68Nematoda 0.06 0.86 3.03 0.91 4.84 81.51Gastropoda 0.48 0.20 2.51 0.86 4.01 85.52Tanaidacea 0.31 0.30 2.27 0.79 3.62 89.14Copepoda 0.00 0.25 1.51 0.37 2.41 91.55Groups WI & NBAverage dissimilarity = 58.21Group WI Group NBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Mytilus spp. 2.16 1.74 5.51 1.36 9.46 9.46Idotea resecata 1.12 1.85 5.30 1.01 9.10 18.57Nematoda 0.44 0.86 5.00 1.38 8.58 27.15Caprella sp. 1.72 0.79 4.94 1.30 8.49 35.64Nereidae 1.77 0.82 4.86 1.28 8.36 44.00Gastropoda 0.97 0.20 3.91 1.35 6.72 50.72Tubeworm 1.20 0.00 3.77 0.95 6.48 57.20Continued on next page59AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERAmphipoda 1.76 1.27 3.50 0.82 6.02 63.22Unidentified Bivalvia 0.81 0.10 3.15 1.01 5.42 68.63Acari 0.07 0.53 3.09 0.70 5.30 73.94Cirolana sp. 0.52 0.09 2.41 0.85 4.14 78.08Copepoda 0.21 0.25 2.28 0.54 3.92 82.00Tanaidacea 0.30 0.30 2.18 0.66 3.74 85.75Phyllaplysia taylori 0.54 0.00 1.64 0.73 2.82 88.57Pycnogonida 0.43 0.00 1.43 0.68 2.45 91.02Groups RP & NBAverage dissimilarity = 66.29Group RP Group NBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 2.65 0.79 10.90 1.91 16.44 16.44Idotea resecata 0.53 1.85 8.44 1.17 12.73 29.17Phyllaplysia taylori 1.92 0.00 7.50 1.37 11.32 40.49Mytilus spp. 1.01 1.74 6.99 0.99 10.55 51.04Continued on next page60AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERNereidae 1.56 0.82 5.59 1.23 8.43 59.47Amphipoda 1.20 1.27 4.83 0.90 7.29 66.76Tubeworm 0.90 0.00 3.95 1.07 5.96 72.72Acari 0.00 0.53 3.39 0.66 5.11 77.83Nematoda 0.55 0.86 3.11 0.62 4.70 82.53Tanaidacea 0.44 0.30 2.68 0.82 4.05 86.57Gastropoda 0.27 0.20 1.87 0.67 2.82 89.39Copepoda 0.00 0.25 1.56 0.36 2.36 91.75Groups DC & CBAverage dissimilarity = 68.08Group DC Group CBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 3.45 0.80 12.95 1.85 19.02 19.02Phyllaplysia taylori 2.82 0.00 10.83 1.74 15.91 34.93Amphipoda 2.25 0.94 8.38 1.15 12.31 47.24Pycnogonida 0.06 1.49 8.12 1.34 11.93 59.17Continued on next page61AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERNereidae 1.23 0.76 4.14 1.04 6.08 65.24Idotea resecata 1.20 1.39 3.74 0.84 5.49 70.73Tanaidacea 0.31 1.09 3.58 1.12 5.26 75.98Mytilus spp. 0.37 1.18 2.73 0.82 4.01 79.99Gastropoda 0.48 0.24 2.41 0.76 3.53 83.53Nematoda 0.06 0.57 1.71 0.60 2.52 86.04Acari 0.00 0.39 1.42 0.61 2.08 88.13Lottia pelta 0.11 0.16 1.07 0.49 1.58 89.70Copepoda 0.00 0.23 0.95 0.43 1.39 91.10Groups WI & CBAverage dissimilarity = 65.12Group WI Group CBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Pycnogonida 0.43 1.49 7.67 1.12 11.77 11.77Amphipoda 1.76 0.94 5.60 0.93 8.59 20.37Nereidae 1.77 0.76 5.31 1.18 8.15 28.52Continued on next page62AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERCaprella sp. 1.72 0.80 5.27 1.30 8.10 36.62Nematoda 0.44 0.57 4.56 0.99 7.01 43.63Tanaidacea 0.30 1.09 4.34 1.32 6.67 50.30Tubeworm 1.20 0.11 3.98 1.12 6.11 56.41Idotea resecata 1.12 1.39 3.79 0.80 5.83 62.24Mytilus spp. 2.16 1.18 3.53 0.98 5.42 67.65Gastropoda 0.97 0.24 3.03 0.93 4.65 72.31Unidentified Bivalvia 0.81 0.21 2.99 0.90 4.59 76.90Cirolana sp. 0.52 0.00 2.60 0.81 3.99 80.89Copepoda 0.21 0.23 2.11 0.56 3.24 84.13Acari 0.07 0.39 1.90 0.48 2.91 87.04Phyllaplysia taylori 0.54 0.00 1.51 0.71 2.33 89.37Lottia pelta 0.22 0.16 1.38 0.58 2.12 91.49Groups RP & CBAverage dissimilarity = 69.94Group RP Group CBContinued on next page63AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 2.65 0.80 11.06 1.43 15.81 15.81Pycnogonida 0.06 1.49 8.93 1.16 12.77 28.58Phyllaplysia taylori 1.92 0.00 7.25 1.41 10.36 38.94Idotea resecata 0.53 1.39 6.24 0.95 8.92 47.87Nereidae 1.56 0.76 5.88 1.15 8.41 56.28Amphipoda 1.20 0.94 5.73 1.06 8.19 64.47Tanaidacea 0.44 1.09 4.63 1.02 6.61 71.08Tubeworm 0.90 0.11 4.06 0.98 5.80 76.88Mytilus spp. 1.01 1.18 2.94 0.52 4.20 81.08Nematoda 0.55 0.57 2.26 0.51 3.24 84.32Gastropoda 0.27 0.24 1.68 0.62 2.40 86.71Acari 0.00 0.39 1.58 0.59 2.25 88.97Unidentified Bivalvia 0.22 0.21 1.32 0.59 1.89 90.86Groups NB & CBAverage dissimilarity = 56.08Continued on next page64AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERGroup NB Group CBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Pycnogonida 0.00 1.49 8.83 1.42 15.74 15.74Mytilus spp. 1.74 1.18 5.62 1.09 10.02 25.76Tanaidacea 0.30 1.09 5.30 1.32 9.46 35.22Acari 0.53 0.39 4.42 0.85 7.89 43.11Amphipoda 1.27 0.94 4.32 0.96 7.70 50.80Idotea resecata 1.85 1.39 4.17 0.68 7.44 58.25Caprella sp. 0.79 0.80 4.03 1.01 7.19 65.44Nereidae 0.82 0.76 4.03 1.04 7.18 72.62Nematoda 0.86 0.57 3.05 0.74 5.44 78.06Copepoda 0.25 0.23 2.44 0.51 4.35 82.41Gastropoda 0.20 0.24 1.93 0.69 3.45 85.86Unidentified Bivalvia 0.10 0.21 1.34 0.56 2.39 88.24Lottia pelta 0.04 0.16 0.99 0.43 1.76 90.00Groups May & AugustAverage dissimilarity = 58.26Continued on next page65AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPERGroup May Group AugustSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Mytilus spp. 0.27 2.31 8.60 1.67 14.76 14.76Caprella sp. 1.22 2.55 7.33 1.12 12.58 27.34Phyllaplysia taylori 0.36 1.75 5.66 0.84 9.72 37.06Nereidae 0.90 1.55 4.38 1.00 7.52 44.58Amphipoda 1.23 1.74 4.14 0.95 7.11 51.68Nematoda 0.20 0.79 4.13 0.92 7.09 58.77Idotea resecata 1.18 1.26 3.32 0.75 5.69 64.47Tanaidacea 0.38 0.59 3.15 0.87 5.41 69.87Tubeworm 0.19 0.69 2.72 0.68 4.67 74.54Gastropoda 0.28 0.58 2.36 0.90 4.05 78.59Pycnogonida 0.37 0.44 1.79 0.51 3.07 81.66Acari 0.26 0.13 1.76 0.50 3.02 84.68Unidentified Bivalvia 0.17 0.38 1.48 0.65 2.54 87.22Copepoda 0.23 0.05 1.16 0.40 2.00 89.22Cirripedia:Sessilia 0.01 0.18 0.82 0.45 1.41 90.63Continued on next page66AppendixA.FirstAppendixTable A.1 ? continued from previous pageAggregated SIMPER67Appendix BSecond AppendixRaw species data analyses.B.1 Diversity & EvennessTable B.1: ANODE-Chi of raw species richness by site and time.Df Deviance Resid. Df Resid. Dev Pr(>Chi)NULL 159.00 271.823site 4 24.16 155.00 247.661 0.0001time 1 55.59 154.00 192.075 0.0000site:time 4 12.82 150.00 179.260 0.0122Table B.2: ANOVA of rarefied species richness by site and time.Df Sum Sq Mean Sq F value Pr(>F)site 4 86.07 21.52 27.456 0.0000time 1 8.89 8.89 11.343 0.0010site:time 4 62.62 15.65 19.974 0.0000Residuals 137 107.37 0.78Table B.3: ANODE-Chi of raw species evenness by site and time.Df Deviance Resid. Df Resid. Dev Pr(>Chi)NULL 153.00 1428.569site 4 248.32 149.00 1180.248 0.0000time 1 338.32 148.00 841.928 0.0000site:time 4 221.24 144.00 620.692 0.000068B.1. Diversity & Evenness# of Species05101520AB BAB AB AMayBCABABAugustSitesRarefied Richness246810DC WI RP NB CBABA AADC WI RP NB CBABCABBCCABFigure B.1: Mean richness of raw species data (A) and rarefied richness ofall raw species (B). All statistical values shown are an interaction betweensite and time. Pairwise comparisons were made within each time amongsites (Tukey and General Linear Hypotheses, p < 0.05)69B.1. Diversity & EvennessSitesDiversity0.20.40.60.81.0DC WI RP NB CBBCA ABBMayDC WI RP NB CBAB ABABCAugustFigure B.2: Simpson?s Index of raw species data diversity across all sitesin May and August 2012. All pairwise comparisons are within time periodamong sites (General Linear Hypotheses, p < 0.05).70B.1. Diversity & EvennessSitesSimpson's Evenness0.20.40.60.8DC WI RP NB CBADBACMayDC WI RP NB CBA AABCAugustFigure B.3: Average community evenness, Simpson?s Evenness measure ofall sites in May and August with raw species data. All pairwise comparisonsare within time period among sites (General Linear Hypotheses, p < 0.05).71B.1. Diversity & EvennessMay?0.4 0.0 0.2 0.4?0.50.00.5DCRPWINBCBAugustPCoA 2?0.4 0.0 0.2 0.4DCRPWICBNBFigure B.4: Beta diversity as Bray-Curtis distance for each site and timewith raw species data.72B.1. Diversity & EvennessDC WI RP NB CB0.00.20.40.60.8MayAABAAADC WI RP NB CBAugustDistance to centroidBAABABABB     SitesFigure B.5: Mean distance to centroid, Bray-Curtis beta diversity of all sitesin May (A) and August (B). Differences in variation among sites within eachtime detected with post-hoc Tukey (p < 0.05).73B.2. Community CompositionB.2 Community Composition74B.2.CommunityCompositionTable B.4: Similarity percentage analysis of raw species datawith pairwise comparisons between sites and times.Raw SIMPERGroups DC & WIAverage dissimilarity = 62.10Group DC Group WISpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 3.45 1.72 6.79 1.31 10.93 10.93Phyllaplysia taylori 2.82 0.54 6.10 1.67 9.83 20.76Aoroides columbidae 2.01 1.10 4.15 1.04 6.68 27.44Mytilus spp. 0.37 2.16 4.05 1.13 6.52 33.95Nereid C 0.55 1.00 2.77 0.99 4.46 38.41Unidentified Amph 2 1.05 0.50 2.65 1.03 4.27 42.69Photis brevipes 0.91 1.05 2.50 0.91 4.03 46.72Tubeworm 0.00 1.20 2.47 0.96 3.98 50.70Corophium sp. 0.24 0.79 2.45 0.88 3.94 54.64Lacuna spp. 0.21 0.66 2.26 1.04 3.65 58.29Continued on next page75B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERNematode A 0.06 0.44 2.22 0.68 3.57 61.86Unidentified Bivalvia 0.03 0.81 2.08 0.95 3.35 65.22Idotea resecata 1.20 1.12 1.88 0.77 3.03 68.25Amphithoe sp. 0.14 0.69 1.87 0.84 3.02 71.27Cirolana sp. 0.00 0.52 1.79 0.85 2.88 74.15Tanaidacea 0.31 0.30 1.68 0.76 2.71 76.85Eogammarus confervicolus 0.28 0.35 1.67 0.69 2.69 79.54Nereid B 0.83 1.06 1.32 0.68 2.12 81.66Nereid D 0.08 0.49 1.21 0.58 1.94 83.60Copepoda 0.00 0.21 1.06 0.46 1.70 85.30Pycnogonida 0.06 0.43 0.92 0.67 1.49 86.79Margarites spp. 0.27 0.47 0.88 0.67 1.42 88.21Nereid H 0.07 0.13 0.81 0.42 1.31 89.53Pugettia spp. 0.22 0.09 0.62 0.49 1.00 90.53Groups DC & RPAverage dissimilarity = 56.08Group DC Group RPContinued on next page76B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Phyllaplysia taylori 2.82 1.92 5.59 0.81 9.97 9.97Aoroides columbidae 2.01 1.03 5.11 1.03 9.11 19.08Caprella sp. 3.45 2.65 4.84 0.69 8.63 27.72Unidentified Amph 2 1.05 0.03 3.88 1.36 6.92 34.63Idotea resecata 1.20 0.53 3.69 1.22 6.59 41.22Nereid C 0.55 1.00 3.40 1.32 6.07 47.28Nereid D 0.08 0.71 3.30 0.96 5.89 53.18Photis brevipes 0.91 0.41 3.11 1.03 5.55 58.72Tubeworm 0.00 0.90 2.99 1.05 5.34 64.06Nereid B 0.83 0.37 2.68 0.99 4.79 68.85Mytilus spp. 0.37 1.01 2.01 0.75 3.58 72.42Tanaidacea 0.31 0.44 1.85 0.80 3.30 75.72Lacuna spp. 0.21 0.27 1.56 0.67 2.79 78.51Eogammarus confervicolus 0.28 0.03 1.39 0.55 2.48 80.99Corophium sp. 0.24 0.04 1.20 0.52 2.13 83.12Nematode B 0.00 0.51 1.16 0.61 2.07 85.19Pugettia spp. 0.22 0.03 0.79 0.46 1.41 86.61Amphithoe sp. 0.14 0.06 0.78 0.40 1.39 88.00Continued on next page77B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERMargarites spp. 0.27 0.00 0.73 0.53 1.30 89.30Unidentified Bivalvia 0.03 0.22 0.59 0.44 1.04 90.34Groups WI & RPAverage dissimilarity = 69.29Group WI Group RPSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 1.72 2.65 7.34 1.24 10.59 10.59Phyllaplysia taylori 0.54 1.92 4.68 1.48 6.75 17.34Nereid D 0.49 0.71 4.38 1.25 6.32 23.66Tubeworm 1.20 0.90 3.71 1.08 5.36 29.02Mytilus spp. 2.16 1.01 3.52 0.79 5.08 34.10Aoroides columbidae 1.10 1.03 3.34 0.94 4.82 38.91Idotea resecata 1.12 0.53 3.27 0.96 4.72 43.64Corophium sp. 0.79 0.04 3.12 0.93 4.50 48.13Nereid C 1.00 1.00 2.96 0.99 4.27 52.41Photis brevipes 1.05 0.41 2.81 1.01 4.05 56.46Continued on next page78B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERNereid B 1.06 0.37 2.77 1.01 3.99 60.45Nematode A 0.44 0.03 2.37 0.63 3.42 63.87Unidentified Bivalvia 0.81 0.22 2.32 0.92 3.34 67.21Lacuna spp. 0.66 0.27 2.13 0.86 3.07 70.29Amphithoe sp. 0.69 0.06 2.12 0.89 3.06 73.35Tanaidacea 0.30 0.44 1.99 0.78 2.87 76.22Cirolana sp. 0.52 0.00 1.98 0.82 2.86 79.08Eogammarus confervicolus 0.35 0.03 1.46 0.58 2.11 81.19Unidentified Amph 2 0.50 0.03 1.42 0.76 2.05 83.24Copepoda 0.21 0.00 1.18 0.44 1.71 84.95Margarites spp. 0.47 0.00 1.09 0.59 1.57 86.52Pycnogonida 0.43 0.06 1.07 0.64 1.54 88.06Nematode B 0.00 0.51 1.01 0.61 1.46 89.52Nereid H 0.13 0.03 0.75 0.39 1.08 90.60Groups DC & NBAverage dissimilarity = 69.34Group DC Group NBContinued on next page79B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 3.45 0.79 10.58 2.40 15.26 15.26Phyllaplysia taylori 2.82 0.00 9.79 1.60 14.12 29.38Aoroides columbidae 2.01 0.28 7.27 1.92 10.48 39.86Mytilus spp. 0.37 1.74 5.53 1.81 7.97 47.84Idotea resecata 1.20 1.85 4.24 0.95 6.12 53.95Unidentified Amph 2 1.05 0.46 3.22 1.16 4.64 58.59Photis brevipes 0.91 1.08 2.76 0.99 3.97 62.57Acari 0.00 0.53 2.65 0.72 3.83 66.40Nematode A 0.06 0.86 2.59 0.91 3.74 70.14Nereid B 0.83 0.57 2.21 0.85 3.19 73.33Nereid C 0.55 0.20 2.10 0.87 3.03 76.36Tanaidacea 0.31 0.30 1.95 0.79 2.81 79.17Corophium sp. 0.24 0.25 1.51 0.64 2.18 81.35Eogammarus confervicolus 0.28 0.03 1.39 0.56 2.01 83.36Lacuna spp. 0.21 0.20 1.39 0.58 2.00 85.36Copepoda 0.00 0.25 1.24 0.37 1.79 87.15Amphithoe sp. 0.14 0.20 1.09 0.53 1.57 88.72Pugettia spp. 0.22 0.00 0.82 0.49 1.18 89.90Continued on next page80B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERMargarites spp. 0.27 0.00 0.79 0.56 1.14 91.04Groups WI & NBAverage dissimilarity = 65.52Group WI Group NBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Mytilus spp. 2.16 1.74 4.59 1.34 7.01 7.01Idotea resecata 1.12 1.85 4.42 1.01 6.75 13.75Nematode A 0.44 0.86 4.22 1.34 6.45 20.20Caprella sp. 1.72 0.79 4.10 1.31 6.25 26.45Nereid C 1.00 0.20 3.77 1.20 5.76 32.21Aoroides columbidae 1.10 0.28 3.39 1.23 5.18 37.39Tubeworm 1.20 0.00 3.11 0.94 4.75 42.14Corophium sp. 0.79 0.25 2.88 0.94 4.40 46.54Unidentified Bivalvia 0.81 0.10 2.60 1.04 3.97 50.51Acari 0.07 0.53 2.59 0.69 3.95 54.46Photis brevipes 1.05 1.08 2.49 0.87 3.81 58.27Continued on next page81B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERLacuna spp. 0.66 0.20 2.48 1.02 3.78 62.05Nereid B 1.06 0.57 2.42 0.94 3.70 65.75Cirolana sp. 0.52 0.09 2.00 0.84 3.06 68.81Copepoda 0.21 0.25 1.92 0.53 2.94 71.74Amphithoe sp. 0.69 0.20 1.86 0.79 2.84 74.58Tanaidacea 0.30 0.30 1.81 0.66 2.76 77.34Unidentified Amph 2 0.50 0.46 1.71 0.85 2.61 79.95Nereid D 0.49 0.04 1.47 0.57 2.24 82.19Eogammarus confervicolus 0.35 0.03 1.47 0.60 2.24 84.44Phyllaplysia taylori 0.54 0.00 1.34 0.73 2.04 86.48Margarites spp. 0.47 0.00 1.18 0.61 1.79 88.27Pycnogonida 0.43 0.00 1.17 0.68 1.78 90.06Groups RP & NBAverage dissimilarity = 78.20Group RP Group NBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Continued on next page82B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERCaprella sp. 2.65 0.79 9.77 1.91 12.49 12.49Idotea resecata 0.53 1.85 7.55 1.14 9.66 22.15Phyllaplysia taylori 1.92 0.00 6.78 1.35 8.66 30.82Mytilus spp. 1.01 1.74 6.23 0.99 7.97 38.78Nereid C 1.00 0.20 4.24 1.29 5.42 44.21Photis brevipes 0.41 1.08 4.17 1.03 5.33 49.53Aoroides columbidae 1.03 0.28 4.13 1.15 5.29 54.82Nereid D 0.71 0.04 3.66 0.92 4.68 59.50Tubeworm 0.90 0.00 3.57 1.07 4.56 64.06Nematode A 0.03 0.86 3.51 0.79 4.49 68.55Acari 0.00 0.53 3.03 0.66 3.88 72.43Nereid B 0.37 0.57 2.94 0.90 3.76 76.18Tanaidacea 0.44 0.30 2.41 0.81 3.08 79.27Unidentified Amph 2 0.03 0.46 2.04 0.78 2.60 81.87Lacuna spp. 0.27 0.20 1.69 0.67 2.16 84.03Nematode B 0.51 0.00 1.50 0.61 1.92 85.95Copepoda 0.00 0.25 1.38 0.35 1.76 87.72Corophium sp. 0.04 0.25 1.38 0.54 1.76 89.48Continued on next page83B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERAmphithoe sp. 0.06 0.20 1.11 0.51 1.42 90.90Groups DC & CBAverage dissimilarity = 74.86Group DC Group CBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 3.45 0.80 10.92 1.89 14.59 14.59Phyllaplysia taylori 2.82 0.00 9.38 1.68 12.53 27.12Aoroides columbidae 2.01 0.22 8.17 1.67 10.92 38.04Pycnogonida 0.06 1.49 6.82 1.33 9.11 47.15Unidentified Amph 2 1.05 0.40 3.54 1.13 4.73 51.88Idotea resecata 1.20 1.39 3.14 0.83 4.19 56.07Tanaidacea 0.31 1.09 3.09 1.11 4.13 60.20Photis brevipes 0.91 0.73 3.07 0.93 4.11 64.31Nereid B 0.83 0.25 2.56 0.94 3.41 67.73Mytilus spp. 0.37 1.18 2.37 0.82 3.17 70.89Nereid C 0.55 0.55 2.29 0.86 3.05 73.95Continued on next page84B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERLacuna spp. 0.21 0.24 1.70 0.68 2.27 76.22Eogammarus confervicolus 0.28 0.04 1.57 0.55 2.10 78.31Nematode A 0.06 0.57 1.50 0.59 2.01 80.32Corophium sp. 0.24 0.11 1.42 0.56 1.89 82.21Acari 0.00 0.39 1.21 0.61 1.62 83.83Amphithoe sp. 0.14 0.10 1.00 0.43 1.33 85.17Lottia pelta 0.11 0.16 0.92 0.48 1.22 86.39Copepoda 0.00 0.23 0.79 0.44 1.05 87.44Pugettia spp. 0.22 0.00 0.78 0.46 1.04 88.49Margarites spp. 0.27 0.00 0.71 0.55 0.95 89.44Nereid D 0.08 0.06 0.71 0.35 0.94 90.38Groups WI & CBAverage dissimilarity = 72.23Group WI Group CBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Pycnogonida 0.43 1.49 6.54 1.07 9.06 9.06Continued on next page85B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERCaprella sp. 1.72 0.80 4.44 1.30 6.15 15.20Nematode A 0.44 0.57 3.93 0.97 5.44 20.65Nereid C 1.00 0.55 3.86 1.02 5.34 25.99Tanaidacea 0.30 1.09 3.63 1.32 5.03 31.02Aoroides columbidae 1.10 0.22 3.56 1.07 4.92 35.94Corophium sp. 0.79 0.11 3.43 0.92 4.75 40.69Tubeworm 1.20 0.11 3.37 1.12 4.66 45.36Idotea resecata 1.12 1.39 3.19 0.78 4.41 49.77Mytilus spp. 2.16 1.18 2.99 0.98 4.14 53.90Nereid B 1.06 0.25 2.73 0.97 3.77 57.68Unidentified Bivalvia 0.81 0.21 2.49 0.94 3.45 61.12Amphithoe sp. 0.69 0.10 2.33 0.92 3.23 64.36Photis brevipes 1.05 0.73 2.32 0.76 3.22 67.57Cirolana sp. 0.52 0.00 2.17 0.80 3.01 70.58Lacuna spp. 0.66 0.24 1.97 0.80 2.73 73.31Copepoda 0.21 0.23 1.77 0.57 2.45 75.77Acari 0.07 0.39 1.68 0.44 2.33 78.10Eogammarus confervicolus 0.35 0.04 1.65 0.58 2.29 80.38Nereid D 0.49 0.06 1.59 0.59 2.20 82.59Continued on next page86B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERUnidentified Amph 2 0.50 0.40 1.49 0.75 2.06 84.65Phyllaplysia taylori 0.54 0.00 1.25 0.72 1.73 86.38Lottia pelta 0.22 0.16 1.16 0.56 1.61 87.99Margarites spp. 0.47 0.00 1.07 0.61 1.48 89.47Nereid H 0.13 0.13 1.00 0.47 1.39 90.86Groups RP & CBAverage dissimilarity = 77.51Group RP Group CBSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Caprella sp. 2.65 0.80 10.02 1.47 12.93 12.93Pycnogonida 0.06 1.49 8.15 1.12 10.52 23.44Phyllaplysia taylori 1.92 0.00 6.60 1.40 8.51 31.96Idotea resecata 0.53 1.39 5.72 0.91 7.38 39.34Nereid C 1.00 0.55 4.50 1.21 5.80 45.15Aoroides columbidae 1.03 0.22 4.36 1.02 5.63 50.78Tanaidacea 0.44 1.09 4.22 1.02 5.44 56.22Continued on next page87B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERNereid D 0.71 0.06 3.97 0.86 5.12 61.34Tubeworm 0.90 0.11 3.69 1.00 4.76 66.10Photis brevipes 0.41 0.73 2.71 0.71 3.49 69.59Mytilus spp. 1.01 1.18 2.70 0.52 3.48 73.07Nematode A 0.03 0.57 2.08 0.47 2.68 75.75Nereid B 0.37 0.25 1.85 0.64 2.39 78.14Lacuna spp. 0.27 0.24 1.52 0.62 1.96 80.10Unidentified Amph 2 0.03 0.40 1.46 0.62 1.89 81.99Acari 0.00 0.39 1.43 0.58 1.84 83.83Nematode B 0.51 0.03 1.38 0.61 1.77 85.60Unidentified Bivalvia 0.22 0.21 1.21 0.59 1.56 87.17Copepoda 0.00 0.23 0.89 0.43 1.14 88.31Lottia pelta 0.00 0.16 0.80 0.37 1.04 89.35Amphithoe sp. 0.06 0.10 0.79 0.38 1.02 90.37Groups NB & CBAverage dissimilarity = 61.65Group NB Group CBContinued on next page88B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Pycnogonida 0.00 1.49 8.19 1.35 13.29 13.29Mytilus spp. 1.74 1.18 5.18 1.07 8.39 21.69Tanaidacea 0.30 1.09 4.80 1.31 7.79 29.48Acari 0.53 0.39 4.10 0.84 6.65 36.12Idotea resecata 1.85 1.39 3.87 0.65 6.29 42.41Caprella sp. 0.79 0.80 3.70 1.00 5.99 48.40Photis brevipes 1.08 0.73 3.29 0.89 5.33 53.73Nematode A 0.86 0.57 2.77 0.72 4.50 58.23Nereid B 0.57 0.25 2.69 0.87 4.36 62.59Nereid C 0.20 0.55 2.22 0.77 3.60 66.20Copepoda 0.25 0.23 2.22 0.50 3.59 69.79Unidentified Amph 2 0.46 0.40 2.07 0.81 3.36 73.15Lacuna spp. 0.20 0.24 1.77 0.68 2.87 76.02Aoroides columbidae 0.28 0.22 1.71 0.66 2.77 78.79Corophium sp. 0.25 0.11 1.59 0.58 2.58 81.37Amphithoe sp. 0.20 0.10 1.30 0.56 2.11 83.48Unidentified Bivalvia 0.10 0.21 1.22 0.56 1.98 85.46Lottia pelta 0.04 0.16 0.92 0.42 1.50 86.96Continued on next page89B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERNemertea 0.11 0.11 0.76 0.42 1.24 88.20Nereid D 0.04 0.06 0.65 0.30 1.06 89.26Cirripedia:Sessilia 0.13 0.07 0.64 0.42 1.04 90.30Groups May & AugustAverage dissimilarity = 65.27Group May Group AugustSpecies Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.%Mytilus spp. 0.27 2.31 7.48 1.65 11.46 11.46Caprella sp. 1.22 2.55 6.25 1.12 9.57 21.03Phyllaplysia taylori 0.36 1.75 4.77 0.83 7.31 28.34Photis brevipes 0.50 1.17 3.48 1.07 5.33 33.67Nereid B 0.18 1.05 3.37 1.20 5.16 38.83Nematode A 0.20 0.58 3.06 0.73 4.68 43.51Idotea resecata 1.18 1.26 2.91 0.70 4.46 47.97Tanaidacea 0.38 0.59 2.79 0.85 4.27 52.24Nereid C 0.53 0.79 2.78 0.90 4.26 56.50Continued on next page90B.2.CommunityCompositionTable B.4 ? continued from previous pageRaw SIMPERAoroides columbidae 0.78 1.07 2.54 0.79 3.90 60.40Tubeworm 0.19 0.69 2.29 0.68 3.50 63.90Unidentified Amph 2 0.27 0.71 2.14 0.89 3.29 67.19Nereid D 0.30 0.25 1.91 0.55 2.92 70.11Lacuna spp. 0.28 0.35 1.70 0.77 2.60 72.71Pycnogonida 0.37 0.44 1.60 0.48 2.45 75.16Acari 0.26 0.13 1.58 0.50 2.43 77.59Corophium sp. 0.41 0.16 1.38 0.66 2.11 79.70Amphithoe sp. 0.16 0.31 1.30 0.66 2.00 81.70Unidentified Bivalvia 0.17 0.38 1.27 0.64 1.94 83.64Copepoda 0.23 0.05 1.01 0.39 1.55 85.19Margarites spp. 0.00 0.30 0.81 0.50 1.24 86.43Eogammarus confervicolus 0.24 0.05 0.76 0.50 1.17 87.59Cirripedia:Sessilia 0.01 0.18 0.72 0.44 1.10 88.69Nematode B 0.00 0.22 0.71 0.37 1.08 89.77Lottia pelta 0.05 0.16 0.64 0.43 0.99 90.7691B.2. Community CompositionTable B.5: PERMANOVA of raw species composition data.Source df SS MS Pseudo-F P(perm) Unique perms P(MC)site 4 37.893 9.4731 2.2745 0.0148 7302 0.0028time 1 16.457 16.457 33.561 0.0001 9929 0.0001site x time 4 16.66 4.165 8.4941 0.0001 9827 0.0001Res 150 73.551 0.49034Total 159 144.5692B.2. Community CompositionFigure B.6: nMDS of raw species data epifaunal communities across all sitesin May and August.93

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.24.1-0166841/manifest

Comment

Related Items