UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Morphological and functional diversification during kelp evolution Starko, Samuel 2019

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

Item Metadata

Download

Media
24-ubc_2019_may_starko_samuel.pdf [ 15.78MB ]
Metadata
JSON: 24-1.0378025.json
JSON-LD: 24-1.0378025-ld.json
RDF/XML (Pretty): 24-1.0378025-rdf.xml
RDF/JSON: 24-1.0378025-rdf.json
Turtle: 24-1.0378025-turtle.txt
N-Triples: 24-1.0378025-rdf-ntriples.txt
Original Record: 24-1.0378025-source.json
Full Text
24-1.0378025-fulltext.txt
Citation
24-1.0378025.ris

Full Text

  Morphological and functional diversification during kelp evolution   by   Samuel Starko BSc., University of British Columbia, 2013   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  The Faculty of Graduate and Postdoctoral Studies  (Botany)  THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)  April 2019  © Samuel Starko, 2019           ii The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:     Morphological and functional diversification during kelp evolution    submitted by Samuel Starko in partial fulfillment of the requirements for the degree of  Doctor of Philosophy             in Botany             Examining Committee: Dr. Patrick Martone Supervisor  Dr. Sean Graham Supervisory Committee Member  Dr. Christopher Harley Supervisory Committee Member Dr. Quentin Cronk  Supervisory Committee Member Dr. Brian Leander  University Examiner Dr. Mary O’Connor  University Examiner Dr. Heroen Verbruggen  External Examiner      iii Abstract  Kelps are highly successful ecosystem engineers that substantially increase the productivity of nearshore ecosystems, forming nursery habitat for other species. Despite the importance of kelps to the modern ecology of temperate ecosystems, we have a limited understanding of their evolutionary relationships, the diversification dynamics that led to their modern distributions, and the factors that determine where they are found in nature. In this dissertation, I quantitatively explore how distributions and modern ecological strategies have arisen through kelp evolution. I utilize phylogenomic approaches to determine the evolutionary relationships between kelp species and the timing of their diversification, elucidating many previously unknown phylogenetic relationships between kelps (Ch. 2). I then investigate the strategies that kelps use to survive on wave-swept shores and test for trade-offs that may have constrained their broad range of morphological variation. I use flow tank experiments and field mechanical testing to investigate how external forces from the environment interact with morphology. I show that kelps with high mechanical support also experience greater environmental forces than weaker, more streamlined species, consistent with well known trade-offs of stress resistance in other organisms (Ch. 3). I also investigate the interspecific scaling of biomass allocation in organs of kelp that are analogous to those of land plants and showed that there are shared features of how size influences morphology in both groups (Ch. 4). Lastly, I combine trait and phylogenetic information to explore patterns of trait evolution and determine how species that specialize in different environments are distributed across the phylogeny. I assess whether phylogenetic relatedness or trait differences explain the assembly of kelp   iv communities and demonstrate that kelp communities are composed of distantly related species that have converged on similar traits (Ch. 5). Taken together, these studies offer a multifaceted perspective on the morphological and ecological diversification of kelps and demonstrate that ecological strategies are convergent among phylogenetically divergent lineages.                    v Lay Summary  Kelps are highly successful marine photosynthesizers that substantially increase the productivity of nearshore ecosystems and form nursery habitat for other species. Despite the importance of kelps to the modern ecology of temperate ecosystems, we have a limited understanding of their evolutionary relationships, the evolutionary dynamics that led to their modern distributions, and the factors that determine where they are found in nature. In this dissertation, I quantitatively explore how environmental conditions have shaped the evolutionary dynamics and ecological strategies of kelps. I do so across multiple spatial and temporal scales, from the patterns of kelp evolution through deep time, to the development of individual species. Together, these results illustrate how the evolutionary dynamics of kelps have been shaped by environmental drivers.             vi Preface  Much of the work in this dissertation has been published or submitted to peer-reviewed scientific journals and some of the work was done in collaboration with other researchers. Below is a description of each research chapter, outlining my contributions.  Chapter 2: The sampling scheme for this chapter was designed by me, with input from K. Demes, M. Soto Gomez, S. Graham and P. Martone. The work was carried out by me and M. Soto Gomez with support from S. Graham and P. Martone. H. Darby also contributed to laboratory work. H. Kawai, N. Yotsukura, S. Lindstrom provided samples and guidance, and H. Kawai conducted some (n = 3) of the DNA extractions. The paper was written by me with input from all authors. A version of this chapter has been submitted to a peer reviewed journal as: Starko, S., Soto Gomez, M., Darby, H., Kawai, H., Yotsukura, N., Lindstrom, S.C., Keeling, P., Graham, S.W., Martone, P.T. (In revision) A comprehensive kelp phylogeny sheds light on the evolution of an ecosystem.   Chapter 3: Under the guidance of P. Martone, I designed this study, collected and compiled the data, performed statistical analysis and visualization. I also wrote the paper with input from P. Martone. A version of this chapter is published as: Starko, S., & Martone, P. T. (2016). Evidence of an evolutionary-developmental trade-off between drag avoidance and tolerance in wave-swept, intertidal kelps. Journal of Phycology. 52 (1): 54 – 63.   vii Chapter 4: Under the guidance of P. Martone, I designed this study, collected and compiled the data, performed statistical analysis and visualization. I also wrote the paper with input from P. Martone. A version of this chapter is published as: Starko, S., & Martone, P. T. (2016). An empirical test of ‘universal’ biomass scaling relationships in kelps: evidence of convergence with seed plants. New Phytologist, 212 (3), 719 – 729.  Chapter 5: This chapter will be submitted for publication to a peer reviewed journal. This study was designed by me with input from K. Demes and P. Martone. Under the guidance of P. Martone, I collected samples, measured quantitative traits, compiled and analyzed the data. K. Demes provided some of the mechanical data used. I wrote the manuscript with input from all authors.            viii Table of Contents Abstract	..................................................................................................................................	iii	Lay	Summary	...........................................................................................................................	v	Preface	....................................................................................................................................	vi	Table	of	Contents	..................................................................................................................	viii	List	of	Tables	.........................................................................................................................	xiii	List	of	Figures	........................................................................................................................	xiv	List	of	Symbols	&	Abbreviations	............................................................................................	xvi	Acknowledgements	...............................................................................................................	xix	Dedication	............................................................................................................................	xxi	1.	Introduction	.........................................................................................................................	1	1.1	General	background	..............................................................................................................	1	1.2	A	little	known	branch	on	the	tree	of	life	................................................................................	3	1.3	Morphogenesis	and	morphological	diversity	........................................................................	5	1.4	Water	motion	by	waves	and	currents	...................................................................................	6	1.5	Niche	evolution	.....................................................................................................................	8	1.6	Dissertation	structure	&	overview	.........................................................................................	9	2.	A	comprehensive	kelp	phylogeny	sheds	light	on	the	evolution	of	an	ecosystem	.................	11	2.1	Synopsis	...............................................................................................................................	11	2.2	Introduction	.........................................................................................................................	12	  ix 2.3	Materials	&	Methods	..........................................................................................................	17	2.3.1	Taxon	and	gene	sampling	............................................................................................	17	2.3.2	Library	preparation	and	sequencing	............................................................................	18	2.3.3	Assembly,	alignment	and	phylogenetic	inference	.......................................................	19	2.3.4	Divergence	time	estimation	.........................................................................................	21	2.3.5	Diversification	analyses	...............................................................................................	22	2.3.6	Historical	biogeography	...............................................................................................	23	2.3.7	Morphological	ancestral	state	reconstruction	.............................................................	23	2.4	Results	and	Discussion	........................................................................................................	24	2.4.1	Data	characteristics	.....................................................................................................	24	2.4.2	Phylogenetic	relationships	among	kelps	.....................................................................	25	2.4.3	Temporal	and	spatial	patterns	.....................................................................................	31	2.4.4	Evolution	and	assembly	of	kelp	forest	ecosystems	.....................................................	39	2.4.5	Taxonomic	implications	...............................................................................................	44	2.4.6	Conclusions	..................................................................................................................	47	3.	Drag	avoidance	and	tolerance	strategies	of	sympatric	kelp	species:	evidence	of	a	trade-off48	3.1	Synopsis	...............................................................................................................................	48	3.2	Introduction	.........................................................................................................................	49	3.3	Methods	..............................................................................................................................	52	3.3.1	Kelp	collections	and	area	quantification	......................................................................	52	3.3.2	Tenacity	.......................................................................................................................	53	3.3.4	Drag	.............................................................................................................................	54	3.3.5	Streamlining	and	tolerance	indices	.............................................................................	55	  x 3.3.6	Predicting	avoidance	and	tolerance	at	different	sizes	.................................................	56	3.3.7	Second	moment	of	area	..............................................................................................	56	3.3.8	Phylogenetic	comparative	methods	and	statistics	......................................................	57	3.4	Results	.................................................................................................................................	58	3.5	Discussion	............................................................................................................................	64	3.5.1	Support	for	a	streamlining-tolerance	trade-off	...........................................................	64	3.5.2	Factors	driving	the	trade-off	........................................................................................	65	3.5.3	Ecoevolutionary	implications	.......................................................................................	68	3.5.4	Conclusions	..................................................................................................................	69	4.	An	empirical	test	of	"universal"	biomass	scaling	relationships	in	kelps:	evidence	of	convergence	with	seed	plants.	...............................................................................................	70	4.1	Synopsis	...............................................................................................................................	70	4.2	Introduction	.........................................................................................................................	71	4.2.1	Predicting	‘universal’	scaling’	.......................................................................................	72	4.2.2	Organ	biomass	scaling	in	seed	plants	..........................................................................	73	4.2.3	Testing	‘universal’	relationships	with	kelps	.................................................................	75	4.2.4	Aims	.............................................................................................................................	78	4.3	Methods	..............................................................................................................................	79	4.3.1	Allometric	analyses	......................................................................................................	79	4.3.2	Sample	collection	.........................................................................................................	80	4.3.3	Dry	weight	quantification	............................................................................................	81	4.3.4	Quantification	of	A	and	regression	with	MT	................................................................	84	4.3.5	Statistical	analysis	........................................................................................................	85	  xi 4.4	Results	and	Discussion	........................................................................................................	86	4.4.1	Scaling	of	A	and	MT	......................................................................................................	86	4.4.2	Biomass	partitioning	across	kelp	taxa	.........................................................................	87	4.4.3	Influence	of	habitat	on	biomass	allocation	..................................................................	94	4.4.5	“Diminishing	returns”	with	increasing	biomass	...........................................................	95	4.4.6	Conclusion	.................................................................................................................	100	5.	Convergent	evolution	of	niche	structure	in	northeast	Pacific	kelp	communities	...............	101	5.1	Synopsis	.............................................................................................................................	101	5.2	Introduction	.......................................................................................................................	102	5.3	Methods	............................................................................................................................	107	5.3.1	Trait	data	...................................................................................................................	107	5.3.2	Phylogenetic	relationships	.........................................................................................	110	5.3.1	Community	data	........................................................................................................	110	5.3.1	Testing	for	a	phylogenetic	signal	on	traits	and	community	assembly	.......................	111	5.4	Results	...............................................................................................................................	114	5.4.1	Ordination	and	phylogenetic	signal	of	functional	traits	............................................	114	5.4.2	Community	assembly	................................................................................................	118	5.5	Discussion	..........................................................................................................................	125	5.5.1	Niche	partitioning	along	a	disturbance	gradient	.......................................................	128	5.5.4	Conclusions	................................................................................................................	131	6.	Conclusions	......................................................................................................................	133	6.1	Key	findings	.......................................................................................................................	133	6.1.1	Environmental	factors	have	influenced	kelp	evolution	.............................................	134	  xii 6.1.2	Physical	limitations	have	constrained	morphological	evolution	...............................	136	6.1.3	Adaptations	are	convergently	distributed	across	the	phylogeny	..............................	138	6.2	Limitations	and	future	directions	......................................................................................	139	6.2.1	Intraspecific	variation	................................................................................................	139	6.2.2	Traits	and	productivity	...............................................................................................	140	6.2.3	Niche	modeling	..........................................................................................................	141	6.3	Final	thoughts	...................................................................................................................	142	Bibliography	.........................................................................................................................	144	Appendix	A	–	Supplementary	Information	for	Chapter	2	......................................................	168	Appendix	B	-	Supplementary	Information	for	Chapter	3	.......................................................	195	Appendix	C	-	Supplementary	Information	for	Chapter	4.	......................................................	203	References	for	Appendix	C	...................................................................................................	210	          xiii List of Tables  Table 3.1. Additional collection information .................................................................... 59	Table 4.1. Collection information for biomass partitioning analyses. .............................. 83	Table 4.2. Parameter estimates of the reduced major axis regression analyses ................ 90	Table 4.3. Standing biomass scaling exponents ................................................................ 92	Table 4.4. Log-transformed standing biomass scaling constants ..................................... 92	Table 5.1. Locations of field sites from which trait data were collected. ....................... 108	Table 5.2. Statistical testing of phylogenetic signal for quantitative traits. .................... 118	Table 5.3. PGLS models testing for correlations between traits and exposure. ............. 122             xiv List of Figures Fig 1.1. The three vegetative organs of a kelp. ................................................................... 6	Fig 2.1. Phylogenetic reconstruction of the kelps. ............................................................ 27	Fig 2.2. Historical biogeography of the complex kelps. ................................................... 29	Fig 2.3. Patterns of ocean climate and kelp diversification through deep time.. .............. 33	Fig 2.4. Likelihood reconstructions of morphological complexity in the kelps ............... 38	Fig 2.5. Lineage through time plots. ................................................................................. 41	Fig 3.1. Photographs of the species used in this study. ..................................................... 53	Fig 3.2. Schematic of the recirculating flume set-up used to measure drag. .................... 55	Fig 3.3. An example of closely related species with alternate strategies. ......................... 60	Fig 3.4 Negative correlation between tolerance and streamlining indices ....................... 61	Fig 3.5. Drag and tenacity for all species at 70 and 1150 cm2 .......................................... 62	Fig 3.6 Predicted drag and tenacity for 1150 cm2 kelps, plotted against I ....................... 62	Fig 3.7. Time calibrated phylogeny of the eight species used in this study ..................... 63	Fig 4.1. A comparison of the three convergent organs of land plants and kelps .............. 78	Fig 4.2. Pictorial and graphical representations of allometry applied to kelps ................. 80	Fig 4.3. Interspecific scaling relationships for dry mass versus total area ....................... 91	Fig 4.4. Interspecific scaling of kelp organ biomasses ..................................................... 93	Fig 4.5. Biomass scaling relationships of kelps from different habitats. .......................... 96	Fig 5.1. Phylogenetic distribution of trait axes in northeast Pacific kelp species ........... 116 Fig 5.2. Ancestral state reconstruction of continuous traits ............................................ 117 Fig 5.2. Principal component 1 versus streamlining and tolerance indices .................... 118   xv  Fig 5.3. Correlation matrix of species pairs .................................................................... 120	Fig 5.4. Species covariation versus phylogenetic and trait dissimilarity ........................ 120	Fig 5.5. The standard effect size of phylogeny-based indices ........................................ 121	Fig 5.6. Relationship between wave exposure and principal component 1 .................... 123	Fig 5.7. Local regression (loess) of species occupancy across a gradient of exposure. . 124	Fig 5.8. NMDS plot of community presence data .......................................................... 125	                  xvi List of Symbols & Abbreviations  A  Area ACP  Akkesiphycaceae, Pseudochordaceae and Chordaceae families AIC  Akaike information criterion BC  British Columbia BMA  Blade mass per area c.  Circa CD  Drag coefficient d  Diameter DEC  Dispersal-extinction-cladogenesis model DMC  Dry matter content D  Drag EI  Flexural stiffness E  Modulus of elasticity EO  Eocene-Oligocene boundary HPD  High probability density HMF  Holdfast mass fraction Hz  Hertz I  Second moment of area J  Jump dispersal parameter K  Blomberg's K statistic of phylogenetic signal   xvii LMA  Leaf mass per area LR  Likelihood ratio ML  Maximum likelihood ML  Leaf or blade mass MNTD Mean nearest taxon distance MPD  Mean paired distance MR  Root or holdfast mass MS  Stem or stipe mass MT  Total mass MY  Millions of years MYA  Millions of years ago N  North NMDS  Non-parametric multidimensional scaling OLS  Ordinary least squares PCA  Principal component analysis PC1  First principal component PC2  Second principal component PGLS  Phylogenetic generalized least squares RMA  Reduced major axis  S  Streamlining index SA  Surface area SES  Standard effect size SMA  Standardized major axis    xviii SMF  Stipe mass fraction T  Tolerance index U  Fluid velocity V  Volume W  West WBE  West, Brown and Enquist model α  Scaling exponent β  Scaling/allometric constant !18O  Proportion of oxygen consisting of 18O isotope ε  Tensile extensibility; breaking strain λ  Pagel's Lambda statistic of phylogenetic signal π  Mathematical constant (ratio of circle circumference to diameter) ρ  Fluid density ∂  Tensile breaking stress            xix Acknowledgements  I am truly grateful to have had the opportunity to undertake a PhD at a top-tier research university. The academic environment at UBC has been highly stimulating both academically and socially and I am lucky to have had the opportunity to study here. I would like to first acknowledge my supervisor, Patrick Martone, who has provided me with endless support and opportunity while allowing me the academic freedom to ask my own questions, design my own studies and grow as a scientist. I will always look back on my time under his supervision with fondness and gratitude. In the same vein, it is important to thank other members of the Martone lab who have both directly and indirectly made this dissertation possible and helped me to maintain my sanity through the ups and downs of a graduate degree. I acknowledge my committee, made up of Sean Graham, Quentin Cronk and Christopher Harley each of whom have served as mentors for me during my degree and have contributed to the conceptual development of this dissertation. I thank departmental staff and faculty for creating a safe, open and nurturing environment where students can focus on their studies with little worry of daily, departmental bureaucracy. I also thank the staff at Bamfield Marine Sciences Centre who have made my research experiences out there not only smooth, but ultimately life-changing. I thank my family for unconditionally supporting my decision to pursue graduate work. I thank my parents (Brooke Hatfield and Robert Starko), grandparents (Peter & Sally Hatfield and Joseph and Jean Starko), siblings (Molly & Ben Starko) and extended family for instilling a sense of wonder in me from a young age and for encouraging me to follow my dreams. In particular, I would like to acknowledge my grandfather, Peter Hatfield, who exposed me to the   xx west coast of British Columbia, helped inspire my passion for the ocean and taught me marine navigation and safety skills that have been crucial to my success in the field. I thank partners, friends and colleagues that have always been there for me and helped me to troubleshoot personal and professional challenges through my decade of post-secondary education. In particular, I thank my partner Lianna Gendall who has provided unconditional support through the (often stressful) process of writing and defending this dissertation. The work described in this dissertation was funded by a UBC 4-Year Fellowship, an Isaak W. Killam Doctoral Scholarship, Natural Sciences and Engineering Research Council (NSERC) Postgraduate Scholarship and Canadian Graduate Scholarship awards, Catalyst Paper Corporation Scholarship, NSERC Discovery Grants to P.T. Martone, S.W. Graham and P. Keeling. Collections conducted in Pacific Rim National Park were authorized by the Canadian Ministry of the Environment (Permit No. PRN-2015-18843).              xxi Dedication  This dissertation is dedicated to my family who inspired my sense of exploration at a young age and helped guide me to be who I am today.                     1 1. Introduction 1.1 General background Brown algae in the order Laminariales, commonly referred to as “kelps,” include the world’s largest and fastest growing algae. The giant kelp, Macrocystis pyrifera, can reach 50 m in length (Hoek et al., 1995) and the bull kelp, Nereocystis luetkeana, can grow up to 14 cm per day (Kain et al., 1987). The enormous structures formed by kelps act as underwater forests that are often compared to their terrestrial counterparts in terms of their productivity (Lüning & tom Dieck, 1990) and the diversity that they support (Steneck et al., 2002; Teagle et al., 2017). These marine forests are essential habitat for many nearshore organisms (Steneck et al., 2002; Graham, 2004; Teagle et al., 2017; Miller et al., 2018) including commercially important fish, shellfish and echinoderms (Teagle et al., 2017). Moreover, the rapid growth of kelps fuels coastal productivity through herbivory, secondary production of detritus and trophic transfer of kelp-derived carbon (Duggins et al., 1989; Fredriksen, 2003). Thus, kelps are foundation species around the world and are strongly tied to the ecology of a wide range of coastal organisms.   Kelps provide a range of ecosystem services to humans and have done so for a very long time (Steneck et al., 2002). Kelps are eaten all over the world and their extracts are used as animal feed, cosmetic products, medical products, health supplements and laboratory supplies (Bartsch et al., 2008; Bennett et al., 2016). Kelps further provide indirect services through wave attenuation and by playing an essential role in the life cycles of commercially important fish and invertebrate species (Steneck et al., 2002; Graham, 2004; Teagle et al., 2017). Even in regions with one or two species, kelps have substantial ecological and economic importance (Bennett et   2 al., 2016; Wernberg et al., 2019). For example, in Australia where only two kelp species (Macrocystis pyrifera and Ecklonia radiata) dominate subtidal regions, the economic value of this “Great Southern Reef” has been compared to that of the Great Barrier Reef (Bennett et al., 2016). This exemplifies the important role that kelps play in driving coastal productivity and the intimate connection between kelp forests and the coastal people that rely on the ecosystems that they support.  Despite the importance of kelps to the modern ecology of temperate ecosystems, we have an incomplete understanding of the evolutionary relationships between kelp species (Lane et al., 2006; Bolton, 2010; Kawai et al., 2013, 2016), which has limited our ability to draw inferences about the diversification dynamics that have led to their modern distributions (Estes & Steinberg, 1988). While not exceptionally speciose (~120 species), kelps are ecologically and morphologically diverse, and capable of living in a wide range of environments (Steneck et al., 2002). Studies describing the morphological adaptations of kelps and other seaweeds to various habitats have become increasingly common and have illustrated the importance of morphology in determining where species are found in the environment (e.g. Carrington, 1990, 2013; Gaylord et al., 1994; Friedland & Denny, 1995; Utter & Denny, 1996; Denny et al., 1997; Gaylord & Denny, 1997; Boller & Carrington, 2006; Martone, 2007a; Koehl et al., 2008b; Demes et al., 2011; Martone et al., 2012). However, we lack a thorough understanding of the evolutionary dynamics underlying morphological variation across kelps and about the factors that have influenced niche formation.     3 1.2 A little known branch on the tree of life Although kelp forests are often compared to terrestrial forests, kelps have an evolutionary history that is entirely distinct from land plants. Kelps (and all brown algae) are in the Kingdom Chromista and the Division Ochrophyta (Cavalier-Smith, 1998, 2018; Keeling, 2004). Where red and green algae share a common photosynthetic ancestor with the land plants, brown algae acquired photosynthesis horizontally, through endosymbiosis of a red alga by a eukaryotic ancestor of the ochrophytes. This is an example of a larger pattern of ‘serial endosymbiosis’ that has led to the dispersion of photosynthetic organelles (i.e., plastids) across the tree of life (McFadden, 2001; Keeling, 2004, 2010; Archibald, 2009). This endosymbiotic event between an ancestral ochrophyte and a red alga is believed to have also resulted in the lateral transfer of several metabolic pathways, such as the production of cellulose, that has resulted in similarities in cellular structure between ochrophytes and distantly related land plants (Cock et al., 2010; Michel et al., 2010).  Within the brown algae, kelps make up one of 19 currently recognized orders and 33 of the approximately 300 genera found worldwide (Silberfeld et al., 2010, 2014). Brown algae likely originated in the Jurassic (approximately 180 MYA) with ancestors of ten of these orders, known as the brown algal crown radiation, diversifying in the Lower Cretaceous (Silberfeld et al., 2010). All members of the brown algae are multicellular (Fritsch, 1935), but taxa possess an enormous range of morphologies and tissue structures, from thin, microscopic filaments to large parenchymatous or pseudoparenchymatous thalli (Graham et al., 2008). Kelps are regarded as the most productive of the major brown algal clades and are believed to have evolved recently as oceans cooled during the Cenozoic (Estes & Steinberg, 1988; Bolton, 2010). Alternative   4 hypotheses have been posed, however, with some authors arguing that kelps evolved as early as the Cretaceous (Domning, 1989) and others hypothesizing a mid-Miocene radiation (Estes & Steinberg, 1988, 1989). Kelps are found mostly in temperate regions of the world (Bolton, 2010; Teagle et al., 2017) and at broad scales are primarily limited in distribution by temperature and nutrients (Lüning & tom Dieck, 1990; Teagle et al., 2017; Filbee-Dexter & Wernberg, 2018). Kelps are generally intolerant of warm temperatures (Lüning & Freshwater, 1988; tom Dieck, 1993) and have high nitrogen demands (Gerard & Mann, 1979; Henley & Dunton, 1997; Stephens & Hepburn, 2016) keeping them largely out of the tropics (with the exception of deep water refugia; Graham et al., 2007). Along temperate coastlines that offer high nutrients and cool waters, kelps are ubiquitous globally (Bolton, 2010). Although widespread, their diversity is unambiguously greater in the North Pacific than anywhere else in the world (Estes & Steinberg, 1988; Bolton, 2010). In fact, all but one of the 33 currently recognized genera are found in the North Pacific (Bolton, 2010). This has led to the generally accepted hypothesis that kelps originated in the North Pacific and spread globally from there (Estes & Steinberg, 1988; Bolton, 2010). This would suggest that there have been multiple invasions of different oceans by kelps, because species outside of the North Pacific are likely not phylogenetically clustered (Bolton, 2010). Yet, a comprehensive phylogeny at an appropriate taxonomic scale is needed to understand the biogeographical history of kelps globally and to make inferences about lineage dispersal pathways.    5 1.3 Morphogenesis and morphological diversity Kelps have a heteromorphic sporic life cycle, whereby a large macroscopic sporophyte alternates with male and female gametophytes that are filamentous and microscopic (Graham et al., 2008). While gametophytes are largely cryptic (Fritsch, 1935; Kawai et al., 2013), sporophytes are highly diverse morphologically and exhibit a wide range of morphological and ecological strategies (Abbott & Hollenberg, 1976; Estes & Steinberg, 1988; Steneck et al., 2002; Druehl & Clarkston, 2016). Sporophytes are composed of three vegetative organ types: a blade, stipe and holdfast (Fig 1.1). Although many kelps are blade-like in morphology, with only one of each organ (as in Fig 1.1), some kelps are highly modular and have complex morphologies with several blades or stipes (Druehl & Clarkston, 2016). Some species even grow clonally along a rhizoid (Abbott & Hollenberg, 1976). Complex morphologies of kelps are strongly tied to their ability to exhibit apoptosis (Fritsch, 1935), forming longitudinal slits along blades and stipes to produce branching patterns by bisecting meristematic tissue. Some kelps also have buoyant floats which are produced through cell death and tissue cavitation within the medulla of stipe or blade tissue (Fritsch, 1935; Liggan & Martone, 2018). This forms air pockets that are filled with by-product gases of metabolism and allow kelp tissues to float towards the surface (Fritsch, 1935; Liggan & Martone, 2018). Regardless of the form that adult kelps eventually take, all kelps germinate as unbranched blades and some species increase in morphological complexity as they develop, leading to a wide range of possible adult morphologies. However, the drivers and trade-offs associated with this morphological diversity remain poorly understood.   6  Fig 1.1. The three vegetative organs of a kelp. Image of Laminaria yezoensis (suction cup kelp) showing the the blade, stipe and holdfast.  1.4 Water motion by waves and currents Water motion is an important environmental driver for virtually all organisms living in the marine photic zone (Hurd, 2000). Wave action can induce immense physical forces on organisms that live under crashing waves, influencing fitness by limiting both survivorship and reproductive success (Koehl, 1999; Denny & Gaylord, 2002; Duggins et al., 2003). At exposed sites on the outer coast, velocities under breaking waves are commonly between 10 and 20 ms-1, sometimes reaching as high as 25 ms-1 (Denny et al., 2003a). Organisms that cannot survive the immense forces imposed by breaking waves or flowing currents are dislodged or broken (Denny,   7 2006). This is believed to limit which species can live in wave-exposed areas (Koehl, 1999; Denny & Gaylord, 2002; Denny, 2006). In spite of potential for dislodgement or tissue damage, kelps and other seaweeds have a complex relationship with water motion because flow can also positively affect macroalgae growing on rocky shores (Hurd, 2000, 2017). Moving water can limit the effects of nutrient depleted boundary layers, that form around organisms as they absorb nutrients from the surrounding environment (Hurd, 2000, 2017). Thus, wave-exposed areas tend to be highly productive (Leigh et al., 1987) and host a variety of kelp species (Abbott & Hollenberg, 1976; Druehl & Clarkston, 2016).  The ability of kelps to survive in wave-exposed areas depends on the magnitude of the physical forces that they experience relative to the forces that they are able to resist. One key strategy used by kelps to resist dislodgement from waves is flexibility (Johnson & Koehl, 1994; Friedland & Denny, 1995; Utter & Denny, 1996). Kelps produce materials that are flexible, which allows them to fold up and reconfigure under breaking waves or in tidal currents, reducing the magnitude of forces that they experience (Johnson & Koehl, 1994; Utter & Denny, 1996). This may explain how kelps can grow to enormous sizes despite living in environments of extreme wave energy (Johnson & Koehl, 1994; Denny et al., 1997). Kelps are also capable of structural reinforcement as they develop, thickening stipes and growing larger holdfasts (Johnson & Koehl, 1994; Friedland & Denny, 1995). Different kelp species evolved distinct morphological strategies to resist dislodgement from waves. Exploring how physical forces and the ability of kelps to resist them scale through development, across different species, may lend insights into the factors that drive and/or constrain morphological variation on wave-exposed shores.    8 1.5 Niche evolution We have a limited understanding of the processes underlying assembly of kelp communities and the factors that determine where species are found on fine spatial scales. Given the importance of wave exposure as an environmental driver, and the strong selective pressures that wave forces impose on rocky shore organisms, it is likely that waves strongly influence where species are found in nature (Koehl, 1999; Denny & Gaylord, 2002; Denny, 2006) and likely play an important role in the assembly of communities. Indeed, previous work has suggested that certain kelp species tend to specialize in areas with high wave exposure and other species specialize in low wave areas (Dayton, 1973, 1985; Abbott & Hollenberg, 1976; Druehl & Elliot, 1996; Druehl & Clarkston, 2016). Nonetheless, it is unclear exactly what allows species to specialize in these different types of habitats. Morphological and biomechanical traits may influence the types of habitats in which species can live  (Wernberg, 2005; Wernberg & Thomsen, 2005; Demes et al., 2013). This hypothesis has mostly been tested by examining intraspecific variation caused by phenotypic plasticity (Johnson & Koehl, 1994; Duggins et al., 2003; Wernberg & Thomsen, 2005; Koehl et al., 2008b) or local adaptation (Roberson & Coyer, 2004; Lane et al., 2007; Augyte et al., 2017); correlations between traits and environmental distributions have yet to be assessed across kelp species. Studies documenting intraspecific morphological variation between non-breeding populations (e.g. ‘incipient species’) have further demonstrated that adaptation to different regimes of water motion may drive speciation through divergent selection (Roberson & Coyer, 2004; Lane et al., 2007; Augyte et al., 2017). If the dynamics observed within species are also responsible for morphological variation across the broader diversity of kelps, then we would expect closely related species to differ in habitat,   9 partitioning their distributions across different parts of the environment (as described in Cavender-Bares et al., 2004a; Silvertown et al., 2006 for embryophytes). Yet, the range of morphological strategies used by co-occurring kelp species on wave-swept shores is poorly understood, and it is unknown whether patterns of divergent selection observed across populations of the same species are widespread across the phylogeny.   1.6 Dissertation Structure & Overview The goal of this dissertation is to compare kelp species in a phylogenetic context to understand the factors that have driven and constrained their morphological and ecological diversification. I begin by reconstructing evolutionary relationships between kelps and investigating how changes in global climate and trophic structure may have influenced the diversification patterns of kelps through deep time. Next, I examine the traits and ecological strategies that kelps use to survive environmental stress from waves, and ask whether structural trade-offs or geometric constraints have limited the morphological evolution of kelps. Lastly, I explore whether traits and habitat use are phylogenetically conserved or convergent in northeast Pacific kelp communities. I do this over four chapters that are briefly described below. In Chapter 2, I utilize phylogenomic approaches to determine the evolutionary relationships between kelp species and the timing of their diversification; I elucidate many previously unknown phylogenetic relationships between kelps and demonstrate that their diversification initiated with an increase in speciation rate coinciding with a major climate event and mass extinction (the Eocene-Oligocene boundary, c. 32 mya). This timeline highlights the role that changes in ancient climate have played in the evolution of kelps and supports the   10 hypothesis that kelp diversification occurred before the evolution of benthic foraging mammals like sea otters. In Chapter 3, I investigate the morphological strategies that kelps use to survive environmental disturbance from waves and test for trade-offs that may explain the high level of morphological diversity across the lineage. I use flow tank experiments and field mechanical testing to investigate how external forces from the environment interact with morphology. I show that kelps with high mechanical support also experience greater environmental forces than weaker, more streamlined species, consistent with well-known trade-offs of stress resistance in other organisms.  In Chapter 4, I investigate how kelp species that differ in size allocate biomass to their organs, in the context of widely studied patterns in land plants. I relate the organs of kelps (holdfast, stipe and blade) to analogous organs in land plants (roots, stem, leaves) and determine whether biomasses of these organs scale in similar ways between the two groups despite large differences in evolutionary trajectory. I demonstrate that, although there are fundamental differences in how kelps and embryophytes allocate carbon, constraints of size on morphology may be ubiquitous. In Chapter 5, I combine trait and phylogenetic information to determine whether either can be used to predict species distributions across a gradient of wave action. I explore whether phylogenetic relatedness or trait differences explain the assembly of kelp communities by making use of field data from 87 field sites on Vancouver Island, BC. The results of this analysis demonstrate that, although trait differences correlate with the distribution of species, these traits are labile and do not correspond to phylogenetic relationships. Thus, kelp communities are composed of distantly related species that have converged in morphology. 11 2. A comprehensive kelp phylogeny sheds light on the evolution of an ecosystem 2.1 Synopsis Reconstructing phylogenetic topologies and divergence times is essential for inferring the timing of radiations, the appearance of crucial adaptations, and the historical biogeography of key lineages. In temperate marine ecosystems, kelps (Laminariales, Phaeophyceae) drive coastal productivity and create habitat for many other organisms. However, an incomplete understanding of their phylogeny has limited our ability to infer their evolutionary origins and the spatial and temporal patterns of their diversification.  In this chapter, I present the results of a study aimed at reconstructing the evolutionary diversification of habitat-forming kelps using a global genus-level phylogeny inferred primarily from organellar genome data sets and dated using fossil and biogeographical calibrations. Several new phylogenetic features are resolved, including relationships among the morphologically simple kelp families (sometimes viewed as “ancestral”) and the broader radiation of complex kelps, demonstrating that the initial radiation of the latter resulted from an increase in speciation rate around the Eocene-Oligocene boundary. This burst in speciation rate is consistent with a possible role of recent climatic cooling in triggering the kelp radiation and pre-dates the origin of sea otters and other benthic-foraging carnivores.  Historical biogeographical reconstructions point to a northeast Pacific origin of complex kelps, with subsequent colonization of new habitats likely playing an important role in driving their ecological diversification. We infer that complex morphologies associated with modern kelp forests (e.g. branching, pneumatocysts) evolved several times over the past 15-20 MY,  12 highlighting the importance of morphological convergence in establishing modern upright kelp forests. Our phylogenomic findings provide new insights into the geographical and ecological proliferation of kelps and provide a timeline along which feedbacks between kelps and their food webs could have shaped the structure of temperate ecosystems.  2.2 Introduction Brown algae in the order Laminariales (commonly referred to as ‘kelps’) are the most productive and one of the most ecologically significant clades of macroalgae in the world, playing essential roles as habitat-forming species in temperate regions worldwide (Leigh et al., 1987; Steneck et al., 2002; Teagle et al., 2017). The biomass-rich communities formed by kelps provide a marine analog to terrestrial forests, reaching tens of metres high and producing dynamic and essential habitats along temperate coastlines worldwide (Steneck et al., 2002; Graham, 2004; Bolton, 2010; Kawai et al., 2016). Kelps increase the productivity of nearshore ecosystems both as a source of food (Duggins et al., 1989) and by providing habitat for a variety of other organisms (Teagle et al., 2017; Miller et al., 2018). Kelps have also long provided ecosystem services to humans both directly as a harvestable resource (e.g. for food and valuable extracts: Bartsch et al., 2008), and indirectly by increasing ecosystem productivity (Steneck et al., 2002; Kremen, 2005; Smale et al., 2013). There is even evidence that kelps facilitated transcontinental travel of people to the Americas (Erlandson et al., 2007; Braje et al., 2017). Despite their global distribution and essential role as foundation species, the evolutionary history of the kelps is not entirely understood, and many questions remain about their early evolution and diversification (Bolton, 2010; Kawai et al., 2013, 2016).   13  Kelps are commonly thought to have originated in the North Pacific (e.g. Estes & Steinberg, 1988), but the timing and geographical paths by which their subsequent diversification occurred has been contested (Domning, 1989; Estes & Steinberg, 1989; Bolton, 2010; Kawai et al., 2016). Two of the nine Laminariales families (Akkesiphycaceae and Pseudochordaceae) are endemic to the coast of North Japan, which has the highest kelp species richness of any coastline in the world (Bolton, 2010). Based on this relative richness, it was hypothesized that kelps originated in the western Pacific and spread globally from there (Lüning & tom Dieck, 1990; Bolton, 2010). However, the eastern Pacific has several more genera than the western Pacific, many of which are endemic to the coastline between Alaska and California (Estes & Steinberg, 1988; Bolton, 2010). Rather than basing these arguments on relative taxon richness in different areas, we need to explicitly reconstruct how the spatial distribution of kelps has changed over time using a comprehensive understanding of their phylogenetic history. Kelps are a cold-water lineage found primarily in temperate regions (Graham et al., 2007; Bolton, 2010; Assis et al., 2017), and their global range is currently decreasing as climate change and extreme marine heat waves drive range contractions around the world (Wernberg et al., 2012, 2016; Filbee-Dexter & Wernberg, 2018). Yet, for much of the earth’s history, the ocean (including the North Pacific) was warmer than it is today (Cramer et al., 2009). Until the Late Cenozoic, much of the current range of kelps (e.g. California) was tropical, and unlikely to support kelp communities in their current form (Estes & Steinberg, 1988), except perhaps in deep water refugia (Graham et al., 2007). Thus, it has been proposed that the radiation of complex kelps, currently distributed globally, occurred only recently (Estes & Steinberg, 1988). This is supported by molecular phylogenetic evidence showing relatively little sequence divergence among distantly related kelps (Saunders & Druehl, 1992). Moreover, kelp-associated species, such as stipe-grazing  14 limpets, which have specific shell morphologies allowing them to graze cylindrical kelp stipes, do not appear in the fossil record until the Pliocene, further suggesting that kelp-dominated ecosystems are not ancient (Estes & Steinberg, 1988). Since the Eocene-Oligocene boundary, global climate has cooled substantially, as evidenced by shifts in the distribution of marine taxa (Lindberg & Marincovich, 1988; Ivany et al., 2000; Vermeij et al., 2018) and changes in polar ice cover (Cramer et al., 2009). It is therefore generally thought that kelps evolved and diversified during periods of global cooling, perhaps taking advantage of newly available mid-latitude habitats (Bolton, 2010; Vermeij et al., 2018). This hypothesis is consistent with early molecular clock analyses that suggested that most kelp diversity evolved within the past 30 million years (Saunders & Druehl, 1992; Silberfeld et al., 2010). However, it has yet to be tested explicitly using a dated phylogeny inferred using a broad taxon sampling of Laminariales.  Although not particularly species-rich (~120 species), the kelps are morphologically highly diverse, varying by several orders of magnitude in thallus size (Starko & Martone, 2016a) and exhibiting a wide range of ecological strategies, even among closely related species (Steneck et al., 2002; Lane et al., 2006; Starko & Martone, 2016b). While vicariance and geographical isolation play important roles in speciation of marine taxa (Cánovas et al., 2011; Cowman & Bellwood, 2013; Neiva et al., 2018), adaptive radiation–the diversification of lineages to fill different ecological niches–is also responsible for the production of marine biodiversity (Palumbi, 1994; Schluter, 2000; Ingram, 2011; Aguilar-Medrano et al., 2015) and may explain the wide diversity of kelps. Grazers (especially sea urchins) and their predators (especially sea otters in the North Pacifc) control kelp populations and represent one of the most studied examples of trophic cascades (Estes & Duggins, 1995a), leading to the hypothesis that top predators created an ecological opportunity for the kelps by providing an environment with  15 unusually low herbivory rates (Estes & Steinberg, 1988). Although ancestral sea otters did not appear until the Late Miocene (and modern Enhydra in the Pliocene), top predators have a long history in the North Pacific (Estes & Steinberg, 1988; Vermeij et al., 2018). In particular, Kolponomos which appeared in the North Pacific in the Late Oligocene (Vermeij et al., 2018) and benthic-foraging odobenids that radiated in the Miocene (Repenning, 1976) were both likely similar to sea otters, ecologically. This suggests that there has been potential for predatory mammals to promote kelp populations for much longer than the appearance of sea otters would suggest (Vermeij et al., 2018). Yet, the role that top predators may have played in the success and adaptive radiation of kelps has also been contested (Domning, 1989; Vermeij et al., 2018) and molecular clock analyses of some members of Laminariales suggest that kelps might out-date any of these top predators (Saunders & Druehl, 1992; Silberfeld et al., 2010). An alternative hypothesis is that kelp-dominated ecosystems evolved from the bottom-up: cooling of the North Pacific could have stimulated the adaptive radiation of kelps, fueling coastal productivity and promoting the evolution of complex consumer networks. Thus, there is clear need for a formal analysis of kelp diversification to determine which hypotheses, if any, are consistent with the temporal patterns of kelp evolution. While there has been substantial work on kelp phylogenetics (Boo et al., 1999; Lane et al., 2006; Kawai et al., 2013, 2017; Jackson et al., 2017), we still lack a single analysis that robustly resolves the relationships across a broad enough sampling of Laminariales to explicitly and reliably test hypotheses about the spatial and temporal patterns of kelp origins and radiation. The first extensive molecular kelp phylogeny (Lane et al., 2006) stimulated numerous ecological hypotheses and revolutionized our understanding of the lineage (Bolton, 2010). Although it did not firmly resolve relationships among several key clades, it demonstrated that many important  16 morphological traits, such as branching or production of pneumatocysts (buoyant floats), are distributed throughout kelp phylogeny, suggesting multiple origins of these traits. Jackson et al (2017) added to this work by using a transcriptome dataset to resolve many of the family-level relationships within Laminariales. However, several important kelps were not investigated in that study, directly limiting our ability to make inferences about the evolution of important ecological characters and biogeographical events (Kawai et al., 2016). In particular, previous work has not resolved the relationships between three morphologically distinct kelp families (Akkesiphycaceae, Chordaceae, and Pseudochordaceae) and the other kelp families (Alariaceae, Agaraceae, Arthrothamnaceae, Aureophycaceae, Laminariaceae, and Lessoniaceae), which are often collectively referred to as “ancestral” and “derived” kelps, respectively. The latter terminology is misleading, as it tends to confuse individual traits with the taxa that possess them (Crisp & Cook, 2005); in contrast, a key approach in modern comparative biology is to infer whether individual traits are ancestral or derived. Akkesiphycaceae, Chordaceae and Pseudochordaceae are distinct from other kelps in that they have simple thallus structures and eyespots on their zoospores. Akkesiphycaceae also has anisogamous gametes and monoecious gametophytes (Kawai et al., 2013). All other members of Laminariales produce complex, modular thalli (with clear distinction between stipe and blade), lack eyespots and have unisexual gametophytes and oogamous reproduction (Kawai et al., 2013) (hereafter “complex kelps”). It has been hypothesized that major aspects of the morphology and reproductive biology of Akkesiphycaceae, Chordaceae and Pseudochordaceae are ancestral, but characteristics of these taxa are distinct from those in closely related orders (Silberfeld et al., 2010; Kawai et al., 2013). Thus, the validity of these trait-focused hypotheses concerning ancestral vs. derived traits should  17 be tested using well-supported phylogenetic inferences based on a broadly representative taxon sampling, which remains to be accomplished for kelps (Kawai et al., 2013, 2016).  In this chapter, I present the results of a broadly-sampled phylogenomic analysis that uses gene sets assembled here from organellar genome data (and published transcriptome and organellar genome data sets) and also includes the nuclear rDNA genes (18S and 28S). Most extant kelp genera were included (22 of 29 genera sampled, representing all nine kelp families), the implications of inferred trees for understanding kelp evolution were explored. We dated the kelp phylogeny using fossil and biogeographical calibrations to characterize the spatial and temporal scales over which kelps diversified and specifically ask: What are the relationships among extant kelp genera (including Akkesiphycus, Pseudochorda and Chorda)? Have speciation rates changed through time or across kelp phylogeny, perhaps in response to climate and trophic structure? Where did kelps originate and over what spatial scales have they diversified? And when did kelp forest-forming morphologies first appear?  2.3 Materials & Methods 2.3.1 Taxon and gene sampling Forty species of kelp and eight outgroups were included in the phylogenomic analyses; 34 kelps and six outgroups were included for the plastid data-matrix, and 40 kelps and eight outgroups were included in the mitochondrial and combined data-matrices. New sequence data was acquired for 25 kelp species and three outgroups (Sphacelaria sp., Desmarestia aculeata and Analipus japonicus). Tissues were collected from around the world (Table A1-A3), dried and stored in silica gel for all taxa except for Chorda asiatica, Pseudochorda nagaii and Akkesiphycus lubricus, which were cultured from accessioned collections. Previously published  18 annotated plastomes (n = 4), mitomes (n = 14) and transcriptomes (n = 8) were downloaded from GenBank and incorporated into analyses (Table A4, A5).   2.3.2 Library preparation and sequencing  Next generation sequencing was used to produce phylogenomic datasets for 28 species of brown algae. DNA extractions were performed using one of three methods (Table A2): (a) Qiagen DNeasy Plant Extraction Mini Kit as per the manufacturer’s instructions, (b) brown algal extraction buffer (Saunders & McDevit, 2012) followed by the Qiagen Wizard DNA Clean-Up Kit (Qiagen, Hilden, Germany), or most often (c) a modified CTAB protocol (Doyle, 1987; Rai et al., 2003). DNA libraries were prepared using Bioo NEXTflex Rapid DNA Library Preparation Kit (Bioo Scientific Corporation, Austin, TX, USA), New England Biolab Ultra II DNA Library Preparation Kit (New England Biolabs, Ipswich, MA, USA) or NuGEN Ovation Ultralow Library Preparation Kit (NuGEN Technologies, Inc, San Carlos, CA, USA), following manufacturer instructions (Table A2). Total DNA samples with concentrations between 10 and 1000 ng/uL were sheared (to 400-bp fragments) using a Covaris s220 sonicator (Covaris Inc, Woburn, MA, USA) and size-selected (to 550-650-bp fragments) using magnetic beads (Agencourt AMPure XP Magnetic beads: Beckman Coulter Genomics, Brea CA; NucleoMag NGS Size-selection and Clean-up beads: Macherey-Nagel, Bethlehem, PA, USA). For quality control, libraries were quantified using Qubit fluorometry (Thermofisher Scientific, Waltham, MA, USA) to ensure a minimum DNA mass of 200 ng, verified fragment size profiles using BioAnalyzer (Agilent Technologies, Santa Clara, CA, USA) and measured concentrations of adaptor-ligated fragments using the Kapa Illumina GA library quantification kit (Roche Diagnostics, Basel, Switzerland). Most libraries were sequenced as 125-bp paired-end reads on  19 an Illumina HiSeq 2500 multiplexed with a maximum of 32 samples per lane. However, one sample (Egregia menziesii) was sequenced on an Illumina HiSeq 2000 as 100-bp paired-end reads and multiplexed with 28 other samples. Barcoding sequences (CO1-5P, ITS, or rbcL) were used to check species identities using BLAST (Altschul et al., 1990)(see Table A2).     2.3.3 Assembly, alignment and phylogenetic inference A reference-based assembly approach was used to retrieve individual genes for assembling organellar gene sets. The focus was on protein-coding genes and the large and small ribosomal DNA genes from both organellar genomes. Paired-end reads were combined and mapped to mitochondrial and plastid genome references of Costaria costata (GenBank accessions NC_023506 and NC_028502, respectively) using CLC Genomics workbench v. 6 and 7 (CLC Bio, Aarhus, Denmark). The consensus sequences were extracted using an average minimum read depth of 10x. Organellar data from previously published transcriptomes (Jackson et al., 2017) were also extracted by mapping contigs to the same references. Each gene was aligned individually using MAFFT (Katoh et al., 2002), and manually checked and edited sequences in Geneious R 7.1.9 (http://www.geneious.com; Kearse et al., 2012). Three different concatenated matrices were compiled: one from plastid data, a second from mitochondrial data, and a third from combined plastid, mitochondrial, and nuclear ribosomal genes (28S rDNA, 18S rDNA). Data-matrices were partitioned both by gene and by codon position (first, second and third position for the protein-coding genes, leaving the rDNA genes as individual partitions). PartitionFinder2 (Lanfear et al., 2017)  was used with the random clustering algorithm and AICc to find optimal partitioning schemes and models that were used in subsequent phylogenetic analyses. There were 268 initial partitions for the plastid alignment, 80 for the mitochondrial  20 alignment and 350 for the combined data-matrix, resulting in an optimal partitioning scheme of 114, 55 and 156 partitions, respectively. Maximum likelihood phylogenetic inference was conducted using RAxML 8.2.10 (Stamatakis, 2014), with 20 independent search replicates for the best tree and 1000 bootstrap replicates to assess branch support. Bayesian analyses were conducted in MrBayes (Ronquist et al., 2012) using 5,000,000 Markov Chain Monte Carlo (MCMC) iterations. ESS values were >200 for individual parameters, which were visualized using Tracer (Rambaut & Drummond). Gaps were treated as missing data. Phylogenetic analyses were performed using the CIPRES Science Data Portal (Miller et al., 2011).  A Shimodaira-Hasegawa test (Kishino & Hasegawa, 1989) was performed to determine whether relationships between the Akkesiphycaceae-Pseudochordaceae, Chordaceae and complex kelp clades were significantly more likely than alternative hypotheses for these relationships. Likelihood trees of all three possible topologies were produced by inputting constraint trees into RAxML. Differences in the likelihood of these topologies were then tested using CONSEL (Shimodaira & Hasegawa, 2001). To characterize the phylogenetic placement of five Saccharina species not included in the main phylogenomic dataset, alignments of the 5’-end of cox1 and nad6 were produced using data from this study and previously published sequences (Table A6). Sequences were aligned using MAFFT and edited manually in Geneious. Gene trees were inferred using the GTR+G model in RAxML 8.2.10 (Stamatakis, 2014). These trees allowed for placement of all but six species that have yet to be sequenced extensively, into appropriate genera that are redefined here (see Table A6).   21 2.3.4 Divergence time estimation Molecular clock analyses were performed in BEAST 1.8.4 using a 9,561 bp alignment of plastid (psaA, psbA, rbcL) and mitochondrial (atp6, cox1, cox3, nad6) protein-coding genes, and one ribosomal locus (ITS). To maintain the well resolved tree topology from phylogenomic analyses, a constraint tree was used and the tree estimation operators were unselected when creating the xml file in BEAUTI to restrict MCMC exploration to estimating parameters, rather than tree topology (Drummond et al., 2012). Pseudolessonia laminarioides, which has previously been resolved as the sister group of the giant kelp clade (Postelsia, Nereocystis, Macrocystis and Pelagophycus (Cho et al., 2006)) was included in this analysis by using three previously published sequences (rbcL, psaA, ITS). We used the plastid-inferred topology for Agaraceae; note that preliminary analyses using the mitochondrion-inferred topology for the family yielded similar results for this and subsequent analyses (not shown here). We converted the starting tree into a chronogram (with millions of years as the branch length units) based on the reduced seven-gene matrix using the penalized likelihood method implemented in the R package “ape” with the function “chronos” (Paradis et al., 2004). We used a birth-death model and three calibrations, two were fossil-based calibrations and one was biogeographic. BEAST analyses were run for 10,000,000 generations, sampled every 1000 and three independent runs were combined to achieve convergence and ESS values > 200 for all parameters. The only reliable fossil of a kelp is Julescrania from the Monterey Bay Miocene deposits (Parker & Dawson, 1965; Silberfeld et al., 2010; Bolton, 2010). This was used to set a minimum age of 13 MY for the stem-node of the giant kelps (Nereocystis, Postelsia, Pelagophycus and Macrocystis. The Monterey deposit also included a few species of Cystoseiraceae brown algae (Parker & Dawson, 1965), and so the crown node of the Fucales outgroups (Fucus vesiculosus and  22 Sargassum horneri) was also given a minimum age of 13 MY. In both cases, a log-normal distribution, with a SD of 2MY was used for fossil calibration. We also assigned a uniform prior with a maximum age of 5.3 MY and a minimum age at present to the crown node of Atlantic Laminaria species (L. digitata and L. hyperborea), as they are known to have speciated in the Atlantic, after the opening of the Bering Strait (Rothman et al., 2017).   2.3.5 Diversification analyses To determine whether diversification rates shifted through time in response to changes in ancient climate or biotic conditions such as the appearance of sea otters, we performed two variant analyses. First, we tested for shifts in speciation and extinction through time using BAMM (Bayesian Analysis of Macroevolutionary Mixtures) (Rabosky, 2014). We used this program to fit diversification-through-time models in a Bayesian framework, using priors generated from our time tree (see previous section) in the R package “bammtools” and using 100,000,000 MCMC iterations to obtain ESS values >200. We also assessed lineage-through-time plots to provide an additional line of evidence of temporal patterns of diversification using "phytools” (Revell, 2012). To account for missing taxa in our time tree, we used addTaxa to randomly resolve polytomies within genera that were not completely sampled in our phylogenomic the analysis. We produced a distribution of 100 possible trees that contained these missing species and were therefore nearly complete, missing only six monospecific genera from the Far East Russia whose taxonomic classification is yet to be resolved (i.e., Costularia, Tauya, Undariella, Feditia, Phyllaria, and Streptophyllopsis). Eualaria fistulosa was included as a species on the Alaria stem branch based on morphology and previous molecular analyses (Lane et al., 2007).   23  2.3.6 Historical biogeography Ancestral range reconstruction was performed using a likelihood framework in the R-package BioGeoBears (Matzke, 2013). BioGeoBears is unique in its use of a “J” parameter (jump parameter; founder-event speciation) allowing for simultaneous dispersal and cladogenesis. We ran analyses of dispersal-extinction-cladoenesis models both with (DEC+J) and without (DEC) this parameter (Matzke, 2013). We then used AIC criteria and a likelihood test to determine the model of best fit. Biogeographical areas were chosen to reflect the makeup of regional floras, consistent with the analysis of Bolton (2010). Including a large number of areas can be problematic for ancestral range reconstructions (Landis et al., 2013). Therefore, we simplified Bolton’s scheme by grouping all southern hemisphere areas together as Southern Oceans, and also grouped Atlantic and Arctic regions to reflect dispersal from the Pacific to the Atlantic occurring through the Arctic via the Bering Strait (McDevit & Saunders, 2010; Cánovas et al., 2011). Modern geographical distributions of species were taken from the extensive online taxonomic database, AlgaeBase (Guiry & Guiry; accessed March 2018). Although the modern distribution of some species includes multiple areas, the highest likelihood ranges for all ancestors had two or fewer areas. Thus, to minimize issues associated with a large number of area combinations, we assessed the relative probability of ranges containing a maximum of two areas.  2.3.7 Morphological ancestral state reconstruction To investigate the timing of the evolution of morphological traits, we used the R package “ape” to perform ancestral-state reconstructions of the complex kelps for two traits, growth habit  24 and branching status, with three possible character states for each. For growth habit, species were assigned to categories based on whether they have pneumatocysts, stiff stipes or have no upright feature (i.e., prostrate). We used three states because buoyant floats and stiff stipes both function to hold kelp upright but are analogous. For branching status, we scored species based on whether they had stipe branching (at least one dichotomy in the stipe), blade splitting (i.e., with programmed cell death along the blade but not into the stipe) or no branching (assigned to unbranched). In most cases, stipe branching is a result of blade splitting that extends into the stipe tissue resulting in a branch point, and therefore these are homologous states. Moreover, some species undergo both branching and blade splitting in different parts of their thallus (e.g. Macrocystis pyrifera). We also performed ancestral state reconstruction on all kelps (including Chordales) for seven additional morphological and life history traits (Figs A1-A4) using Ectocarpales (putative sister clade to the kelps; Silberfeld et al., 2010) as an outgroup. We performed ancestral state reconstruction using equal rate (“ER”) and multiple rate (“ARD” and “SYM”) transition models and compared reconstruction fits using AIC. In all cases, equal rate models out-performed more complex transition matrices. To compare phlorotannin content of different kelp species, data were taken from the literature (Steinberg, 1985; Van Alstyne et al., 1999; Dubois & Iken, 2012).  2.4 Results and Discussion 2.4.1 Data characteristics We assembled three data matrices: (a) a 71,153 bp plastid gene alignment that included 88 protein-coding genes shared by all included species (of 127 genes in the plastome of Costaria costata for reference; Zhang et al., 2015) and two plastid rDNA genes (rns and rnl); (b) a 24,490  25 bp mitochondrial gene alignment that included 26 protein-coding (of 63 genes in the mitome of Laminaria digitata for reference; Oudot-Le Secq et al., 2002) and two mitochondrial rDNA genes (rns and rnl); and (c) a 98,840 bp combined alignment for 40 kelps and eight outgroups that also included genes for the nuclear SSU and LSU ribosomal RNAs (i.e., 18S and 28S rDNA genes), the largest data matrix compiled for kelps to date. The plastid alignment comprised 34 kelp species and six outgroup taxa, and the mitochondrial and combined datasets included 40 kelp species and eight outgroups. Sequence data sets were newly recovered for 28 species here (Table A1-A3), with additional data from publicly available genome or transcriptome data sets (Table A4-A5). All alignments include species from 22 currently recognized kelp genera from all nine families, and outgroups from five brown algal orders (Desmarestiales, Ectocarpales, Fucales, Ralfsiales, Sphacelariales). Most of the gene sets are complete or nearly complete, although partial gene sets were recovered for a few taxa (e.g., 44% and 81% recovery here of plastid and mitochondrial data for Aureophycus aleuticus, Table A2; lower recovery or missing organellar gene sets for several previously published taxa, Table A5). Concatenated alignments used in this study are available at http://purl.org/phylo/treebase/phylows/study/TB2:S23066.  2.4.2 Phylogenetic relationships among kelps Using a combined plastid, mitochondrial and nuclear data set comprising up to 120 genes from 40 taxa, we resolved relationships between most extant kelp genera with strong support (Fig 2.1). For the portion of diversity that was previously analyzed by Jackson et al. (2017), we inferred a largely congruent tree. However, our results improve on their analyses by sampling many more taxa within this group and by including three additional families of key interest to the evolution of the kelps. Jackson et al. (2017) resolved relationships among five kelp families  26 using a 16-taxon, 152-gene alignment based on transcriptome data. Our results are consistent with these family-level relationships and differ only by the positions of Ecklonia and Cymathaere, which are strongly supported in both the likelihood and Bayesian analyses presented here. Our analysis swapped the relative positions of these taxa compared to Jackson et al. (2017) who found only moderate support for this node. Family-level circumscriptions inferred here (and in Jackson et al., 2017) also correspond broadly to the clade circumscriptions inferred in earlier five- and eight-gene analyses (Lane et al., 2006; Kawai et al., 2013), although relationships between and within the family-level clades have often been poorly supported or variable in previous studies, and the position of some species in some genera (e.g. Egregia, Ecklonia) drifted depending on sampling scheme (Lane et al., 2006; Kawai et al., 2013, 2017) and phylogenetic inference approaches (Kawai et al., 2013). Thus, the well-supported and consistent (across genomic compartments and inference methods) phylogenetic relationships that we infer here significantly advance our understanding of relationships within the Laminariales. 27  Fig 2.1. Phylogenetic reconstruction of the kelps. (A) Phylogenetic inference based on partitioned maximum likelihood and Bayesian analyses of combined plastid, mitochondrial and ribosomal genes. Support values are shown for both ML bootstrapping and Bayesian posterior probability. Asterisks (*) indicate full support using both methods; other support values are shown beside branches (bootstrap /posterior probability). Thalassiophyllum was not included in analyses of the combined matrix, but its conflicting placements within Agaraceae based on mitochondrial and plastid data are shown in panels (B) and (C), respectively. The topology of Agaraceae from the concatenated alignment excluding Thalassiophylum is shown in panel (D). Scale bars in all panels represent substitutions per site. Our proposed taxonomic scheme is also shown with orders in black boxes and genera identified with brackets (see “Taxonomic implications” section of Results and Discussion). 28 Jackson et al. (2017) also placed Aureophycus (Aureophycaceae) as the sister group of Alariaceae in a separate 24-taxon, eight-gene analysis, with moderate likelihood bootstrap support. This placement of Aureophycus was also moderately supported by a recent study that used an eleven-gene alignment (Kawai et al., 2017).  We confirm this relationship here, but with maximum support (Fig. 2.1). We also resolve most intrafamilial relationships at current taxon sampling with strong support (Fig. 2.1). Minor exceptions include the precise placement of Undaria within Alariaceae (weakly supported by bootstrap analysis, Fig. 2.1), and local relationships among a few members of Saccharina (i.e., S. coriaecea, S. latissima, and a third undescribed species; Fig. 2.1) that are now considered to be a single species (Neiva et al., 2018). Our results confirm that the widespread genus Laminaria and the only genus found exclusively in the southern hemisphere, Lessonia, are sister clades (as reported by Kawai et al., 2017; Jackson et al., 2017), although there are substantial differences in the estimated crown ages of these sister families (Fig. 2.2). This is especially evident when considering that our dataset is likely to cover the entire phylogenetic variation within Lessonia (Martin & Zuccarello, 2012) but not Laminaria (Rothman et al., 2017). Our analyses also uncovered that Arthrothamnus bifidus is nested within the Saccharina clade, a finding with taxonomic implications (see “Taxonomic implications” section below).   29  Fig 2.2. Historical biogeography of the complex kelps. A time-adjusted phylogeny generated in BEAST 1.8.4 using two fossil and one biogeographical time calibration and a subset of genetic data (see text). Pie charts on nodes indicate the most likely biogeographical distribution of that ancestor, as predicted by DEC+J model in BioGeoBears. White colour on the pie chart indicates the combined relative likelihood of ranges that were less than 10%. Node bars near the base of the tree indicate uncertainty in divergence time (95% confidence intervals). Other node bars were removed for clarity but are available in Fig A9.   Our study included three additional kelp families, the morphologically simple Akkesiphycaceae, Chordaceae and Pseudochordaceae, that were excluded (Lane et al., 2006; Jackson et al., 2017) or whose placements were not well resolved (Kawai et al., 2013, 2017) in past phylogenetic studies. Our combined analyses resolve these three families as a clade that is sister to the larger radiation of complex kelps (the “ACP clade,” which we also define as a new order, see “Taxonomic implications” section). We also recovered Akkesiphycaceae and  30 Pseudochordaceae as sister taxa with maximum support (Fig. 2.1). Although support for a broader three-family ACP clade was moderate (82% bootstrap support and 1.0 posterior probability in ML and Bayesian partitioned analyses, respectively), a Shimodaira-Hasegawa test in CONSEL (Shimodaira & Hasegawa, 1999, 2001) finds significantly higher likelihood for this arrangement than alternative hypotheses of ACP paraphyly (LH difference = 8.91, P < 0.001). The existence of the ACP clade has important implications for our understanding of kelp morphological evolution. Species in this clade were previously thought to possess ancestral characteristics (or were viewed as ancestors of other modern lineages). Their collective placement here as a clade comprising three families, sister to the complex kelps, underscores the need to revisit hypotheses of ancestral morphology and historical biogeography in a modern phylogenetic framework. Ancestral state reconstruction of a range of morphological and reproductive characters suggest that many traits that are unique to members of the ACP clade are not ancestral to all of the kelps (Figs A1-A4). For example, sexual dimorphism of gametophytes and unicellular paraphyses are more likely to be derived characteristics of certain species than ancestral traits of the kelps. In contrast, the absence of an eyespot on the zoospore (which is present in members of the ACP clade as well as the sister order Ectocarpales) is a shared, derived trait of the complex kelps only. Our results also suggests that the ancestor of the complex kelps was perennial. This has important implications because phylogenetically disparate members of this clade exhibit similar patterns of seasonal growth including endogenous growth rhythms (Lüning, 1991; Schaffelke & Lüning, 1994; Bartsch et al., 2008).   Separate plastid and mitochondrial analyses were done before combining genes from these different genomic compartments (summarized in Figs. A5-A8). These separate analyses are generally highly congruent with the tree inferred when these data sets are all combined (Fig A9- 31 A10), typically with strong support for individual relationships in each case. Some minor differences that are not strongly supported in individual analyses are highlighted in each supplementary figure. A single strong conflict within ingroup taxa concerns relationships in Agaraceae (Fig 2.1B-D). In the plastid analysis, Thalassiophyllum clathrus is well supported as the sister of Agarum clathratum (Fig 2.1B), but in the mitochondrial analysis it is instead strongly supported as the sister group of the remaining Agaraceae (Fig. 2.1C). This discrepancy was also documented by Kawai et al. (2017), who used a subset of genes from each organelle. Because of these conflicting signals and the possibility of differences in plastid and mitochondrial evolutionary history, we did not include T. clathrus in our combined-matrix analyses, leading to yet a third topology within Agaraceae in likelihood analyses (Fig 2.1D, Fig A9); Bayesian analysis of the combined alignment inferred relationships within the Agaraceae that were consistent with plastid-based analyses (Fig A10). Conflicting evolutionary histories of plastids and mitochondria may arise through a variety of evolutionary processes, such as rapid radiation and incomplete lineage sorting, or hybrid speciation (Maddison et al., 2006; Baack & Rieseberg, 2007; Joly et al., 2009); Agaraceae may therefore provide an opportunity to explore these processes in kelps.   2.4.3 Temporal and spatial patterns   The highly supported and broadly sampled analysis of kelp phylogeny described above allowed us to reliably explore the spatial and temporal diversification of kelps. In order to explore the temporal patterns of kelp evolution, we used three time calibrations to date divergence events: two fossils from the Monterey Bay Miocene deposits (c. 13 MYA) and one trans-arctic biogeographical calibration in the genus Laminaria (see Methods). Time-calibrated  32 molecular clock analyses suggest that while the entire kelp lineage (crown age) is approximately 73 million years old (49 – 120 MY based on 95% highest posterior densities, HPD; Table A7; Fig A11), complex kelps – which comprise most of the diversity of modern kelps - diversified recently, with a crown age around 31.5 (21 – 43) MYA, near the Eocene-Oligocene (EO) boundary (Fig A11). Our clade age estimates are generally consistent with previously published time-calibrated trees where these overlap. For example, Silberfeld et al (2010) examined the entire brown algal radiation based on a subset of kelps, and found divergence times that are consistent with our estimates for the complex kelps and the ACP clade, and also dated the radiation of the complex kelps to c. 30 MYA. Rothman et al. (2017) dated the split between Laminaria and Arthrothamnaceae to approximately 25 MYA, matching a nearly identical estimate here (Fig 2.2; Fig A11). A molecular clock analysis with small subunit rRNA (SSU) (26) is also more or less consistent with our results, with 0.66% sequence divergence between Alaria and Nereocystis, possibly equivalent to c. 16 – 30 MY divergence time (Saunders & Druehl, 1992). The diversification of complex kelps was associated with an increase in speciation rate (Fig 2.3), the timing of which is consistent with large changes in climatic, oceanographic and biotic conditions. During the EO boundary, waters of the North Pacific saw a significant drop in temperature (as demonstrated by ∂O18 in Fig 2.3), which may have provided cool water habitat that were ideal for an initial kelp radiation. Moreover, the EO boundary was associated with a marine mass extinction event that saw a substantial loss of taxonomic diversity in the world’s oceans (Ivany et al., 2000). Although detrimental to some groups, mass extinctions are hypothesized to provide novel ecological opportunities for other lineages (Erwin, 1998; Mahler  33 et al., 2010), and this phenomenon may have contributed to the proliferation of kelps in the North Pacific.   Fig 2.3. Patterns of ocean climate and kelp diversification through deep time. (A) Pacific deep sea ∂18O (a proxy for ocean temperature) over the past 70 million years (data from Cramer et al., 2009). Speciation rate plotted through time (B), and across the kelp phylogeny (C) as estimated using BAMM.   34 Historical biogeographical analysis suggests that much of the ACP and complex kelp clades diversified on opposite sides of the Pacific, although uncertainty remains about the predicted ancestral distributions of several deeper nodes in kelp phylogeny (Fig 2.2). The geographic origin of the common ancestor to the kelps remains unresolved (Fig A12). Ancestral nodes in the ACP clade most likely arose in the northwestern Pacific waters, but much of the early diversification of complex kelps likely occurred on the eastern side of the Pacific Ocean, especially Alaska (Figs 2.2, A12). Although the complex kelps likely diversified in the northeast Pacific, trans-continental, trans-equatorial and polar dispersal have all occurred repeatedly across the tree (Figs 2.2, A12). For example, the most recent common ancestor of the Arthrothamnaceae is predicted to have an eastern Pacific origin, yet Saccharina and allied clades –nested within Arthrothamnaceae- were likely to have originated and largely diversified in Asia (Figs 2.3, A10). Saccharina and Hedophyllum species have subsequently recolonized the eastern Pacific and other parts of the globe several times (Fig 2.3, A10). Kelp lineages have also crossed the equator at least four times independently (Macrocystis, Laminaria, Ecklonia, Lessonia) or separately invaded the North Atlantic through the arctic at least four times (Alaria, Saccharina, Hedophyllum, Laminaria). Resolving the phylogenetic placement of species not included here will improve the resolution at which we understand the historical pathways of kelp dispersal and may reveal even more of these dispersal events. For example, Laminaria may have crossed the equator twice (Rothman et al., 2017) and the dispersal pathways of some widespread genera still remain unclear (e.g.Ecklonia). Future work could also improve on this analysis by incorporating growing knowledge of how geographic features and environmental conditions have changed through deep time. For example, warmer conditions in the past may have prevented kelps from living as far south as they do today and may have made the Bering Sea suitable habitat for a wider range of kelp species. Incorporating this  35 information into the analysis may reduce some of the uncertainty in the dispersal pathways taken by various kelp lineages. These results follow previous work demonstrating the importance of long-distance dispersal events in establishing the contemporary distribution of modern kelp taxa (Bolton, 2010). The DEC+J biographical model, which allows for species to disperse to a new environment concurrently with speciation events, outperformed the more restrictive DEC model (∆ AIC = 11.6296, p < 0.001) but our conclusions were independent of the model of dispersal and speciation used in BioGeoBears (data not shown). Across the phylogeny, our analysis reveals that many kelp taxa are endemic to areas that were likely not inhabited by their recent ancestors. Thus, geographic isolation may have played a fundamental role in the speciation of kelps. This is consistent with previous work on species-rich kelp genera that exist along latitudinal (López-Cristoffanini et al., 2013) or longitudinal clines (Starko et al., 2018a).  A general caveat of these analyses is that we did not include seven monospecific genera due to difficulty in obtaining suitable material (Eualaria, Costularia, Tauya, Undariella, Feditia, Phyllaria, and Streptophyllopsis). With the exception of Eualaria, these taxa are all restricted in distribution to Far East Russia (Bolton, 2010). Recent molecular evidence suggests that Tauya is nested within Arthrothamnaceae (Klochkova et al., 2017); thus, dispersal to Russia from Alaska could explain its distribution. Eualaria is a member of Alariaceae (Lane et al., 2007) and is found exclusively in Alaska and Far East Russia (Lane et al., 2007; Bolton, 2010), and so the exclusion of Eualaria and Tauya is unlikely to have influenced our general conclusions. It is more difficult to predict how inclusion of the remaining five genera would affect tree topology or downstream analyses. None of these taxa have been sequenced (they have only rarely been collected or examined). We speculate that their inclusion might increase the probabilities of Russian origins for some clades, which would support even more back-and-forth dispersal across  36 the Pacific. Inclusion of these species into the phylogeny is a priority, and the essential next step in reconstructing the evolutionary history of Laminariales.  Although geographic isolation and allopatric speciation doubtlessly played an important role in the evolutionary history of kelps, sympatric speciation has also likely occurred, as several kelp subclades appear to have diversified over relatively small spatial ranges (Fig 2.2, Fig A12). For example, the Saccharina clade likely originated in the western Pacific before spreading globally (Figs 2.2, A10). Adaptive radiation is an essential process linking ecology and evolution and may be promoted when resources are underutilized due to mass extinction, during the invasion of new habitats, or because key innovations allow a lineage to exploit a new set of resources (Schluter, 1996, 2000; Givnish, 2015). Kelp species are known to be functionally diverse, and play many different ecological roles, even among sympatric members of the same family (Steneck et al., 2002; Bolton, 2010; Starko & Martone, 2016b). For example, Postelsia palmaeformis invests more than 40% of its dry biomass in attachment to the substratum (Chapter 4), allowing it to live in the high intertidal of extremely wave-exposed sites. In contrast, its sister species, Nereocystis luetkeana, invests only ~10% of its biomass into attachment, and lives in calmer subtidal environments (Chapter 4). These closely related species differ drastically in habitat and morphology, but largely overlap in geographical range. Our phylogenetic reconstruction (Fig 2.1) suggests that this is not an anomaly. Instead, closely related species often differ in morphology (Fig 2.4) and life history (Fig A4). Phenotype-environment correlations are well documented in the literature (Estes & Steinberg, 1988; Druehl & Elliot, 1996; Starko & Martone, 2016a) and studies on survival and growth of genetically distinct intraspecific kelp populations (Kawamata, 2001; Blanchette et al., 2002; Roberson & Coyer, 2004) directly demonstrate that trait differences can influence fitness. Taken together, this  37 collective evidence suggests that ecologically-driven divergent selection may have promoted the kelp diversification and adaptive radiation (Palumbi, 1994; Roberson & Coyer, 2004; Lago-Lestón et al., 2010). 38   Fig 2.4. Likelihood reconstructions of morphological complexity in the kelps. Circles at leafs indicate the morphology of species. Pie charts on internal nodes indicate the relative likelihood (as computed in “ape”) of that ancestor possessing a specific morphology. Panel (A) shows the evolution of traits associated with upright growth (pink: stiff stipes that hold kelps above the substratum; red: buoyant floats that suspend fronds on the water surface. Panel (B) shows the evolution of branching (light blue: blade splitting that forms multiple connected blades; dark blue: stipe branching, in which a blade split progresses into the stipe, causing the branch to form a dichotomy). Both panels include time scales in millions of years. Asterisks (*) indicate significantly higher likelihood of the most likely state than other possible states (LR > 2.0).    39 In many clades, bursts of radiation are followed by a slowing in speciation rate as niche space is filled by newly evolved species (i.e. diversity-dependent speciation (Phillimore & Price, 2008; Rabosky, 2013, 2014)). However, our data suggest that kelps have radiated at a constant rate since the initial burst, with possible evidence of an additional rate increase in the clade containing Saccharina and allies (Fig 2.3). This largely constant speciation rate suggests that diversification did not slow in recent periods as kelps were diversifying ecologically. Therefore, there may be remaining unoccupied niche-space across which kelps will continue to radiate. This hypothesis that kelps have yet to saturate their ecological niche-space in marine environments is consistent with the ubiquity of high intraspecific differentiation (Lane et al., 2007; McDevit & Saunders, 2010), local adaptation (Assis et al., 2017; Neiva et al., 2018) and species complexes (Saunders & McDevit, 2014; Starko et al., 2018a) in modern kelp floras.   2.4.4 Evolution and assembly of kelp forest ecosystems Although our time-calibrated analysis points to a more ancient origin of the kelps than previously proposed by some authors (Estes & Steinberg, 1988; Bolton, 2010), it still suggests that much of the diversity that we associate with modern kelp communities evolved relatively recently. Since the initial burst (c. 31 MYA) near the crown node of the complex kelps, clade-wide speciation rates have not changed through time (Fig 2.3). A constant speciation greater than extinction would lead to exponential growth in the number of lineages through time (Stadler et al., 2014). Indeed, the number of kelp species has approximately doubled in the past 5 million years and increased approximately six-fold in the last 15 million years (Fig 2.5). Upright growth form and branching are two major morphological characteristics associated with complex three-dimensional kelp forest habitats that also appear to have evolved independently and multiple  40 times relatively recently (Fig 2.4). We document that stiff stipes likely evolved from flexible ones at least five separate times, although increased sampling within the most morphologically variable genera (e.g., Laminaria) that contain species both with and without these complex morphological features may reveal additional independent origins of these characters. Branching has also appeared repeatedly across the kelp phylogeny over the past 15-20 MY (Fig 2.4). Evidence that these ecologically-relevant traits evolved in parallel several times across kelps provides additional support for the hypothesis that kelp diversification was driven by adaptive radiation.   41  Fig 2.5. Lineage through time plots for (A) all kelps and (B) complex kelps (excluding ACP clade). Each line represents one of 100 possible, near-complete phylogenies generated using the R package, “addTaxa”.  Modern kelp forest architecture is globally variable and dependent on the species that make up regional pools, but there are general features shared by many kelp forest ecosystems (Teagle et al., 2017). Buoyant kelps produce canopies that reach the surface of the water in many regions around the world. For example, Macrocystis pyrifera, the most widespread canopy  42 species, produces the dominant canopy in the northeast Pacific and in much of the southern Hemisphere (Teagle et al., 2017). Stiff-stiped species are also globally distributed, with species of Ecklonia, Laminaria or Lessonia present in virtually every kelp assemblage around the world (Bolton, 2010). Thus, modern kelp forests almost universally contain species that reach above the benthos. In many subtidal forests, the canopy species, stiff-stiped species and species without upright traits create multiple layers in this highly productive habitat. Traits associated with upright growth appeared much more recently than the initial diversification of kelps (Figs 2.2, A12). Thus, animal species that are adapted to modern kelp forests likely diverged recently from animals that utilized different habitats. This hypothesis is consistent with, for example, the temporal disconnect between the appearance of stipe-dwelling limpets (~ 3 MYA) (Estes & Steinberg, 1988) and the more ancient origin of complex kelps (~ 30 MYA; Figs 2.2, A12).  Our results also refute the hypothesis that sea otters or earlier marine mammals stimulated the early diversification of complex kelps (Estes & Steinberg, 1988). Although the kelps represent a radiation in the North Pacific, where there has historically been a longer marine food chain than in other regions (Estes & Steinberg, 1988; Vermeij et al., 2018), we date the kelp radiation to more than 30 million years ago (Figs 2.3, A10), predating the appearance of both sea otters and earlier benthic feeding mammals (Estes & Steinberg, 1988; Vermeij et al., 2018), as well as sea urchins that evolved in the Miocene (Kober & Bernardi, 2013; Vermeij et al., 2018). 95 % HPD of estimates for the origin of kelps unambiguously exclude overlap with the appearance of sea otters and only narrowly overlap with the presence of Kolpomonos in the North Pacific. Instead, our results are consistent with the hypothesis that kelp diversification and proliferation provided resources that allowed for the diversification of herbivorous invertebrates and benthic-foraging marine carnivores. Kelp presence and abundance increases the productivity  43 of marine animals by providing habitat and increasing food availability for herbivores, detritivores and filter feeders (Duggins et al., 1989). Subtidal invertebrates--such as herbivores and suspension feeders--along coastal shores enriched by kelp diversification may have provided an unexploited resource and encouraged the evolution of a longer food chain. Our timeline also suggests that kelps may have been food for North Pacific sirenians (e.g., Steller’s sea cow) since their initial appearance c. 20 MYA (as proposed in Domning, 1976) and herbivorous desmostylians that lived in the North Pacific from the early Oligocene to the late Miocene and may have fed on coastal algae (Domning, 1989; Vermeij et al., 2018). The intercalary meristem of kelps and their rapid growth rate may have made them ideal sources of food for herbivorous mammals that could have eaten large amounts of kelp tissue without damaging the tissues necessary for growth (Vermeij et al., 2018). The relative importance of top-down and bottom-up forcing on establishing ecosystem structure is a fundamental debate in ecology and evolution (Power, 1992; Hunter & Price, 1992; Estes et al., 2011) and kelp-dominated ecosystems have set the stage for many well-known studies of these phenomena (e.g. Estes & Duggins, 1995a; Estes et al., 2011; Pfister et al., 2017). Although our results demonstrate that otters and earlier carnivores did not stimulate the initial diversification of kelps, top predators clearly exert controls on the structure of coastal ecosystems (Estes & Steinberg, 1988; Estes & Duggins, 1995b; Vermeij, 2012; Estes et al., 2016). The initial argument that kelp diversification may have been stimulated by food chain length was entirely predicated on observations about modern ecology (Estes & Steinberg, 1988). For example, the observation that shallow kelps in the North Pacific are poorly defended against herbivores suggests that kelps in this region may be adapted to longer food-chains (Estes & Steinberg, 1988; Steinberg et al., 1995).  Re-evaluating the phyletic distributions of chemical  44 defenses, however, suggests that defence compound production (phlorotannins) is highly labile –very different percentages found in closely related species (Fig A13). For example, within the genus Laminaria, species differ by more than an order of magnitude in phlorotannin content covering nearly the whole range of this value demonstrated by any kelp (Van Alstyne et al., 1999; data not shown in Fig A13). Labile phlorotannin content and perhaps other traits  (e.g. Fig 2.4, Starko & Martone, 2016b) suggest that food-chain length may have influenced trait evolution, but without influencing the rates of kelp speciation or extinction. Thus the effects of top-down forcing on kelp traits can be understood without invoking a hypothesis that involves the entire duration of kelp evolution. We instead hypothesize that the initial diversification of kelps was promoted by cooling of North Pacific waters and stimulated the lengthening of temperate marine food chains by enhancing coastal productivity (Vermeij et al., 2018). Following this initial burst of diversification, feedbacks between kelps and newly evolving herbivores and carnivores may have shaped the selective landscape of rocky coastlines, influencing trait evolution of kelps and other taxa. For example, despite the finding that the complex kelps radiated ~30 MYA (and diverged from the ACP clade much earlier), the timing of the evolution of many of the traits associated with modern, upright forests is more recent and could have occurred in the presence of a mammalian top predator. In this way, we hypothesize that bottom-up and top-down forcing both played fundamental roles in the evolutionary assembly of modern temperate ecosystems.  2.4.5 Taxonomic implications Our results suggest that taxonomic changes should be made to make the kelp classification system more reflective of lineages’ evolutionary histories and trajectories. The  45 ACP clade is sufficiently distinct (in morphology and phylogenetic distance) from the complex kelps that we describe it here as a new kelp order, Chordales, restricting Laminariales sensu stricto to the complex kelps (Fig. 2.1). This proposed taxonomic change would also exclude the ACP families from being called “kelps”, when restricting usage of the term to the order Laminariales. Chordales Starko, H.Kawai, S.C.Lindstrom & Martone Description & Diagnosis: Heteromorphic life history; gametophyte microscopic and filamentous; sporophyte macroscopic with inconspicuous discoid holdfast. Several plastids per cell with no pyrenoid. Oogamous or anisogamous reproduction. Differentiated from the Laminariales by the lack of differentiation between stipe and holdfast and the presence of an eyespot on the zoospore.  Constituent families: Akkesiphycaceae, Chordaceae, Pseudochordaceae. Type species: Chorda filum (Linnaeus) Stackhouse.  Within Artrothamnaceae, our analyses also provide robust support for the placement of Arthrothamnus bifidus, a branched kelp from the western Pacific, as nested within the Saccharina clade, forming a sister-group relationship to a clade that contains the generitype of the former genus Hedophyllum (S. subsessilis) (Fig. 2.1). Phylogenetic divergence within Saccharina is substantial compared to other genera (with the exception of Laminaria) with divergence time estimates similar in age to the crown node of the giant kelps (Macrocystis, Pelagophycus, Nereocystis and Postelsia; Fig 2.2). Hedophyllum, along with Kjellmaniella (now S. sculpera), were sunk into Saccharina when monophyly of these genera was first reported (Lane et al., 2006). However, their analysis of a small number of genes failed to capture the full phylogenetic diversity of this clade.  Thus, in order to be internally consistent with the taxonomic framework used on other members of the Arthrothamnaceae, we support the reinstatement of Hedophyllum 1901 Setchell (see description in Supplementary Information) and Kjellmaniella 1902 Miyabe, two genera that are morphologically distinct from other Saccharina species, and  46 were recognized for over 100 years. These two resurrected genera were also strongly supported as clades that are distinct from Saccharina sensu stricto in a recent population level analysis with extensive intraspecific sampling (Neiva et al., 2018). We analysed several additional single-gene datasets aimed at including species (n = 5) with less available data (Fig A14-A15, Table A6). These analyses suggest that Kjellmaniella is monospecific (Saccharina gyrata, formerly Kjellmaniella gyrata, is very likely a Saccharina species) and Hedophyllum contains six species (formerly S. bongardiana, S. dentigera, S. druehlii, S. nigripes, S. sessilis, S. subsessilis). A generitype has not been assigned for Kjellmaniella. Thus, we designate K. crassifolia (formerly S. sculpera) as the generitype.  Kjellmaniella Miyabe  Monotypic. Lectotype species: Kjellmaniella crassifolia Miyabe 1902: 134 Homotypic synonym: Saccharina sculpera C.E.Lane, C.Mayes, Druehl, & G.W.Saunders 2006: 962  Hedophyllum Setchell emend. Starko, S.C.Lindstrom & Martone  Description and diagnosis: Young plant with distinct stipe, blade frequently bullate. Mature plants without stipe, haptera developing from basal margin of blade, or plants maintaining a distinct stipe, usually cylindrical near haptera and flattening toward the base of the blade, which often becomes cordate. Mature blades smooth, frequently lacerated into two or more segments. Haptera stout. Lectotype species: Hedophyllum subsessile (Areschoug) Setchell 1901: 122. Designated by Saunders (1901), who attributed the type designation to Setchell.  Hedophyllum bongardianum (Postels & Ruprecht) Yendo Basionym: Laminaria bongardiana Postels & Ruprecht in Illustr. Alg., 1840: 10, pls. XIII, XIV.  Hedophyllum dentigerum (Kjellman) Starko, S.C.Lindstrom & Martone Basionym: Laminaria dentigera Kjellman Om Beringhafyets Algflora 1889: 45.   Hedophyllum druehlii (Saunders & McDevit) Starko, S.C.Lindstrom & Martone Basionym: Saccharina druehlii in Saunders & McDevit, Botany. 2014. 92: 824..   Hedophyllum nigripes (Rosenvige) Starko, S.C.Lindstrom & Martone Basionym: Laminaria groenlandica Rosenvinge Meddelelser om Grønland. 1893: 847   47 Hedophyllum sessile (C. Agardh) Setchell Basionym: Laminaria sessile Agardh. Systema Algarum 1824: 270.   Hedophyllum subsessile (Areschoug) Setchell Basionym: Hafgygia bongardiana f. subsessilis Areschoug in Observ. Phycol. I, 1883: 5.  2.4.6 Conclusions Our phylogenomic study, the largest compiled to date for kelps, allowed us to identify the spatial and temporal patterns underlying their evolutionary diversification. The ACP families, which had previously been described as having ancestral characters, were shown to comprise a sister clade to the complex kelps and are recognized herein as the new and distinct order Chordales. Complex kelps speciated at a substantially increased rate relative to the ACP families, and likely began diversifying during or after the EO boundary. Together, cooling of the ocean combined with the opening of niche space from a mass extinction may have allowed for the rapid radiation and ecological diversification of complex kelps. Historical biogeographical reconstruction suggests that the early diversification of complex kelps occurred in the northeast Pacific, but dispersal to other regions has been a common occurrence, with substantial diversification occurring more recently in Asia. Branched and upright growth forms evolved multiple times independently within complex kelps but did so recently in all cases. Thus, the three-dimensional kelp forest habitats that are widespread along modern temperate coastlines are likely a near-contemporary occurrence. These results provide new insights into niche construction in temperate waters worldwide and lend insight into the timing of major radiations, dispersal events and the origins of ecologically important characters.    48 3. Drag avoidance and tolerance strategies of sympatric kelp species: evidence of a trade-off 3.1 Synopsis Kelps are a clade of morphologically diverse, ecologically important, habitat-forming species. Many kelps live in wave-swept environments and are exposed to chronic flow-induced stress. In order to grow and survive in these harsh environments, kelps can streamline (reducing drag coefficient) to avoid drag, or increase attachment and breakage force to tolerate it. I aimed to quantify the drag tolerance and streamlining strategies of kelps from wave-swept intertidal habitats. I measured drag coefficient and tenacity of populations from eight kelp species over a wide range of sizes to determine whether kelps avoid dislodgement by reducing drag coefficient or by increasing tenacity as they grow, and whether these traits are traded off. I employed phylogenetic comparative methods to rule out potentially confounding effects of species’ relatedness. There was a significant negative relationship between drag avoidance and tolerance strategies, even after incorporating phylogeny. Kelps that were more tenacious were less able to reduce drag, resulting in a continuum from “tolerators” to “streamliners”, with some species demonstrating intermediate, mixed strategies. Drag and tenacity were correlated with geometric properties (ie., second moment of area) of the stipe in large kelps. This suggests that kelps are either strong or streamlined, but not both. This continuum is consistent with avoidance and tolerance trade-offs that have been documented in many different biological systems and may have widespread implications for the evolution of large macroalgae, perhaps driving morphological diversity within this group.  49 3.2 Introduction Trade-offs between avoidance and tolerance strategies of stress resistance have long drawn attention in the fields of evolutionary and functional biology (e.g., Storey & Storey, 1986; Henry & Aarssen, 1997; Fornoni et al., 2003; Iason & Villalba, 2006; Touchette et al., 2007; Puijalon et al., 2011). Avoidance and tolerance trade-offs have been described in a variety of organisms, from amphibians (Storey & Storey, 1986) and mammals (Iason & Villalba, 2006), to plants in salt marshes (e.g., Touchette et al., 2007) and current-swept rivers (e.g. Puijalon et al., 2011). Although these studies addressed different types of stress and resistance, they have repeatedly demonstrated trade-offs, where tolerance or avoidance strategies incur costs to the traits that underlie the opposing strategy. This constraint could be direct, if the use of one trait directly impedes the use of another (Ballhorn et al., 2010) or indirect, if using both avoidance and tolerance strategies together is too costly (e.g., Fornoni et al., 2003; Baucom & Mauricio, 2008). Fluid forces exerted by moving air or water cause chronic stress for many organisms (Koehl, 1982; Vogel, 1994; Denny, 1995; Ennos, 1997; Denny & Gaylord, 2002) and may impart selective pressures on the structure and morphology of plants (Norton, 1991; Ennos, 1997; Demes et al., 2013; Starko et al., 2015). Some plants and macroalgae can reconfigure and streamline, reorienting under waves and “going with the flow” to reduce the drag that they experience (Vogel, 1989; Boller & Carrington, 2006; Martone et al., 2012). Alternatively, plants and macroalgae can increase the thickness and strength (i.e., tenacity) of support tissues (Wainwright et al., 1976; Demes et al., 2013) to resist the drag that they experience.  Drag tolerance traits that involve tissue thickening may directly impede drag reduction traits, if thicker  50 tissues lead to an increase in the rigidity of fronds or basal structures, limiting bending and reconfiguration of macroalgae in flow (Koehl, 1982, 1984; Demes et al., 2011).  Puijalon et al (2011) demonstrated evidence of a streamlining-tenacity trade-off in current-swept embryophytes (see Anten & Sterck, 2012 for a review), in which plants that experienced more drag (i.e., were less streamlined) resisted greater forces before dislodging or breaking. This constituted the first empirical evidence of an avoidance-tolerance trade-off related to mechanical stress. While the generality of this streamlining-tenacity trade-off has been scrutinized and supported (Puijalon et al., 2011; Butler et al., 2012; Anten & Sterck, 2012), it remains an open question whether it can be extended to wave-exposed kelps, which are not land plants and represent a distinct evolutionary trajectory (Keeling, 2004). The wave-swept intertidal zone is of particular interest for studies of mechanical stress resistance, due to the magnitude and unpredictability of wave velocities (Denny et al., 2003b). Among the forces that seaweeds experience in flow, drag is the best documented, and often the most important (Jensen & Denny, 2015). Drag (measured in newtons) is related directly to size through the following formula:         Eq. 3.1 where is the mass density of the fluid (1029 kg m-3 in seawater), U is the fluid velocity (ms-1), A is the planform area (m2), and CD is drag coefficient, a dimensionless parameter that describes the streamlined nature of an organism (Vogel, 1994). Drag forces imposed by crashing waves are not only immense (often greater than hurricane force winds), but also occur frequently, often on the scale of seconds (Denny et al., 2003b). This repetitive loading can cause material fatigue, often leading to dislodgement at otherwise sublethal velocities (Mach et al., 2007, 2011). Thus, selection may not only act to reduce drag under the largest waves, but also under more typical Drag = 12 ρU2ACDρ 51 conditions. Kelps and other brown seaweeds are the largest organisms to inhabit the rocky intertidal zone (Steneck et al., 2002). So how do they grow so large, if size increases drag?  Kelps are flexible, allowing them to reconfigure and reorient in flow (Koehl & Wainwright, 1977; Johnson & Koehl, 1994; Utter & Denny, 1996). Seaweed populations at sites with higher wave exposure may have lower drag coefficients than conspecifics from sheltered areas (e.g., Buck & Buchholz, 2005; Kitzes & Denny, 2005; Fowler-Walker et al., 2006; Koehl et al., 2008a), suggesting a selective advantage to possessing a streamlined form, because streamlining can reduce drag and dislodgement risk.  For example, the bull kelp, Nereocystis luetkeana, streamlines as it grows, adopting a lower drag coefficient (at a given velocity) with increased photosynthetic area to avoid a proportional increase in drag and allowing for decreased investment in support structures (Johnson & Koehl, 1994). Production of increasingly streamlined forms through development may help kelps grow large and experience disproportionally less drag, potentially reducing the need for increased support and attachment. Kelps also have continuously growing holdfasts (Oliveira et al., 1980) and stipes (similar to stalks; see Koehl, 1984) to which they can add additional girth as they grow and develop (DeWreede, 1986; Klinger & DeWreede, 1988; Martone, 2007a), thereby constituting an alternative strategy to resisting wave forces. Mechanical failure occurs when drag equals tenacity (Johnson & Koehl, 1994; Utter & Denny, 1996; Martone et al., 2012). Thus investment in the holdfast or stipe could circumvent the need to streamline, allowing kelps to tolerate rather than reduce drag. Kelp species look similar when they are small, but diversify through development, producing mature plants that are highly variable in morphology (Abbott & Hollenberg, 1976), tenacity (Starko et al., 2015), and streamlining capability (e.g. Koehl, 1984; Johnson & Koehl,  52 1994; Utter & Denny, 1996). Given that streamlining and tenacity are both beneficial for survival in the intertidal zone, are drag-tolerance and streamlining strategies traded off among different kelp species living in wave-swept environments? Could a negative correlation between these capabilities help to explain the wide morphological diversity of kelps? In order to address these questions, I quantified morphogenic changes in streamlining and tolerance over a wide range of sizes for wave-swept populations of eight kelp species, so as to determine whether increases in tenacity are associated with decreases in streamlining abilities, and vice versa.    3.3 Methods  3.3.1 Kelp collections and area quantification  I collected data on intertidal populations of eight kelp species (Fig 3.1) from two wave-swept sites near Bamfield, British Columbia during their peak-growing season. All collections were made between February and September (2013/2014), and included a minimum range of sizes for each species from 70 to 1150 cm2 (see Table 3.1 for additional collection information).  Alaria marginata Postels & Ruprecht, Egregia menziesii (Turner) Areschoug, Macrocystis pyrifera (Linnaeus) Agardh, Saccharina nigripes (J.Agardh) Lontin & G.W.Saunders and Laminaria setchellii Silva were collected from wave-swept, west-facing regions of Eagle (Scott’s) Bay (N 48.83376, W 125.14895). Lessoniopsis littoralis (Tilden) Reinke, Costaria costata (Agardh) Saunders, and Saccharina sessilis (Agardh) Kuntze were collected from a wave-exposed headland at Brady’s Blowhole (N 48.82466, W 125.16148). Offshore swell (commonly 3 – 5 m; LaPerouse Buoy) approaches Barkley Sound from the West, and West-facing sites in Barkley Sound tend to experience similar maximum forces (Boizard, 2007). For a map of the field sites sampled in this study, see Fig B1. I measured planform area as one half of  53 the wetted surface area (Carrington, 1990; Boller & Carrington, 2006; Martone et al., 2012) because total area is a primary predictor of drag (de Bettignies et al., 2013). Specimens were cut up as needed to lay fronds flat and get an accurate measure of planform area, and photos were taken of each specimen from above with a scale of known length. Area was quantified with ImageJ software (version 1.45s; US National Institutes of Health, Berthesda, MD, USA).    Fig 3.1. Photographs of the species used in this study. (A) Alaria marginata, (B) Costaria costata, (C) Egregia menziesii, (D) Laminaria setchellii, (E) Lessoniopsis littoralis, (F) Macrocystis pyrifera, (G) Saccharina nigripes, (H) Saccharina sessilis.  3.3.2 Tenacity I quantified tenacity (in newtons) in situ using tensometers (i.e., spring scales; precision = 0.1 N) that were custom-built using precision springs (Associated Spring Raymond, Ohio, USA). For small kelps, a cable tie was used to attach the stipe to the spring scale, while for larger individuals, the stipe (or entire holdfast in the case of S. sessilis and some M. pyrifera) was wrapped with sanding cushion, and hose clamps were used to attach the tensometer to the kelp  54 (as in Boizard 2007). For some large Macrocystis individuals, the hose clamp was wrapped directly around the rhizomatous holdfast. This method was only performed when a central portion was raised above the substratum and had a gap with no attached haptera. In the case of the strongest kelps (e.g., large L. littoralis, L. setchellii or E. menziesii), a winch, anchored higher in the intertidal zone, was used to apply the forces necessary for dislodgement or breakage. Kelps were pulled perpendicular to the shore, in the direction that they would most likely experience the greatest wave forces. Tenacity was recorded as maximum force to break the stipe or dislodge the holdfast. Any breakage that occurred as a result of cutting from the cable or hose clamp was not recorded.   3.3.4 Drag Drag (in newtons) was measured on kelp specimens at 1 ms-1 in the large recirculating water flume (working section: 1.0 m x 0.8 m x 12 m) at Bamfield Marine Sciences Centre, Bamfield, BC (Fig 3.2). This one velocity (1 ms-1) – the maximum flow rate of the flume – was selected to standardize comparisons among all sizes and species of kelps. Kelps were affixed to a single-axis force transducer (World Precision Instruments, model # FORT5000, 10 Hz sampling rate) with a stainless steel sting that extended 10 cm into the water from above. A small portion of the stipe (1 cm) was affixed to the sting exactly 2 cm from the tip, perpendicular to the direction of flow (Fig. 3.2). For measurements of M. pyrifera and S. sessilis, the holdfast was whittled down as small as possible and affixed directly to the sting at the same position. After each drag measurement, kelp blades were removed, leaving only the holdfast or stipe attached to the sting, as well as any wire or string used to tether the kelp. Drag experienced by this “blank”  55 was subtracted from the initial measurement to account for any drag effects of the attachment method.  Fig 3.2. Schematic of the recirculating flume set-up used to measure drag.  3.3.5 Streamlining and tolerance indices Avoidance and tolerance strategies were quantified as the rate of change in drag coefficient or tenacity versus area, respectively. This allowed for explicit consideration of thallus size when testing for a correlation between avoidance and tolerance strategies. The drag avoidance strategy of each kelp population was quantified as streamlining index (S): the negative scaling coefficient of drag coefficient and area (CD α A-S). This value was used even when the slope was not statistically significant, because S is mathematically equivalent to the difference from unity of the scaling relationship between drag and area (D α A1 – S), which was always significant. Tolerance index (T) was defined as the scaling coefficient of tenacity and area (Tenacity α AT ). S and T were quantified using the slopes of the lines fitted to log-log curves. I 1.0$ms'1$Direc-on$of$Flow$Force$transducer$Stainless$steel$s-ng$Water$line$0.8$m$Kelp$12$m$ 56 tested for a correlation between S and T using a type II linear model conducted in R (v3.0.2; R Foundation for Statistical Computing, Vienna, Austria) with the “lmodel2” package (Legendre 2004). In order to determine whether species had significantly different streamlining and tolerance indices, I conducted ANCOVAs of drag coefficient and tenacity as a function of area, species and a species-by-area interaction.  3.3.6 Predicting avoidance and tolerance at different sizes In order to determine whether more tenacious kelps also experienced more drag, generalized additive models were fit on log-log axis between drag (at 1.0 ms-1) and planform area, as well as tenacity and planform area, using the “mgcv” package in R (Wood, 2011).  Size-specific drag and tenacity were then predicted (with standard error) from these models at two sizes (70 and 1150 cm2); these sizes were chosen because they were minimum and maximum sizes that were collected of all species and thus do not include the maximum range of some larger species (Table 3.1). Type II linear models were used to test for a correlation between tenacity and drag (on log-transformed axis) at these sizes.  3.3.7 Second moment of area To determine whether streamlining and tolerance were predictably related to the tissue thickening of the stipe or base of blade, I quantified second moment of area (I) of basal structures for each species. The tendency of a structure to bend depends on its flexural stiffness (EI), the product of I and material flexibility (E). Second moment of area takes into account both cross-sectional area and shape. Thus it directly describes the structural (i.e., non-material)  57 resistance to bending of a beam. Stipes are approximately elliptical in cross-section; thus I was calculated using the following formula:          Eq. 3.2  where d1 is stipe diameter parallel to the direction of flow, and d2 is stipe diameter measured perpendicular to flow. For S. sessilis, I was quantified for the portion of the blade that was fused with the holdfast. For all other kelps, diameter measurements were taken at the base of the stipe, directly above where the stipe begins to taper out towards the holdfast. Although kelp stipes can vary significantly in material stiffness (E = 5 – 40 MPa; Hale 2001) EI is proportional to the fourth power of stipe thickness (Eq 2).  Thus changes in stipe thickness over the lifetime of a growing kelp would likely have the greatest impact on overall flexural stiffness (EI). For this reason, I was used as an index for the rigidity of basal structures. The relationship between I and planform area was quantified in R using generalized additive models. Estimates of I were then predicted at 70 cm2 and 1150 cm2 (see previous section) and regressed against predicted values of drag and tenacity. For M. pyrifera and S. sessilis, larger kelps may have multiple independent bases, however within the size range used for predictions of I, drag, and tenacity, kelps had a single stipe (for M. pyrifera) or blade (for S. sessilis). Relevant comparisons of I were not possible for larger kelps with multiple stipes or bases.  3.3.8 Phylogenetic comparative methods and statistics In order to account for potentially confounding effects of the kelps’ relatedness, I used the time-calibrated phylogeny generated in Chapter 2 and employed multiple phylogenetic comparative techniques. I used Blomberg’s K (Blomberg et al., 2003) to test for a phylogenetic I = πd13d264 58 signal on tolerance and avoidance traits and Phylogenetic Generalized Least Squares (P-GLS) models, along with linear models, in all multispecies comparisons (S and T; drag and tenacity). These phylogenetic methods were conducted in R using “phytools” (Revell, 2012) and “ape” (Paradis et al., 2004) packages, respectively.   3.4 Results There was a significant interaction of area and species on drag coefficient (CD) (ANCOVA area x species: F = 7.3722, df  = 99, P < 0.0001) and tenacity (ANCOVA area x species: F = 3.8007, df = 129, P = 0.0009). In other words, streamlining and tolerance indices differed between species. For example, Lessoniopsis littoralis was a representative tolerator and Alaria marginata was a representative streamliner (Fig 3.3). As they grow larger, Alaria shows a greater ability to streamline (CD α Area-0.2) than Lessoniopsis (CD α Area0.01) whereas Lessoniopsis becomes much more tenacious (A. marginata: Tenacity α Area0.29; L. littoralis: Tenacity α Area0.56).  Both streamliners and tolerators were identified at each field site, suggesting that the pattern was not site specific. For complete results on modeling streamlining and tolerance, see Tables B1 – B2 and Figs B2 – B3. Streamlining and tolerance indices were negatively correlated (Type II linear model: r = -0.8485, df = 6, P = 0.00774, Fig 3.4) and, although there was no significant relationship between drag and tenacity in the smaller size class (70 cm2; Type II linear model; r = 0.1456, df = 6, P = 0.7308, Fig 3.5a), there was a strong correlation between drag and tenacity at the larger size (1150 cm2; Type II linear model; r = 0.8136, df = 6, P = 0.01401, Fig 3.5b). See Table B3 and Fig B4 for a summary of models used to predict drag at different sizes. There was a significant, positive relationship between second moment of area (I) and drag (Linear model: F = 15.58, df  =  59 6, P < 0.01), as well as I and tenacity (Linear model: F = 20.98, df = 6, P < 0.01) at the larger size class (see Fig 3.6), but not at the small size class (70 cm2; I vs Drag – Linear model: F = 0.9133, df = 6, P > 0.05; I vs Tenacity – Linear model: F = 2.502, dfr = 6, P > 0.05; see Tables B4-B5).   Table 3.1. Additional collection information Species Sample Size1 Min area (cm2) Max area (cm2) Alaria marginata Postels & Ruprecht Costaria costata (C.Agardh) De A.Saunders Egregia menziesii (Turner) Areschoug Laminaria setchellii P.C.Silva Lessoniopsis littoralis (Farlow & Setchell ex Tilden) Reinke Macrocystis pyrifera (Linnaeus) C.Agardh Saccharina nigripes (J.Agardh) Longtin & G.W.Saunders Saccharina sessilis (C.Agardh) Kuntze NS = 21; NT = 24 NS = 12; NT = 15 NS = 15; NT = 14 NS = 12; NT = 13  NS = 14; NT = 19   NS = 13; NT = 23 NS = 13; NT = 23   NS = 15; NT = 14  50 70 60 50 20  25 40  30 4,660 2,220 10,380 1,150 3,580  20,860 2,220  2,010 1NS is the sample size for streamlining quantification; NT is the sample size for tenacity quantification           60  Fig 3.3. An example of closely related species (Alariaceae family) with alternate strategies. Drag coefficient versus area for Alaria marginata (A) and Lessoniopsis littoralis (B). Tenacity versus area for A. marginata (C) and L. littoralis (D). Solid lines indicate regression fits and dotted lines indicate confidence intervals of fitted regressions. Streamlining (S) and Tolerance (T) indices are shown in each panel and were quantified using the regression slope.          61  Fig 3.4 Negative correlation between tolerance (T) and streamlining (S) indices. The dotted line represents the line of best fit from a type II model (T = -1.288*S + 0.623). Note that a significant PGLS model was also fit to this data (see text).          62  Fig 3.5. Drag and tenacity of young thalli (70 cm2) (A) and large thalli (1150 cm2) (B) of all species, as predicted by generalized additive models. Error bars represent standard error, as predicted with a Bayesian algorithm in “mgcv”.  Fig 3.6 Predicted tenacity (A) and drag (B) for large kelp thalli (1150 cm2) , plotted against second moment of area (I). Error bars represent standard error, as predicted with “mgcv”.  63  There was no significant phylogenetic signal for either streamlining (K = 0.4663, df = 6, P = 0.711) or tolerance (K = 0.5174, df = 6, P = 0.608) indices, and there remained a significant negative relationship between streamlining and tolerance indices, even when correcting for phylogeny (PGLS model; t= -7.6714, df = 6, P = 0.0003). Size-dependent correlations between tenacity and drag were also robust to phylogenetic corrections (PGLS model 1150 cm2: t= 4.844, df = 6,  P = 0.0029; PGLS model 70 cm2: t = -0.0002, df = 6,  P = 0.9998).  Fig 3.7. Time calibrated phylogeny of the eight species used in this study constructed from a phylogenomic data in Chapter 2. Streamlining (S) and Tolerance (T) indices are represented by the size of the circle. White circles represent positive values, while black dots represent negative values.   64 3.5 Discussion 3.5.1 Support for a streamlining-tolerance trade-off In this chapter, I present evidence of a mechanical trade-off between streamlining and tolerance strategies of wave-swept kelps, consistent with that demonstrated in aquatic plants (Puijalon et al., 2011; Anten & Sterck, 2012). Species that became more tenacious through development (high values of T) were also those that showed smaller reductions in drag coefficient (low values of S), and species that were better at streamlining (high values of S) did not increase their tenacity as much through development (low values of T) (Fig 3.4). By incorporating phylogeny into our analysis, I demonstrate that relatedness among kelp species cannot explain the observed patterns of streamlining and tolerance (Blomberg et al., 2003) and that low tolerance-high streamlining or high tolerance-low streamlining strategies have appeared repeatedly throughout the evolution of these kelp species (see Fig 3.7). In addition, species that are not strictly “tolerators” or “streamliners” (e.g., S. nigripes and E. menziesii) fall along the same regression line, suggesting that some kelps have intermediate strategies, but none can exist at both ends of the avoidance-tolerance spectrum simultaneously (i.e., achieving maximum values for streamlining and tolerance). This is consistent with the concept of an avoidance-tolerance continuum, as described in previous studies (e.g. Iason & Villalba, 2006; Puijalon et al., 2011).  Perhaps not surprisingly, the trade-off is principally observed in adult thalli. At the large size class (1150 cm2), more tenacious kelps were predicted to be less able to streamline and, therefore, experience more drag. However, this trend did not hold true for kelps that were small (70 cm2).  Small kelps possess similar morphologies, possibly explaining why drag estimates and tenacities were similar among species at this size. Moreover, some factors underlying  65 streamlining and tolerance may be size-dependent. For example larger individuals may have a greater scope for reconfiguration, because they have more area that can potentially reorient, fold up, and compress (see Wernberg, 2005).   3.5.2 Factors driving the trade-off In previous studies, there have generally been two potential explanations for the evolution of avoidance-tolerance trade-offs. One possibility is that avoidance and tolerance strategies are partially or fully redundant and that both incur costs on the function and/or productivity of the organism. Under this condition, use of both strategies would be at an unnecessary cost, and selection could act to reduce redundancy, explaining the occurrence of the trade-off (Fineblum & Rausher, 1995; Fornoni et al., 2003; Baucom & Mauricio, 2008). Indeed, streamlining and tolerance are partially redundant, in that they both aim to resist dislodgement from wave action. So selection to reduce extraneous costs could explain the trade-off between drag avoidance and tolerance. However, drag and tenacity both increase with increased I (Fig 3.6). Consequently, these data suggest that the tolerance-streamlining trade-off results, at least in part, from a direct interaction between the traits that underlie each strategy (as documented by Ballhorn et al., 2010): kelps with high tolerance indices may be physically less able to streamline. Strength and stiffness tend to covary in plant materials (e.g., Niklas, 1993), and the rigidity (ie flexural stiffness) of a structure depends largely on its thickness (e.g., Demes et al., 2011, Eq. 2).  As tenacity increases, tissues near the holdfast thicken, increasing second moment of area at the base of the kelp (Fig 3.6); in this way, species that are drag tolerant may be more rigid and less capable of reconfiguring to reduce drag. Indeed, I varies by more than four orders of magnitude  66 in large kelps of the same size (Fig 3.6). For example, two of the most tolerant species, Laminaria and Lessoniopsis, possess thick, stiff stipes capable of holding the kelps upright, which could hinder reconfiguration and increase forces experienced by the holdfast. These were both the strongest kelps and those that experienced the most drag at 1150 cm2. Saccharina sessilis, another tolerant species lacks a stipe altogether and instead has dissected blades arising directly from a large holdfast. This too could prevent reconfiguration, because the wide blade base might prevent complete bending of the frond into the direction of fluid flow. The three most streamlined species, M. pyrifera, A. marginata, and C. costata all have thinner, less rigid stipes that can bend and allow for increased reconfiguration in flow.   Morphological responses of kelps to environmental factors may also drive the persistence of a streamlining-tolerance continuum. Frond shape and material properties are known to directly affect drag (Boller & Carrington, 2006; Demes et al., 2011; Martone et al., 2012; de Bettignies et al., 2013; Starko et al., 2015), and seaweeds with drag-prone morphologies tend to be more strongly attached to the substratum (e.g., Starko et al., 2015).  If thallus morphologies develop to improve some aspect of ecophysiological performance, such as light capture or nutrient acquisition (Hurd et al., 2014), this may result in thalli that are inherently poor at streamlining (e.g., Koehl et al., 2008a). These kelps might counterbalance this increase in drag by augmenting tenacity to tolerate these increased forces. Kelp morphology is known to vary across gradients of wave exposure (e.g., Duggins et al., 2003; Roberson & Coyer, 2004; Koehl et al., 2008a), depth (e.g., Gerard, 1988; Molloy & Bolton, 2009; Demes et al., 2013), and various other biotic and abiotic factors (e.g., Thomsen et al., 2005a; Charrier et al., 2012), with marked intraspecific variation occurring over both small and large spatial scales (Wernberg et al., 2003; Thomsen et al., 2005b; Fowler-Walker et al., 2006; Lane et al., 2007; Wernberg & Vanderklift, 2010), and  67 between seasons (e.g., Milligan & DeWreede, 2000; Starko et al., 2018b). Thus, avoidance and tolerance strategies may differ between populations of the same species due to combined effects of plasticity, local adaptation and random disturbance from the environment. Future work should document the intraspecific variability in ontogenetic streamlining and tolerance strategies in order to better understand the factors contributing to each strategy, and to the trade-off.  All drag measurements in this study were made at 1.0 ms-1, a water velocity commonly experienced along wave-swept coastlines, but representing only low velocity wave surge. Nevertheless, interspecific comparisons of wave-swept seaweeds (Boller & Carrington, 2007; Martone et al., 2012) show that differences in drag measured at 1 ms-1 tend to be maintained as water velocity increases, suggesting that measurements taken at 1 ms-1 may be sufficient to compare hydrodynamic performance at higher velocities. Conversely, drag experienced by certain species at higher velocities can be difficult to predict (Martone et al., 2012), and differences in drag observed in morphological variants of some species may disappear at higher velocities (de Bettignies et al., 2013). These results cannot rule out convergence upon a narrow range of drag coefficients as water velocity increases.  At a minimum, “streamliners” likely experience less drag than “tolerators” under sublethal conditions. This could, in turn, reduce the risk of long-term fatigue failure for the former cases (Mach et al., 2007, 2011).   The phylogenetic comparative methods used in this study allowed me to conclude that the streamlining-tolerance continuum is not an artefact of data non-independence (i.e., species relatedness; Felsenstein, 1985; Blomberg et al., 2003). Unlike previous work on avoidance and tolerance of aquatic tracheophytes (Puijalon et al., 2011), these methods failed to detect any effect of phylogeny on either streamlining or tolerance strategies. Some closely related species demonstrate very different streamlining and tolerance strategies (see Fig 3.7). For example,  68 Alaria and Lessoniopsis are relatively closely related but exhibit entirely opposing strategies (see Fig 3.3). Furthermore, even the two congeneric Saccharina species fail to fall together on the continuum (Fig 3.4). Thus, streamlining and tolerance strategies may be highly evolvable in the kelps, unlike aquatic tracheophytes, and past disruptive selection may have caused some closely related species to exhibit divergent strategies.   3.5.3 Ecoevolutionary implications One major goal in the study of evolution is to recognize and understand the factors selecting for and maintaining the diversity of traits, particularly those that are ecologically important. For habitat-forming kelps, morphology is particularly important, since it can determine which organisms can be supported by a new habitat, ultimately influencing community composition (e.g., Hughes, 2010). The selective advantages of both tolerance and streamlining strategies are clear. In the absence of a trade-off, it would be expected that species growing on wave-exposed coasts would be both strong and streamlined; this could limit morphological diversity and drive convergence towards a form that is both hydrodynamically optimal (i.e., experiencing less drag than other forms) and maximally strong to tolerate forces generated by the most powerful waves. My results suggest, however, that use of tolerance strategies may preclude the adoption of streamlined fronds and vice versa. Given that streamlining and tolerance are partially redundant strategies, investing in one strategy could free up diversification of the other.  Kelps that invest in tenacity could be free to develop any morphology because they can endure the large forces associated with that structure. Kelps that specialize in streamlining, on the other hand, need not invest much in attachment. This kind of  69 relief from selection could allow for diversity in the opposing trait, perhaps permitting morphologies that are favourable in some other ecological or physiological sense.   3.5.4 Conclusions In this study, I provide evidence for a negative correlation between avoidance and tolerance strategies of intertidal kelps to resist dislodgement under breaking waves. These data suggest that there is a trade-off between these strategies. This trade-off could have had substantial consequences for the evolution of brown algae, since selection for one strategy may lead to a reduced ability or need to use the alternative strategy. This, in turn, may play an important role in maintaining morphological diversity within this group. For example, investing in tenacity could release fronds from such strict morphological selection, allowing them to generate less streamlined structures and contributing to the diversity observed today.  70 4. An empirical test of "universal" biomass scaling relationships in kelps: evidence of convergence with seed plants. 4.1 Synopsis Biomass allocation patterns have received substantial consideration in biology, leading to the recognition of several “universal” interspecific trends.  Despite efforts to understand biomass partitioning among embryophytes, few studies have examined macroalgae, which evolved independently, yet function ecologically in much the same ways as plants. Kelps allocate photosynthate among three organs (blade, stipe and holdfast) that are superficially convergent with organs of land plants, providing a unique opportunity to test the limits of “universal” trends. In this study, I use an allometric approach to quantify interspecific biomass partitioning patterns in kelps and assess whether embryophyte-based predictions of biomass scaling can be applied to marine macrophytes that lack root-to-leaf hydraulic transport. Photosynthetic area and dry mass are found to scale to approximately the ¾ power and kelp biomass allocation patterns are shown to match closely to empirical measures of allometric scaling among woody plants. Larger kelp species are found to have increased relative stipe and holdfast mass than smaller species, highlighting important consequences of size for marine macroalgae. This study lends insights into the evolution of size in the largest marine macrophytes and corroborates previous work suggesting that the morphology of divergent lineages of photoautotrophs may reflect similar selective pressures.  71 4.2 Introduction The consequences of size for living organisms have long been a focus of studies in evolutionary biology (e.g. Huxley, 1932; Kleiber, 1932; Gould, 1966; Peters & Peters, 1986; LaBarbera, 1989; Hanken, 1993; Brown et al., 1993). While many aspects of size evolution are highly variable and lineage specific (Huxley, 1932; Gould, 1966), several common patterns have emerged that may reflect convergent responses of divergent lineages to increased size (West et al., 1997, 1999a,b). Perhaps the most influential and wide-spread consequence of size is the relationship between surface area and volume; while most organisms must inhabit three dimensional space within their environment, they must also interact with it across a two-dimensional surface area. This “curse of dimensionality”, generally causes larger organisms to have decreased surface area to volume ratios (SA:V) relative to smaller organisms, resulting in unavoidable consequences that may drive many widespread patterns in both physiological and morphological evolution (Kleiber, 1932; Niklas, 1994, 2004; West et al., 1997, 1999a,b). In biological systems, surface area to volume scaling rarely matches that of traditional dimensional analysis, which predicts a scaling exponent of 2/3 (Niklas, 1994, 2004; West et al., 1999a). Instead, organisms are intricate and are believed to evolve body plans that maximize exchange area with the environment, while simultaneously maintaining structural integrity and internal transport efficiency (Niklas, 1994; West et al., 1997, 1999a,b).  Presumably because the consequences of SA:V scaling exist for all three-dimensional organisms regardless of phyletic affiliations, several “universal” ¼ - power scaling relationships have surfaced that may represent evolutionary ‘compromises’ between scaling as a plane to maximize surface area and scaling as a Euclidean solid to minimize transport distance (West et al., 1997, 1999b). Although exceptions do exist (Price et al., 2007), these are common scaling relationships, within a broad range of  72 possible geometries (Price & Enquist, 2006; Price et al., 2007), that are supported by large-scale datasets and can be predicted mathematically using models of fractal-like branching (West et al., 1997, 1999a; Price & Enquist, 2006; Enquist et al., 2007; Savage et al., 2010).  4.2.1 Predicting ‘universal’ scaling’ The model of West, Brown, and Enquist (1997, 1999a, b; hereafter WBE model), and extensions of it (e.g. Price & Enquist, 2006; Price et al., 2007; Savage et al., 2010) rationalize some of these universal exponents and predict that effective exchange area (e.g., photosynthetic surface area) should scale to the ¾ power of volume (and therefore total mass: MT) across species (West et al., 1997, 1999a,b). Indeed, the ¾ exponent has been commonly recovered in biological studies related to dimensional scaling (Niklas & Enquist, 2001; Savage et al., 2004; Niklas, 2004) and is believed to be attributed to the diverse, volume-filling branching patterns of plant and animal resource distribution networks (West et al., 1997, 1999a,b). For example, across a large data set that spans several orders of magnitude of embryophytes, net photosynthetic production (NPP) has been shown to scale with the ¾ power of total mass (Niklas & Enquist, 2001; Niklas, 2004; Enquist et al., 2007). This potential consequence of surface area to volume scaling has received substantial attention over the past two decades and may have far-reaching implications for the evolution and ecology of photosynthetic organisms, from single cells to entire forest communities (Enquist & Niklas, 2002b; Enquist, 2002; Savage et al., 2004; Niklas, 2004, 2006). The size-dependence of SA:V and the ¾ exponent has an apparent influence on the standing-organ biomass of plants, whereby increasing size of photosynthetic organs or whole organisms tend to produce “diminishing returns” (Enquist & Niklas, 2002b; Niklas & Enquist,  73 2002; Niklas & Cobb, 2008; Koontz et al., 2009). Because plants typically exhibit a clear division of labour between photosynthetic and non-photosynthetic organs, and leaf mass and surface area are generally proportional within a given species (Roderick & Cochrane, 2002), these diminishing returns manifest themselves as ¾ relationships between leaf biomass and total biomass in both conifers and angiosperms (Enquist & Niklas, 2002b; Niklas & Enquist, 2002; Enquist et al., 2007). Thus, larger plants have relatively more non-photosynthetic biomass and require increased input (carbon allocation) with disproportionately less gain in photosynthetic biomass.  As with the predictable relationship between surface area and mass, ‘universal’ biomass allocation patterns are believed to represent selectively advantageous ways in which standing biomass of leaves, stems and roots should scale across seed plants to balance resource uptake and light interception across two dimensions, while maintaining internal transport and biomechanics in three dimensions. Although substantial consideration has been given to the ways in which seed plants allocate photosynthate and how this can determine standing organ biomass, virtually no work has focused on photosynthetic area-dry mass scaling or organ biomass scaling in aquatic macrophytes. This is especially true when considering marine macroalgae, which are ecologically similar to embryophytes, yet are phylogenetically divergent and face different biophysical limitations. If we are to understand the limits of “universal” scaling relationships and the factors driving them, then incorporation of such taxa may be critical to doing so.   4.2.2 Organ biomass scaling in seed plants Refined extensions of WBE are believed to explain patterns of standing biomass partitioning in embryophytes (Enquist & Niklas, 2002b; Niklas & Enquist, 2002) and,  74 specifically, predict that a universal two-phase scaling relationship can approximate interspecific biomass partitioning across the embryophytes (Niklas & Enquist, 2002; Niklas, 2006). In order to maximize both water delivery and photosynthetic area (such that A ~ MT3/4), across large plants with fractal-like geometries, leaf mass (ML) should scale to the ¾ power of both stem (MS) and root (MR) mass (ML = β1MS3/4 = β2MR3/4, where β is the lineage-specific allometric constant of each relationship), while MR and MS should scale with approximate linearity (MR = β3MS) (Enquist & Niklas, 2002b; Niklas & Enquist, 2002). This scaling relationship is believed to arise due to the accumulation of metabolically inactive wood in the body of large plants, while leaves are periodically lost or turned over (Niklas, 2006). By producing large amounts of non-living tissues, large trees may alleviate potential respiratory ‘costs’ that would otherwise be associated with size increase and allow for isometric relationships between living photosynthetic and non-photosynthetic tissues (Sillett et al., 2010, 2015). This growth strategy reduces the ‘diminishing returns’ associated with increased size and ensures that respiratory metabolism scales with photosynthetic production (such that metabolic rate, B, also scales with MT3/4; Enquist et al., 2007; Mori et al., 2010). In reality, scaling theory tends to underestimate allocation to stems and overestimate allocation to roots of large woody plants, thus broad-scale interspecific scaling relationships only coarsely fit these predictions (e.g., Niklas & Enquist, 2002; Cheng et al., 2007; Poorter et al., 2012, 2015; Zhang et al., 2015). Nonetheless, analyses of large datasets have consistently yielded large coefficients of determination (R2), suggesting a substantial degree of invariance across taxa and convergence of distantly related species towards similar biomass partitioning patterns (Enquist & Niklas, 2002b; Niklas & Enquist, 2002).  In contrast to large plants, herbaceous and young plants (up to ~10-3 kg) that lack substantial, or any, secondary tissue may partition linearly (ie. with an allometric exponent of 1)  75 to each of their three organs (Niklas & Enquist, 2002; Poorter et al., 2012). This is acceptable within the framework of WBE, since many herbaceous or young plants possess stems that are generally photosynthetic (Enquist & Niklas, 2002a), have leaves that tend to increase in thickness through development (Sack et al., 2002) and are incompletely volume-filling (Enquist et al., 2007; Koontz et al., 2009). Additionally, gravity is less important for smaller plants that can elongate without increasing stem diameter to the same extent as larger plants (Enquist et al., 2007). Together this allows for departure from the ¾ scaling of leaf mass with root mass while possibly maintaining A ~ MT~3/4 on average (as in Niklas & Enquist, 2001; Sack et al., 2002; Niklas, 2004). Nonetheless, morphological scaling is generally believed to be less predictable in smaller plants due to the wide diversity of forms that often violate model assumptions upheld by large trees (Price & Enquist, 2006; Enquist et al., 2007; Koontz et al., 2009). Like young and herbaceous plants, macroalgae are photosynthetic along their entire thalli, lack xylem and heartwood, and are not restricted in height by either gravitational or hydraulic constraints. Together this led Niklas (2006) to hypothesize that the only predictions that are applicable to macrophytes are those drawn from the scaling of non-woody, herbaceous plants (ie. ML ~ MS ~ MR ~ MA; isometric/linear scaling of all organs). Although a preliminary investigation provided support for these predictions (Niklas, 2006), results were equivocal due to limited sampling effort.  4.2.3 Testing ‘universal’ relationships with kelps Brown algae (Phaeophyceae, Ochrophyta) are a clade of multicellular protists that have independently evolved a plant-like habit (Keeling 2004). They are perhaps the most three-dimensional macroalgal taxa and form complex underwater forests that are the foundation of  76 temperate nearshore communities (Steneck et al., 2002). Kelps (Laminariales) are the largest and most anatomically complex clade of brown macroalgae (Steneck et al., 2002) with biomass partitioned between three integrated organ systems (holdfast, stipe and blades; Fig 4.1) superficially similar to those of embryophytes (roots, stalk and leaves, respectively) (Niklas, 2006). Kelps therefore provide the ideal “outgroup” with which to test hypotheses about invariance generated from data on land plants. Despite this convergence on a tripartite body plan, kelps are structurally different from land plants in several fundamental ways. Firstly, the importance of gravity in the aquatic environment is substantially less than that of the terrestrial environment, due to the high density of water compared to air. Additionally, kelps (like most, if not all algae) lack non-living tissues and would therefore presumably experience increased respiratory costs, relative to photosynthetic production, if area and mass scale with less than unity. Finally, kelps obtain water and nutrients from the environment and therefore do not rely on root-to-leaf water transport.  In spite of these differences, there are several reasons to draw comparisons between kelps and land plants; kelps possess internal transport systems that are analogous to phloem (Lobban 1978, Graham et al 2008, pg 297; Drobnitch et al 2015) and supply sugars to non-photosynthetic tissues, allowing some species to produce thick tissues and large non-photosynthetic organs (ie. holdfasts). Kelps must also be mechanically supported against water movement (i.e., drag; see Starko et al 2015, Starko & Martone 2016) and therefore often thicken (Martone 2007) and may invest increasing amounts of material into metabolically active support material as they grow larger. However, there is evidence that, despite a lack of non-living tissue, supportive tissues (i.e., holdfasts and stipes) may have reduced metabolic demands compared to actively photosynthesizing tissues (Arnold & Manly 1985). Thus, increasing size may be accompanied by  77 a decrease in relative oxygen consumption, similar to respiratory scaling of large woody plants (but not herbaceous plants; Mori et al 2009).  For these reasons, selection for transport distance minimization and structural support may still compete with selection for photosynthetic area maximization and converge on scaling relationships that are similar to those predicted by scaling theory.  78  Fig 4.1. A comparison of the three convergent organs of sporophytes of land plants and kelps. Fraxinus sp. and Postelsia palmaeformis depicted here. 4.2.4 Aims In this study, I examine interspecific biomass scaling in kelps. I use this ecologically and economically important lineage as a phylogenetic “outgroup” to test whether “universal” scaling principles demonstrated in land plants can in fact be extended to independently-evolved aquatic lineages. I find that kelp photosynthetic area scales to approximately the ¾ power of total dry mass, as predicted by WBE, providing support for some universal biomass partitioning patterns.  This study providess insights into the evolution of size in the largest marine macrophytes and  79 corroborates previous work suggesting that divergent lineages of photoautotrophs may face similar morphological selective pressures.  4.3 Methods 4.3.1 Allometric analyses Interspecific patterns of biomass allocation are modeled as allometric power scaling relationships, such that:          Eq. 4.1 where Y and X are the masses of two organs or parameters with coordinated growth, β is the allometric constant (absolute magnitude or intercept of the relationship) and α is the scaling exponent (see Fig 4.2). Although there are many uses for allometric analyses, in the context of this study I use allometric relationships to describe scaling patterns among species of differing sizes.   Y = βXα 80  Fig 4.2. (A) Pictorial and (B) graphical representations of allometry applied to kelps. Examples shown are for scaling between blade mass (analogous to ML) and holdfast mass (analogous to MR).   4.3.2 Sample Collection  Whole individuals (N = 114) of adult kelps from 23 populations spanning 19 different species were collected for interspecific allometric analyses from eight sites along the Pacific  81 coast of British Columbia (Table 4.1 & C1). Kelps were collected both subtidally (by a combination of SCUBA and free-diving) and intertidally at low tide. Holdfasts were carefully removed from the substratum by means of a knife or paint scraper in order to ensure complete collection of holdfast tissue. Samples that were prone to breakage (e.g., blades of Neoagarum fimbriatum or holdfasts of Egregia menziesii) were kept separate so as to prevent loss of tissue prior to weighing.  4.3.3 Dry Weight Quantification Kelps were air-dried at room temperature for at least 12 hours prior to formal drying. All but the largest of these kelps were placed in a 60o C drying oven for at least 24 hours. Individuals of the same population were dried simultaneously and subsamples from each population were weighed at intervals to determine appropriate drying time (until dry weight had stabilized within 1%). For eight species with three-dimensional morphologies that tended to retain water (e.g., Ecklonia arborea, Lessoniopsis littoralis, Postelsia palmaeformis), all individuals, or subsamples of each individual, were weighed at intervals to confirm that dry weight had stabilized. Due to their large size, Egregia and Macrocystis samples were dried at 37-39oC for 24 hours in an industrial kelp drier (Canadian Kelp Services, Bamfield, BC). Three kelp populations (Nereocystis luetkeana, Costaria costata, Pterygophora californica) collected from Victoria were also too large for oven drying. These populations were sun-dried for approximately 20 hours (over the course of two days), and then dried in a small room that was heated by four space heaters. The room was kept at 32OC for 48 hours and then 37-39OC for 18 hours. Samples from all sites were separated into blade(s), stipe(s), and holdfast; each organ was weighed separately. Regardless of method, all samples were inspected before weighing to ensure complete drying of  82 the tissue samples. In order to determine whether the methods used to dry the largest kelps were effective compared to traditional 60 o C drying oven techniques, subsamples were placed in a 60 o C oven for 24 hours following each of these protocols. Error generated from different drying methods was generally 2-3 % (always less than 4.5 %) which was considered inconsequential given the large range of biomasses and the log-transformed nature of analyses.                   83 Table 4.1. Collection information for biomass partitioning analyses. All sites are located in British Columbia, Canada Habitat Site Location Species Sample Size Intertidal (N = 70) Brady’s Blowhole Barkley Sound Alaria nana H.F.Schrader N = 5  Brady’s Blowhole Barkley Sound Costaria costata Greville N = 5  Brady’s Blowhole Barkley Sound Lessoniopsis littoralis (Farlow & Setchell ex Tilden) Reinke N = 5  Brady’s Blowhole Barkley Sound Saccharina sessilis (C.Agardh) Kuntze N = 5  Cape Beale Barkley Sound Postelsia palmaeformis Ruprecht N = 5  Eagle Bay Barkley Sound Alaria marginata Postels & Ruprecht N = 5  Eagle Bay Barkley Sound Egregia menziesii (Turner) Areschoug N = 5  Eagle Bay Barkley Sound Laminaria setchellii P.C.Silva N = 5  Eagle Bay Barkley Sound Macrocystis pyrifera (L.) C.Agardh N = 5  Eagle Bay Barkley Sound Saccharina nigripes (J.Agardh) Lontin & G.W.Saunders N = 5  Edward King Island Barkley Sound Laminaria ephemera Setchell N = 5  Kitsilano Beach Vancouver Saccharina latissima (L.) C.E.Lane, C.Mayes, Druehl & G.W.Saunders N = 5  West Beach Calvert Island Laminaria yezoensis Miyabe N = 5  West Beach Calvert Island Saccharina nigripes (J.Agardh) Lontin & G.W.Saunders N = 5 Subtidal (N = 44) Bamfield Inlet Barkley Sound Neoagarum fimbriatum (Harvey) H.Kawai & T.Hanyuda N = 5  Bamfield Inlet Barkley Sound Ecklonia arborea (Areschoug) M.D.Rothman, Mattio & J.J.Bolton N = 5  Bamfield Inlet Barkley Sound Saccharina latissimi (L.) C.E.Lane, C.Mayes, Druehl & G.W.Saunders N = 4  Ogden Point Victoria Alaria tenufolia Setchell N = 5  Ogden Point Victoria Costaria costata Greville N = 5  Ogden Point Victoria Nereocystis luetkeana (K.Mertens) Postels &  Ruprecht N = 5  84  Ogden Point Victoria Pleurophycus gardneri Setchell & Saunders ex Tilden N = 5  Ogden Point Victoria Pterygophora californica Ruprecht N = 5  Ogden Point Victoria Saccharina nigripes (J.Agardh) Lontin & G.W.Saunders N = 5  4.3.4 Quantification of A and regression with MT In order to determine how photosynthetic area and total dry mass scale across the kelps, a subsample (N=43 observations, across nine kelp species; Table C2) of kelps were cut into small pieces with scissors, laid flat, and photographed with a scale from above. This “planform” area measurement was then multiplied by two to produce an estimate of total photosynthetic surface area.  No global analysis of total leaf area –total dry weight scaling has ever been conducted on seed plants and few studies report the raw data for both parameters.  However, total leaf area and dry weight were estimated as follows.  First, smaller datasets on leaf area – dry weight scaling were obtained from the literature and from publically available datasets: data on 15 species were taken from studies that directly reported both leaf area and dry mass (Table C3). Data for the two largest species, Eucalyptus regnans and Sequoia sempirvirens, were presented by Sillett et al (2010) and Sillett et (2015) as above-ground mass. Root masses for these species were estimated as a proportion from the Niklas & Enquist (2004) dataset (on Eucalyptus spp.) and Burger et al (2012), respectively. Second, additional data for 23 species were used from the Niklas & Enquist (2004) dataset containing both leaf mass and total mass. Assuming that leaf mass generally scales with leaf area in full-sized individuals (as does WBE; West et al., 1997, 1999b), leaf mass was converted to estimates of leaf area using average leaf mass per area (LMA) or specific leaf area (SLA) measurements taken from Wright et al (2004) and other sources (see Table C4). In  85 studies where data were only presented graphically, data-points were extracted using the software GraphClick (Arizona Software, 2010). When data were presented as one-sided surface area, values were multiplied by two in order to determine estimates of total (two-sided) leaf area. Interspecific allometric analysis of seed plants was conducted on species averages of A and MT.  Intraspecific area-mass scaling data of the cactus, Pachycereus priglei, are also presented as adapted from Price and Enquist (2006).   4.3.5 Statistical analysis Interspecific scaling relationships were determined for three pair-wise comparisons (1) blade (“ML”) and holdfast (“MR”), (2) stipe (“MS”) and holdfast (MR), and (3) blade (ML) and stipe (MS), as well as between frond (stipe + blade; analogous to above-ground mass: “MA”) and holdfast (analogous to below-ground mass: “MB”). All regression analyses were performed using reduced major axis (RMA) slopes of log-log data. This is the standard statistical technique used in allometric analyses because it aims to minimize residual size across both axes, rather than just the y-axis as with ordinary least squares (OLS) regression (see Niklas, 1994). All scaling analyses were performed in R v 3.1.3 (R Development Core Team., 2008) using the “lmodel2” package (Legendre, 2014). Differences from unity (α = 1) were evaluated by calculating confidence intervals of allometric exponents (α) to determine whether parameter estimates differed significantly from one. Data for Saccharina sessilis were excluded from allometric analyses that included MS because adult individuals of this species do not have stipes. In order to determine whether scaling relationships differed between subtidal and intertidal kelps, I compared the slopes and intercepts of each scaling relationship using the “smatr” package (Warton et al., 2012) in R, which was designed to fit and compare allometric  86 relationships. Data from L. ephemera (the smallest species) were excluded from analyses such that data were compared across the same range of values. This, however, had no effect on the interpretation of any of our results.   4.4 Results and Discussion 4.4.1 Scaling of A and MT Despite the phyletic and ecological dissimilarity of land plants and brown algae, the relationship posed by WBE, A ~ MT3/4, appears to hold approximately true even among these distantly related taxa (Fig 4.3). Across more than three orders of magnitude in both area and mass, photosynthetic area of kelps scales to the 0.78 power of total dry weight (RMA regression: A = 2.38MT 0.78±0.09, P < 0.001, df = 47, Table 4.2, Fig 4.3), similar to the ¾ relationship predicted by WBE. The ¾ scaling of A and MT has been previously demonstrated not only in herbaceous and woody plants (as number of leaves; see Niklas, 2004), but also among some unicellular algae (as number of chloroplasts or pigment content; Niklas, 1994; Niklas & Enquist, 2001) and succulent plants (Price & Enquist, 2006; Fig 4.3) that lack substantial branching or an external fractal-like morphology, suggesting that this relationship may be common throughout Chlorophyta. Indeed, interspecific scaling data presented here (Fig 4.3) suggest that area tends to scale to the ~3/4 power of total biomass in seed plants. On average, kelps tend to have greater photosynthetic area per unit dry mass than land plants (as noted by the higher y-intercept; Fig 4.3, Table C5), but they share similar scaling exponents. Thus, our results further corroborate the findings of Price and Enqust (2006) and the predictions of WBE (West et al., 1999b) by demonstrating that a near ¾ scaling relationship has evolved independently in a lineage of aquatic macroalgae. The y-intercept and slope of the kelp dataset are similar to intraspecific data  87 from Arabidopsis thaliana, clearly demonstrating that kelp area-biomass scaling relationships are not greater than all plants, just the average trends (Fig C1). Weedier species, like Arabidopsis, may closely match kelps in terms of the intercept of area-dry mass scaling, likely contributing to their fast growth and competitive ability. The convergence of seed plants and kelps on a near-3/4 relationship may suggest that the form and function of plants with divergent phyletic affiliations are influenced by similar biophysical selective pressures despite the many phylogenetic, ecological and biophysical differences between these lineages.  4.4.2 Biomass partitioning across kelp taxa  In spite of near ¾ scaling of A and MT, kelps differ substantially from herbaceous embryophytes in all of the organ biomass scaling relationships examined in this study and partition considerably more biomass to blades than predicted from embryophytes across all sizes (as noted by log β > 0 in ML vs MR and ML vs MS; Table 4.2, Fig 4.4). On average, the biomass of kelps are 78.6% blade, 11.1% stipe, and 10.8% holdfast, which are notably different proportions from biomass allocation in land plants (8%, 67% and 25% for leaf, stem and root, respectively; Niklas and Enquist 2002). Additionally, none of the organ biomass scaling relationships examined in this study follow clear ¼ power scaling relationships, and demonstrate scaling exponents that are not easily interpreted as the outcome of dimensional scaling rules (e.g., 2/3 , ¾, 1). Organ biomass scaling exponents of kelps do, however, match quite closely to the actual (observed) values of biomass allometry for large woody plants (and not herbaceous plants; Table 4.2, Table 4.3). Specifically, blade mass scales with negative allometry towards both stipe and holdfast biomasses (RMA regression: ML α MR0.85, P < 0.001, df = 114, see Table 4.2, Fig 4A; RMA regression: ML α MR0.71, P < 0.001, df = 109, see Table 4.2, Fig 4B) with  88 confidence intervals that exclude unity (95% C.I.: 0.69 – 0.97), but not ¾ (the prediction from woody species). In general, larger kelps, like large embryophytes, have increased relative MR and MS than species of lesser biomass. In contrast, however, stipe mass and holdfast mass scale with significant positive allometry (RMA regression: MS α MR1.15, P < 0.001, df =109, see Table 4.2, Fig 4C; 95% C.I.: 1.02 – 1.26) that excludes the linear predictions from both herbaceous and woody taxa but not the observed MS – MR scaling (~1.10) of woody taxa. Together these results suggest that biomass scaling exponents in kelps do in fact match up well with observed interspecific allometric exponents from woody plants, but not herbaceous plants. Moreover, both woody plants and kelps deviate from the model of Niklas and Enquist in similar ways. The predictive model of Niklas and Enquist (2002) is founded on several assumptions that have been more or less supported empirically among seed plants (Niklas, 2003), at least for large individuals (Price et al 2009), but are believed to be a result of hydraulic rather than mechanical (i.e., gravitational) constraints (Niklas & Enquist, 2001; Niklas, 2003; Niklas & Spatz, 2004). Intuitively, any assumption that is based on hydraulic requirements cannot be assumed to apply to marine algae, which obtain water and nutrients along their entire thallus by means of simple diffusion (Graham et al., 2008). Because macroalgae are not restricted in structure by the internal transport of water, most kelps are largely blade (~78% by mass; note the strongly positive allometric constants in Table 4.2). However, the scaling exponents of all biomass partitioning patterns were statistically indistinguishable from those observed for woody land plants (see Table 4.2).  How kelps closely match the scaling exponents of large trees, which are highly vascularized and must transport water great distances, remains an open question. However, one possible explanation is that both holdfasts and roots interact with their environment across two-dimensional exchange areas (holdfast attachment area and root  89 exchange area). Therefore, if holdfast attachment area in kelps scales somewhat proportionally with blade area to resist drag due to waves, then this may be analogous to embryophyte root exchange area and leaf area scaling proportionally to provide water for photosynthesis. Thus, although differences in the actual magnitude of organs (i.e., allometric constant) may reflect functional differences between roots and holdfasts, similar allometric exponents may result from shared consequences of dimensional scaling. Larger kelps also tended to have disproportionately massive stipes, which is similar to patterns seen in land plants. For example, two of the largest species, Nereocystis luetkeana (bull kelp) and Macrocystis pyrifera (giant kelp), form underwater canopies with stipes that can grow tens of metres long. While larger holdfasts are likely associated with resisting hydrodynamic forces, larger relative stipe biomass may provide kelps with the competitive advantage required to grow to a larger total body mass. Stipes lift kelps up off the substratum, and thus large stipes could both improve light capture and minimize space requirements along the substratum. This parallels closely selection for uprightness among land plants, which must also compete for light in a forest canopy (Falster & Westoby, 2003). However, stipes also play an important role in resistance to mechanical forces (e.g. Koehl & Wainwright, 1977; Johnson & Koehl, 1994; Utter & Denny, 1996; Denny et al., 1997). Because mechanical forces on the stipe are generally related to tension (see Utter & Denny, 1996), increases in blade size may require concurrent increases in stipe diameter or length in order to resist breakage (see Johnson & Koehl, 1994; Denny et al., 1997; Martone, 2007a; Starko & Martone, 2016b for discussions of stipe diameter allometry).     90  Table 4.2. Parameter estimates (± 95 % confidence intervals) of the reduced major axis regression analyses for each interspecific comparison. Regression Samples Allometric exponent (α)* Allometric constant (log β) R2 A vs MT Subsample (N = 43)  0.78 (0.69 –0.86) 0.71 (0.54 – 0.90) 0.89 ML vs MR All (N = 114) 0.85 (0.74 – 0.96) 0.60 (0.32 – 0.90) 0.67  Subtidal (N = 44) 0.87 (0.64 – 1.15) 0.82 (0.24 – 1.54) 0.54  Intertidal (N = 70)  0.79 (0.68 – 0.92) 0.35 (0.06 – 0.68) 0.73 MS vs MR All (N = 109) Subtidal (N = 44) Intertidal (N = 65)  1.14 (1.02 – 1.26) 1.58 (1.28 – 1.95) 1.01 (0.90 – 1.14)  0.24 (-0.06 – 0.58) 1.38 (0.62 – 2.33) -1.00 (-0.40 - 0.25) 0.76 0.72 0.80 ML vs MS All (N = 109) Subtidal (N = 44) Intertidal (N = 65)  0.71 (0.58 – 0.86) 0.51 (0.30 – 0.79) 0.77 (0.61 – 0.96) 0.32 (-0.04 – 0.74) -0.06 (-0.60 – 0.65) 0.42 (-0.04 – 0.96) 0.48 0.35 0.53 MA = L + S vs MR All (N = 114) 0.85 (0.77 – 0.95) 0.68 (0.46 – 0.93) 0.75  Subtidal (N = 44) 0.89 (0.69 – 1.12) 0.92 (0.43 – 1.51) 0.63  Intertidal (N = 70) 0.81 (0.72 – 0.91) 0.46 (0.22 – 0.73) 0.81 *Bold values significantly exclude null hypotheses for organ scaling (α = 1)  91  Fig 4.3. Interspecific scaling relationships for total thallus area of kelps and total leaf area of land plants reveal similar scaling relationships, but different absolute values (y-intercepts). RMA regression analysis of kelps indicate that A α MT0.78 (N = 43) similar to the interspecific scaling exponent of 0.73 in land plants and the 0.75 intraspecific scaling exponent of the cactus, Pachycereus pringlei (from Price and Enquist 2006).         −4 −2 0 2 4−4−2024Log Total Dry Weight (kg)Log Photosynthetic Area(m2 )Kelps (A = 5.13MT0.78)Seed Plants (A = 1.95MT0.73)Pachycerus  (A = 1.02MT0.75)hycere 92 Table 4.3. Standing biomass scaling exponents (α) for land plants (data from the literature) and kelps Regression Herbaceous Plants1 Woody Conifers2 Woody Angiosperms2 Kelps ML vs MR 0.93 0.86 0.76 0.85 ML vs MS 0.90 0.78 0.73 0.71 MS vs MR 1.01 1.10 1.10 1.14 1From Niklas (2006); 2From Enquist and Niklas (2002a)  Table 4.4. Log-transformed standing biomass scaling constants (log β) for land plants (data from the literature) and kelps Regression Herbaceous Plants1 Woody Conifers2 Woody Angiosperms2 Kelps ML vs MR -0.03 -0.12 -0.89 0.60 ML vs MS -0.19 -0.47 -0.52 0.32 MS vs MR  0.10  0.44  0.42 0.24 1From Niklas (2006); 2From Enquist and Niklas (2002a)        93  Fig 4.4. Interspecific scaling of kelp organ biomasses. (A) Blade mass (ML) plotted as a function of holdfast mass (MR). (B) Stipe mass (MS) plotted as a function of MR. (C) ML as a function of MS. Shaded polygons represent the 95% confidence intervals of scaling relations in herbaceous land plants (from Niklas 2006). Solid black lines are fit to all data (N = 114 for A; N =109 for B, C), but data points represent population averages. Dark blue data points represent populations collected from subtidal sites and light blue points represent intertidal populations. Outliers from general trends are labelled (LL = L. littoralis, PP = P. palmeformes, LE = L. ephemera, EM = E. menziesii, NL = N. luetkeana).  −5 −4 −3 −2 −1−5−4−3−2−10Log Holdfast Mass, MR(kg)Log Blade Mass, ML(kg)−5 −4 −3 −2 −1−5−4−3−2−10Log Holdfast Mass, MR(kg)Log Stipe Mass, MS(kg)−5 −4 −3 −2 −1−5−4−3−2−10Log Stipe Mass, MS(kg)Log Blade Mass, ML(kg)ABCLLPPNLEMLELEPPEM 94 4.4.3 Influence of habitat on biomass allocation Habitat has a strong effect on organ biomass scaling relationships. Subtidal and intertidal kelps differ in allometric constant (ie. intercept) of blade-holdfast, blade-stipe and frond (“above ground”) - holdfast scaling regressions but not the exponents of these relationships (Fig 4.5, Table C6). Across almost three orders of magnitude in holdfast (MR) and stipe (MS) mass, subtidal kelps have significantly more blade mass (ML) and frond mass (ML + MS) than intertidal kelps. Additionally, I found a significant effect of habitat on the slope of stipe-holdfast scaling, with more biomass allocated to stipe biomass in subtidal kelps (Table C6). This is likely a response to selection for increased light interception in deep, subtidal species (eg. Nereocystis luetkeana, Pterygophora californica, Ecklonia arborea).   Wave-induced forces, which are likely also a strong source of natural selection, have imposed mechanical limitations on the structure and function of kelps (Wernberg, 2005; de Bettignies et al., 2013; Starko et al., 2015; Starko & Martone, 2016b), and play an important role in size limitation of marine macroalgae (e.g., Martone & Denny, 2008). Subtidal kelps generally do not experience forces applied by breaking waves, but instead experience slower and more predictable currents (Gaylord et al., 2008). These results demonstrate that kelps growing at subtidal sites develop larger blades relative to their holdfasts (Fig 4.5, Table 4.2, Table C3). However, blade-stipe and blade-holdfast scaling exponents (α) remain constant across environments (Fig 4.5). Additionally, two species that are obligate to highly wave-swept coastlines, Postelsia palmaeformis (the sea palm) and Lessoniopsis littoralis (the pom pom kelp), have the highest relative MR of any species evaluated here (47% and 27% of dry mass, respectively), and closely matched seed plants in terms of absolute organ mass (see Fig 4.4).  Larger blades in subtidal or less wave-exposed kelps could result from differences in allometric  95 growth patterns, whereby certain species could have evolved larger or smaller holdfasts in response to their environment, or it could be a result of increased blade breakage in the intertidal zone due to wave stress. Nevertheless, differences among subtidal and intertidal species highlight the importance of hydrodynamic forces in influencing biomass allocation patterns among kelps.   4.4.5 “Diminishing returns” with increasing biomass In this study, I provide multiple lines of evidence for diminishing returns of net productivity with increased body size, as seen in embryophytes (Niklas and Enquist 2002, Niklas and Cobb 2008, Niklas et al 2009). Larger kelps have larger stipes and holdfasts and thus likely face increased relative metabolic costs compared to smaller kelps. Additionally, total photosynthetic area scaled with approximately the ¾ power of total dry mass, suggesting that increases in size are associated with reductions in the relative proportion of photosynthetic biomass.  Organ biomass scaling relationships in this study suggest that larger kelps have greater proportions of less productive organs similar to trends seen in large land plants.           96    Fig 4.5. Biomass scaling relationships of kelps from different habitats. Regressions in A and C are fit separately between kelps from intertidal and subtidal environments (significant effect of environment covariate), while regressions in B-C are fit to all species (no significant effect of covariate).   Changes in relative blade mass alone cannot explain the ¾ scaling relationship between photosynthetic area and dry mass. Instead, this relationship must also be influenced by changes in blade thickness. For example, at reproductive size the smallest kelp in this study, Laminaria −6 −5 −4 −3 −2 −1−5−4−3−2−101Log MR(kg)Log ML(kg)−6 −5 −4 −3 −2 −1−5−4−3−2−101Log MS(kg)Log ML(kg)−6 −5 −4 −3 −2 −1−6−5−4−3−2−10Log MR(kg)Log MS(kg)−6 −5 −4 −3 −2 −1−5−4−3−2−101Log MR(kg)Log ML+MS(kg)IntertidalSubtidalA BC D 97 ephemera, is generally 0.7 mm or less in thickness. Even in subtidal environments, larger bladed species (Agarum fimbriatum, Saccharina latissima) tend to have central portions that are twice this thick (1.25 – 1.45 mm) (Starko, unpublished). Despite this size-dependence of blade thickness, kelps and other macroalgae may grow more in length than either thickness or width (Scrosati 2006), thus the ¾ scaling exponent of A vs MT likely reflects an intermediate between Euclidean scaling (equal growth in all dimensions: A vs V2/3) and growth in only two dimensions (no change in thickness or relative holdfast contribution: A vs V1) that is accomplished without fractal-like external branching. Unlike embryophytes, kelps do not produce non-living tissues, analogous to the “hoarding of wood” observed in large trees. Because kelps lack dead structural tissue, size-dependent effects of SA:V likely reduce the relationship between living photosynthetic and living non-photosynthetic tissues with increased size. Perhaps larger kelps are able to offset some of these diminishing returns by reducing the metabolic requirement of some structural tissues (Arnold & Manly 1985). Indeed, holdfast and stipe tissues from Macrocystis have substantially reduced respiratory demands (as little as 1/5 of the O2 consumption, by weight) than actively photosynthesizing blade tissues (Arnold & Manly 1985). Moreover, thicker parts of the blade may have reduced respiratory rate than thinner tissues (Arnold & Manly 1985). Thus, despite the apparent lack of non-living tissues within kelp thalli, larger species may accumulate metabolically less demanding tissues, similar to the inner sapwood and heartwood of trees. Future work should address the scaling of respiration and photosynthesis with body size in kelps, in order to determine the extent to which diminishing returns exist, physiologically, and how these scaling parameters compare to seed plants.   98 Despite these diminishing returns, the large size of kelps may still be favourable for many reasons. Smaller or deeper kelps may be more light-limited than larger kelps, due to competition and light attenuation, and may be poorer competitors for space, making them likely to be overgrown by larger species. Thus large size may allow increased photosynthetic production by allowing kelps to reach the surface, improving light interception (see Colombo-Pallotta 2006). Kelps are also highly productive (Mann 1973, Steneck 2001), and the y-intercept of interspecific area-mass scaling across the kelp lineage is higher than the interspecific of plants (Fig 3). This suggests that, even at large sizes, the SA:V may still be relatively high compared to seed plants. Reductions in SA:V with increasing size may therefore not be particularly disadvantageous if the initially high area-mass ratio allows a substantial surplus of carbon production, despite the increase in respiratory metabolism. In addition to this, reproduction requires little extra cost for the kelps. With the exception of Alaria spp, which produce metabolically demanding reproductive blades (Pfister, 1992), most kelps reproduce by forming soral patches on pre-existing blades, rather than on separate structures (Graham et al., 2008). Thus, where seed plants must reserve energy for the production of specialized reproductive structures, kelps likely require little additional energy, beyond blade elongation, in order to reproduce. In this way, larger size may be selectively advantageous, despite increased metabolic demands, because reproductive output likely depends on available blade area for soral production.  Thicker tissues together with increased investment in stipe and holdfast may explain why A and MT scale with negative allometry, but why three-quarters? According to the WBE model, fractal-like structures can reach a maximum A – MT scaling relationship of ¾. However, many of these kelps do not have volume-filling, fractal-like body plans, but instead often possess only one of each organ. Price and Enquist (2006) argue that despite the simple (and not fractal-like)  99 external morphologies of succulent plants, for example, volume-filling internal transport systems are still required in order to deliver photosynthate and water throughout the plant. With simple adjustments to WBE, these authors were able to rationalize why A ~ MT3/4 in succulents despite their lack of a fractal-like external morphology and selection for branch minimization. Similar to succulents, many kelps are unbranched or minimally branched, perhaps due to negative hydrodynamic consequences associated with proliferation of branching (Starko et al., 2015). Despite this, our results provide phylogenetically independent evidence for the universality of the ¾ scaling relationship between A and MT (see Fig 3). Kelps possess phloem-like internal transport systems that are ‘optimized’ in certain species, from the perspective of conduit diameter and packing scaling-relationships (Drobnitch et al., 2015). Thus, although kelps may possess morphologies that reduce mechanical stress, internal transport systems must still work as complex supply networks within the thallus, in order to maintain physiological function or improve growth. For this reason, a ¾ scaling relationship may still be predicted. Alternatively, this relationship may have arisen as a result of mechanical selection: larger species must resist greater drag forces than smaller kelps from the same water velocities and likely require thicker tissues that can resist these increased mechanical forces (see Demes et al., 2011; Starko & Martone, 2016b). Future work on red algae or other brown algal orders that lack vasculature could help to tease apart the effects of vasculature and mechanics.  Regardless of the mechanism, our findings suggest that the relationship that likely has the greatest influence on plant productivity (i.e., A vs MT) may be remarkably similar (and nearly invariant) among decidedly divergent photosynthetic lineages. All plants and macroalgae, with the exception of some crusts and biofilms, photosynthesize and exchange nutrients across a two-dimensional surface area while necessarily occupying three dimensions in their environment.  100 Because of this dimensional constraint associated with surface area to volume scaling, diminishing returns may be an important consequence of size evolution across all plants and macroalgae regardless of evolutionary history. Given this predictable effect of size on surface area to volume scaling, accumulation of metabolically less active structural tissue (as in Arnold & Manley, 1985), somewhat analagous to heartwood and sapwood of seed plants, may partially explain why kelps (but not other macroalgae), are capable of growing so large.  4.4.6 Conclusion In our study, I test current hypotheses about biomass partitioning in an independently evolved lineage of photosynthetic macrophytes. I provide support for the general prediction of the West, Brown and Enquist model that photosynthetic area scales with the ¾ power of dry mass, and establish that interspecific organ biomass scaling patterns match closely to those of woody plants despite substantial differences in the absolute magnitude of these organs. Larger kelps were found to have increased relative holdfast and stipe biomass similar to leaf-stem-root scaling in land plants. The results of this study improve our interpretations of previous models and highlight important consequences of size in a group of organisms that, to date, has been understudied.    101 5. Convergent evolution of niche structure in northeast Pacific kelp communities  5.1 Synopsis The sub-discipline of phylogenetic community ecology is a rapidly developing field that seeks to utilize the growing availability of phylogenetic data to make ecological inferences about species distributions and interactions. Increasingly, approaches have relied on the assumption that traits are more similar among closely related species (phylogenetic signal), which has been largely supported by empirical data. However, adaptive evolution can drive closely related species to differ in ecological traits, and so critics have suggested that a more nuanced approach would be to test for phylogenetic signal to determine the relative importance of phylogenetic constraints and adaptive trait evolution. Doing so may allow us to better understand the possible feedbacks between ecological interactions and evolutionary processes. Here, I combine phylogenetic information generated for kelps in Chapter 2 with a dataset of seven quantitative functional traits and a community dataset from Barkley Sound, British Columbia, to investigate phylogenetic patterns of trait evolution and community assembly for 17 northeast Pacific kelp species. I show that functional traits do not have significant phylogenetic signal in kelps and that closely related species pairs are as likely to have similar traits as distantly related ones. Traits associated with whole-kelp structural reinforcement and material properties are significantly correlated with species distributions along a wave disturbance gradient but no clear predictions about the trait combinations of species could be garnered from the phylogeny alone. Communities are made up of species that are more phylogenetically distinct than in randomly  102 assembled “pseudo-communities” (i.e., phylogenetic overdispersion) and species niches have evolved repeatedly across the phylogenetic tree, suggesting that adaptive morphological evolution has eroded any effect of relatedness on the species traits examined here. These results suggest that environmental filtering by waves plays an essential role in the distribution of kelps on fine scales and that species niches are largely convergent across subclades. This study demonstrates how ecologically-relevant traits may change rapidly in evolutionary time, and shows that assumptions of phylogenetic signal in such traits could mischaracterize observable patterns of variation across species.  5.2 Introduction A key goal among ecologists is to understand patterns of species distributions in the environment and the mechanisms or processes that allow certain combinations of species to coexist. It is a commonly held view that species distributions are determined in part by environmental gradients, such that stress or disturbance from the physical environment exceeds the tolerances of some species and so excludes them from communities (e.g. van der Valk, 1981; Menge & Sutherland, 1987; Webb et al., 2002; Kraft et al., 2014). This process, often referred to as environmental filtering (Kraft et al., 2014), can occur across a range of spatial scales, reflecting latitudinal patterns in temperature and precipitation (Reich & Oleksyn, 2004; Swenson & Enquist, 2007; Kraft et al., 2011; Cavalheri et al., 2015) to finer scale patterns of disturbance or other environmental variation (Enquist et al., 2015; Ulrich et al., 2017). Biotic factors, such as competition or predation, can further limit or promote species distributions by excluding species that are poor competitors (Hardin, 1960; Levin, 1970) or that are unable to withstand physical disturbance from predators or grazers (Harley, 2003; Ishii & Crawley, 2011). Identifying the  103 differences among species that allow them to live in particular environments has become a key approach used by ecologists to understand species distributions and community assembly (Kraft et al., 2014). There is a long history of using quantitative measurements of traits to examine differences among species (e.g. Macarthur & Levins, 1967; Poff, 1997; Westoby & Wright, 2006; Kraft et al., 2008). Quantitative trait measurements describing species morphology, physiology or ecology have been used successfully to predict the outcome of environmental filtering or species interactions in many taxa (Westoby & Wright, 2006; Litchman & Klausmeier, 2008; Messier et al., 2010) including land plants (Swenson & Enquist, 2007; Enquist et al., 2015), fish (Frimpong & Angermeier, 2009, 2010), invertebrates (Best & Stachowicz, 2013), zooplankton (Thompson et al., 2015) and phytoplankton (Litchman & Klausmeier, 2008). It is therefore widely thought that studying traits may allow for a mechanistic understanding of ecological niches and species distributions through time and space (Westoby et al., 2002; Westoby & Wright, 2006; Messier et al., 2010; Weber et al., 2017).  The processes of competition and environmental filtering provide alternative expectations for how functional traits should differ within and across communities. On one hand, theory suggests that differences among species are necessary for coexistence among competitors (Macarthur & Levins, 1967), so if competition dominates as a driver in the assembly of communities, species are expected to differ in ways that allow for coexistence (Macarthur & Levins, 1967). This would lead to communities composed of species that differ in traits to avoid competition (Losos, 1995; Cavender-Bares et al., 2004a). On the other hand, if environmental filtering dominates, then we might expect species with similar traits to cluster at locations that are habitable due to specific combinations of traits (Kraft et al., 2014). While the value of traits  104 as predictors of species ecological niches has been widely demonstrated (e.g. Macarthur & Levins, 1967; Poff, 1997; Westoby & Wright, 2006; Kraft et al., 2008), we have a limited understanding of how ecological processes act on fine spatial scales to influence patterns of trait evolution across whole clades. A growing number of studies have assumed that ecological niche and traits are both similar among closely related taxa (e.g. Webb, 2000; Kraft et al., 2007; Burns & Strauss, 2011) in order to infer whether environmental filtering or competitive exclusion are important drivers of community assembly. That is, if coexisting species are more closely related than expected by chance, then traits are often assumed to be clustered, providing tentative evidence that environmental filtering drives community assembly (Webb et al., 2002; Kraft et al., 2007; Cavender-Bares et al., 2009); if community members are more phylogenetically distant than expected by chance, then traits are assumed to be different among species, providing evidence that other species are competitively excluded (Webb et al., 2002; Kraft et al., 2007; Cooper et al., 2008; Cavender-Bares et al., 2009). Hereafter, I will refer to this correlation between traits or niche and phylogeny as phylogenetic signal, following the definition of Losos (2008). Phylogenetic signal is distinct from phylogenetic niche conservatism, which refers to when traits are more similar among close relatives than expected through stochastic evolution (Losos 2008). The assumption of phylogenetic signal is based on the concept that closely related species are non-independent and share a common ancestor with a specific set of traits. Random walk simulations predict that closely related species will share similar trait values (Pagel, 1999; Blomberg et al., 2003) unless other processes cause them to differ in traits (Losos, 2008; Ackerly, 2009). Although there is reasonable support for the assumption of phylogenetic signal among certain taxa, it is clear that not all ecologically important traits have phylogenetic signals  105 (Losos, 1995, 2008; Böhning-Gaese & Oberrath, 1999; Cavender-Bares et al., 2004a; Best et al., 2013; Best & Stachowicz, 2013; Blonder et al., 2016), suggesting that adaptive evolutionary processes may eliminate the expected correlation between species relatedness and trait similarities. To avoid misinterpretation of the phylogenetic structuring of communities, it is critically important to determine which traits allow species to inhabit different environments, and how prevalent phylogenetic signal is among divergent taxa. An alternative hypothesis to that of correlation between traits and phylogenetic distance is that closely related species may differ due to divergent selection (Cavender-Bares et al., 2004a; Losos, 2008). Competition between closely related species might drive divergent selection, character displacement and niche partitioning (Losos, 2008), resulting in closely related species that are more ecologically different than expected by chance (Cavender-Bares et al., 2004a). For example, throughout a region, closely related species may diverge to adapt to different parts of the environment (Schluter, 2000). This may occur concurrently in multiple subclades, leading to communities that are composed of ecologically similar but distantly related species (Cavender-Bares et al., 2004a; Losos, 2008). One example of this pattern is the oak trees (Quercus), which invaded North America coincident with the extirpation of tropical lineages around the Eocene-Oligocene boundary (Cavender-Bares et al., 2004a). Across the entire region, species are phylogenetically overdispersed in the environment, with closely related species occupying different ecological niches and specializing in different types of habitats (Cavender-Bares et al., 2004a). Partitioning of habitat may help to maintain coexistence of closely related species across broad geographic scales by limiting direct competition among close relatives (Cavender-Bares et al., 2004b,a). However, the prevalence of this type of pattern in nature remains unclear (Losos, 2008).   106 Kelps are potentially an excellent model system for testing ecological concepts involving functional morphology. They are ecologically and morphologically diverse, inhabiting a wide range of environments from calm bays to exposed outer coasts, and from the temperate rocky intertidal to deep, cold waters in tropical regions (Steneck et al., 2002; Bartsch et al., 2008; Bolton, 2010). Despite this extensive ecological diversity, kelps diversified quite recently (Saunders & Druehl, 1992; Bolton, 2010; Chapter 2), restructuring temperate ecosystems as they spread worldwide (Steneck et al., 2002; Vermeij, 2012). Around the same time that oaks were invading temperate North America (~30 MYA; Cavender-Bares et al., 2004b), kelps began diversifying in the northeast Pacific Ocean (Chapter 2) as waters cooled and tropical species went extinct (Ivany et al., 2000; Goldner et al., 2014). From ancestors that were unbranched and simple in morphology, complex anatomical traits (e.g. branching, float formation) evolved many times across the kelp phylogeny (Chapter 2). The wide morphological variation of kelps and the short time over which this arose suggests that strong divergent selection and adaptive radiation (defined here as the rise of ecological diversity within a lineage; Givnish 2015) shaped the evolution of this clade, perhaps linked to ecological opportunities associated with cooling temperatures in the north Pacific (Chapter 2). Thus, kelps provide an opportunity to evaluate whether adaptive evolutionary processes can erode the expected phylogenetic signal of traits and niche. Such loss of phylogenetic signal would suggest that knowledge of the processes underlying trait evolution is required to interpret the causes of phylogenetic structure in ecological communities.  In this chapter, I use kelps as a model system to test the persistent hypothesis that closely related species are more ecologically similar than distantly related species. I begin by investigating whether there is a detectable phylogenetic signal on kelp functional traits. Then, I  107 determine whether phylogenetic distance or trait differences can be used as predictors of kelp distributions, by using a community dataset from Barkley Sound on Vancouver Island, British Columbia. I explore species co-occurrences and test for trait-environment correlation in an evolutionary context to determine whether closely related species have similar traits and ecological niches, or alternatively, whether kelps underwent niche partitioning as they radiated in the northeast Pacific.   5.3 Methods 5.3.1 Trait data I compared seven quantitative traits for all 17 kelp species of interest, many of which are analogous to commonly measured traits in land plants; these included two traits describing whole individual biomass allocation (stipe mass fraction or SMF, holdfast mass fraction or HMF) and five traits describing mechanical and structural properties of blade tissues. SMF and HMF describe the proportion of total biomass that exists as stipe or holdfast material, respectively. HMF is analogous to root-shoot ratios in land plants. Each of these traits were measured as described in Chapter 4. In short, organs (holdfast, blades, stipes) of individual kelps were carefully separated and dried in a 60oC drying oven or 37-39oC drying room and were weighed using electronic scales. Blade mass per area (BMA; analogous to leaf mass per area) was defined as the amount of dry biomass per unit area of blade tissue, and dry matter content (DMC) was defined as the ratio of dry weight to wet weight. Both BMA and DMC were measured by taking hole punches of standardized area out of the base of the blade and measuring the wet mass and dry mass of each hole punch. Mechanical properties of blade material: breaking stress (σ), stiffness (E) and extensibility (ε), were measured using an Instron (Instron, Massachusetts,  108 USA), a portable tensometer, or were taken from the literature (Tables 5.1-5.2). With the exception of the few material properties measurements taken from the literature, trait data represent average measurements taken from adult individuals of populations in southern British Columbia (Barkley Sound, Port Renfrew, Vancouver or Victoria; see Tables 5.1-5.2).   Table 5.1. Locations of field sites from which trait data were collected Site Name Location Latitude Longitude Bamfield Inlet Barkley Sound, BC 48.8345 -125.13682 Brady's Blowhole Barkley Sound, BC 48.82329 -125.16151 Edward King Island Barkley Sound, BC 48.82235 -125.21731 Scott's Bay Barkley Sound, BC 48.83413 -125.14775 Prasiola Point Barkley Sound, BC 48.81751 -125.16926 Cape Beale Barkley Sound, BC 48.78537 -125.2165 Ogden Point Victoria, BC 48.41399 -123.38572 Botanical Beach Port Renfrew, BC 48.52753 -124.44877 Whytecliff Park Vancouver, BC 49.37226 -123.29212              109 Table 5.2. Sources of trait data used in this study. σ = breaking stress, E = tensile modulus (stiffness), ε = extensibility, SMF = stipe mass fraction, HMF = holdfast mass fraction, DMC = dry matter content of blades, BMA = blade mass per area     Species Materials (σ, E & ε) Biomass (SMF & HMF) Blade Properties (DMC & LMA) Alaria marginata This study, Botanical Beach (n = 8) Chapter 4  (n = 5) This study, Blowhole  (n = 10) Lessoniopsis littoralis This study, Brady's Blowhole (n = 5) Chapter 4  (n = 5) This study, Blowhole  (n = 9) Pleurophycus gardneri This study, Ogden Point (n = 5) Chapter 4  (n = 5) This study, Ogden Point (n = 6) Pterygophora californica This study, Botanical Beach (n = 8) Chapter 4  (n = 5) This study, Ogden Point (n = 2) Costaria costata This study, Whytecliff Park  (n = 8) Chapter 4  (n = 5) This study, Scott's Bay  (n = 3) Neoagarum fimbriatum This study, Whytecliff Park  (n = 8) Chapter 4  (n = 5) This study, Bamfield Inlet (n = 4) Egregia menziesii Demes et al 2013  (n = 39) Chapter 4  (n = 5) This study, Scott's Bay  (n = 5) Ecklonia arborea Hale 2001 This study, Scott's Bay  (n = 5) This study, Scott's Bay  (n = 5) Cymathaere triplicata This study, Ogden Point (n = 7) Chapter 4  (n = 5) This study, Ogden Point (n = 5) Nereocystis luetkeana This study, Botanical Beach (n = 8) Chapter 4  (n = 5) This study, Scott's Bay  (n = 3) Macrocystis pyrifera Hale 2001 Chapter 4  (n = 5) This study, Scott's Bay  (n = 13) Postelsia palmaeformis This study, Botanical Beach  (n = 8) Chapter 4  (n = 5) This study, Cape Beale  (n = 3) Saccharina sessilis This study, Botanical Beach (n = 8) Chapter 4  (n = 5) This study, Prasiola Point (n = 7) Saccharina nigripes This study, Scott's Bay (n = 6) Chapter 4  (n = 5) This study, Scott's Bay  (n = 12) Saccharina latissima This study, Bamfield Inlet (n = 4) Chapter 4  (n = 4) This study, Bamfield Inlet (n = 6) Laminaria setchellii Starko et al 2018 (Blowhole) (n = 6) Chapter 4  (n = 5) This study, Blowhole  (n = 3) Laminaria ephemera This study ((n = 4) Chapter 4  (n = 5) This study, Edward King (n = 4)  110 5.3.2 Phylogenetic relationships The phylogeny of kelps, a group with more than 100 species, is complex and has been studied previously in considerable detail (e.g. Lane et al., 2006; Jackson et al., 2017; Chapter 2). In this study, I used the time-calibrated phylogeny inferred in Chapter 2 to represent phylogenetic divergence in millions of years for the 17 co-occurring Northeast Pacific kelp species of interest. This dated phylogenomic analysis is the most well supported and comprehensive to date and included all 17 species except Laminaria setchellii, which was incorporated into the analysis by substituting it for L. digitata, the latter not found in the northeast Pacific but included in the phylogenetic analysis in Chapter 2. This substitution relies on the assumption that L. setchellii had an equivalent divergence time from Laminaria ephemera as L. digitata, which is well supported by previous work on intrageneric relationships between Laminaria species. These show less than 1 million years difference in divergence time between L. ephemera and L. setchellii vs. L. ephemera and L. digitata (Rothman et al., 2017).   5.3.1 Community data To determine how trait or phylogenetic differences influence community assembly, I used a community dataset of intertidal kelp distributions in Barkley Sound, British Columbia that was published in a government report (Druehl & Elliot, 1996). Data from sites sampled in 1995 (n = 87 sites), the most extensive year of this survey, were combined. The dataset included all of the species examined in the trait analysis except two (Laminaria ephemera and Cymathaere triplicata). Although a coarse categorical abundance measurement is given in this report, only presence and absence data were used here. At a subset of sites (n = 55) that could be located by photographs in the 1996 report, the upper limit of barnacles was measured in the summer of  111 2018, and these values were used as continuous proxy for wave exposure. The upper limit of barnacles is known to increase in elevation at more wave exposed sites (Harley & Helmuth, 2003; Neufeld et al., 2017). The upper limit of barnacles was measured by using a stadia rod and sight level, along with tide predictions from Bamfield Inlet, Effingham Island or Mutine Point, depending on proximity. A categorical measure of wave exposure provided by Druehl & Elliot (1996) was used for analyses of all 87 sites. Barnacle upper limit was significantly different between these wave exposure categories (ANOVA: F2,52 = 19.5815, P < 0.0001) for all means (Tukey HSD < 0.05), suggesting that barnacle data are an appropriate proxy for wave exposure.  5.3.1 Testing for a phylogenetic signal on traits and community assembly  All statistical analyses were performed in R in the packages “ape” (Paradis et al., 2004), “phytools” (Revell, 2012), “picante” (Kembel et al., 2010), “qvalue” (Bass et al., 2018), and “cooccur” (Griffith et al., 2016). Principal component analyses were conducted on the correlation matrix of centered and scaled data. The phylogenetic signal of traits was measured using Blomberg’s K (Blomberg et al., 2003) and Pagel’s λ (Pagel, 1999). In order to visualize patterns of trait evolution across the kelp phylogeny, I conducted continuous trait ancestral state reconstruction. Ancestral states were reconstructed using maximum likelihood and states were interpolated along each edge using the method of Felsenstein (1985). In order to test for significant associations between species, observed co-occurrence probabilities were calculated for each pair of species, which were then compared to a null expectation of species co-occurrence that was generated using randomizations that considered only the number of sites at which each species was found. In cases where species were expected to co-occur at less than one site, these species pairs were excluded due to insufficient sample size. Deviations from  112 expectations were measured using a log response ratio of observed vs. expected outcomes, calculated as:  !"#	%&'(")'&	%*+," = !"#./ 012345367893:;36 + 1    Eq. 5.1 where “Observed” refers to the actual number of co-occurrences in the community matrix, and “Expected” refers to the number of sites that species were expected to be found together given the null model. Species association analyses were corrected for false discovery rate using the methods of  Benjamini and Hochberg (1995) and were considered significant when q-values were less than 0.05. In order to determine whether phylogenetic distance or trait differences (first and second trait-derived principal components) influenced the co-occurrence probability of species, linear regressions were fit between each predictor (phylogenetic distance, PC1 distance and PC2 distance) and this log response ratio.  The relationship between species presence and wave exposure was measured in two ways using the subset of sites (n = 55) for which continuous wave exposure (barnacle upper limit) were measured. This subset did not include any sites with S. latissima, which was therefore excluded from these analyses. It also included only one observation of P. palmaeformis at one of the most wave-exposed sites in our dataset. This species is well known to occur only on the most wave-exposed shores (Nielsen et al., 2006) and so this site was deemed representative of the niche of P. palmaeformis and the species was included. To assess the relationship between traits and species habitat use, average wave exposure was measured for each species from all sites in which that species was present. A phylogenetic least-squares (PGLS) regression was then used to test for an effect of PC1, PC2 and all seven quantitative traits on average wave exposure. In order to further visualize differences in species habitat use, the probability of species presence was plotted against wave exposure, as modeled using loess local regression (i.e., moving  113 averages). This was done separately for members of the two subclades with the most species included here, the families Arthrothamnaceae and Alariaceae.  To test for an effect of phylogeny on community assembly, the standard effect sizes (SES) of mean paired distance (MPD) and mean nearest taxon index (MNTD) were calculated for each site (Kembel et al., 2010). MPD is the average distance between all species in a community, and MNTD is the average distance between each species and the nearest species with which it coexists at each site. Each of these metrics was calculated and compared to pseudo-communities randomly drawn from the species pool (n = 10,000). The standard effect size was then calculated by using the mean and standard deviation of randomly assembled pseudo-communities:     >?> = @ABCD@EFGGHIEFGG      Eq. 5.2 where Xobs represents the observed MPD or MNTD value of a community and Xnull and SDnull represent the mean and standard deviation, respectively, of pseudo-communities simulated using random draws of species. Negative SES values indicate that communities are more phylogenetically similar than expected by chance (clustering), while positive SES values indicate that communities are more different than expected by chance (overdispersion). The significance of trends in phylogenetic structure was evaluated in two ways. First, at a community level, sites (i.e. individual communities) were considered to be significantly structured by phylogeny if SES values ranked among the 500 most extreme values (97.5th or 2.5th percentiles) of the 10,000 randomly generated communities. A second approach was used to determine if, across the whole dataset, there were significant trends in phylogenetic community structure. SES (MPD) and SES (MNTD) are both expected to be approximately normally distributed with a mean of zero. Therefore, in order to determine whether the mean of the distribution of kelp communities  114 differed from this null expectation, t-tests were also performed, treating sites as replicates (as in Cooper et al., 2008).   5.4 Results 5.4.1 Ordination and phylogenetic signal of functional traits Principal component analysis resulted in seven component axes with the first two explaining 63.9% of the variation in trait values (Fig 5.1A). Principal component 1 (PC1) correlated with structural characteristics of the whole kelp (HMF and SMF), as well as the blade (DMC, BMA). Principal component 2 explained mainly the properties of materials (σ, E and ε). These two components explained 35.3% and 28.6% of the total variation in functional traits, respectively. There was no significant phylogenetic signal on any of the traits investigated in this study, including principal components (Table 5.1, Fig 5.1B). However, this analysis revealed a possible but not significant phylogenetic signal on SMF (Blomberg K: 0.860, P = 0.063; Pagel’s λ = 1.128, P = 0.085). Some pairs of closely related species were somewhat similar in at least some traits (e.g., Pleurophycus gardneri and Pterygophora californica), but for the most part, closely related species differed as much as distantly related ones (Fig 5.1B). This observation was confirmed by the lack of a significant relationship between PC1 and PC2 trait distances and phylogenetic distance (PC1 Mantel test: Z-stat = 6450.835, p = 0.589; PC2 Mantel test: Z-stat = 6449.193, p = 0.691). Ancestral state reconstructions indicate that trait extremes have evolved multiple times over the course of kelp evolution, indicative of convergent evolution (Fig 5.2). For the eight species that overlapped with sampling in Chapter 3, PC1 correlated strongly with  115 streamlining (S; Linear model: F = 10.65, df = 6, P = 0.01718) and tolerance (T; Linear model: F = 11.92, df = 6, P = 0.0136) indices (Fig 5.3). 116    Fig 5.1. Phylogenetic distribution of trait axes in northeast Pacific kelp species. Panel A shows the first two principal component axes. Panel B shows PC1 and PC2 plotted on the phylogeny. Dot size indicates the value and of the principal component; black dots are negative and white dots are positive.   Egregia_menziesiiEcklonia_arboreaCymathaere_triplicataSaccharina_latissimaSaccharina_nigripesSaccharina_sessilisMacrocystis_pyriferaPostelsia_palmaeformisNereocystis_luetkeanaLaminaria_setchelliiLaminaria_ephemeraNeoagarum_fimbriatumCostaria_costataAlaria_marginataLessoniopsis_littoralisPleurophycus_gardneriPterygophora_californicaP C1P C2Egregia_menziesiiEcklonia_arboreaCymathaere_triplicataSaccharina_latissimaSaccharina_nigripesSaccharina_sessilisMacrocystis_pyriferaPostelsia_palmaeformisNereocystis_luetkeanaLaminaria_setchelliiLaminaria_ephemeraNeoagarum_fimbriatumCostaria_costataAlaria_marginataLessoniopsis_littoralisPleurophycus_gardneriPterygophora_californicaPC1PC21Pterygophora	californica2Pleurophycus	gardneri3Lessoniopsis	littoralis4Alaria	marginata5Costaria	costata6Neoagarum	fimbriatum7Laminaria	ephemera8Laminaria	setchellii9Nereocystis	luetkeana10Postelsia	palmaeformis11Macrocystis	pyrifera12Saccharina	 sessilis13Saccharina	 nigripes14Saccharina	 latissima15Cymathaere	triplicata16Ecklonia	arborea17Egregia	menziesiiEgregia_menziesiiEcklonia_arboreaCyathaere_triplicataSaccharina_latissimaSaccharina_nigripesSaccharina_sessilisMacrocystis_pyriferaPostelsia_palmaeformisNereocystis_luetkeanaLaminaria_setchelliiLaminaria_ephemeraNeagarum_fimbriatumCostaria_costataAlaria_marginataLessoniopsis_littoralisPleurophycus_gardneriPterygophora_californicaP C1P C2-3 -2 -1 0 1 2 3-3-2-10123PC1PC2StiffnessStrengthExtensibilityMass	per	areaDry	weight	contentHoldfast	fractionStipe	fractionPrincipal	Component	 1	(35	%	variation)Principal	Component	2	(29	%	variation)25614103168129119141513177A BDry	matter	content 117    Fig 5.2. Ancestral state reconstruction of continuous traits, PC1 (A) and PC2 (B). Colour indicates trait value (given in each inset legend). Values at internal nodes were estimated using maximum likelihood and values along each branch were estimated using the methods of Felsenstein (1985).   118  Fig 5.3. Principal component 1 versus streamlining and tolerance indices for the eight species of northeast Pacific kelps investigated in Chapter 3.  Table 5.2. Statistical testing of phylogenetic signal for quantitative traits.   Phylogenetic Signal    Functional Traits Blomberg’s K P-value Pagel’s λ P-value PC1 0.538 0.610 <0.01 >0.99 PC2 0.612 0.425 <0.01 >0.99 HMF 0.353 0.693 <0.01 >0.99 SMF 0.860 0.063* 1.128 0.085* BMA 0.718 0.190 <0.01 >0.99 DMC 0.521 0.649 <0.01 >0.99 Strength 0.584 0.457 0.108 0.737 Stiffness 0.720 0.197 0.303 0.437 Extensibility 0.285 0.962 <0.01 >0.99 *Trending towards significance (P < 0.10)  5.4.2 Community assembly Communities were assembled non-randomly with several significant associations between species (Fig 5.4). Positive and negative species associations occurred between both closely and distantly related species pairs. For example, closely related species Macrocystis −2 −1 0 1 2 30.00.20.40.60.81.0PC1Tolerance Index−2 −1 0 1 2 3−0.10.00.10.20.30.4PC1Streamlining IndexA B 119 pyrifera and Nereocystis luetkeana were negatively associated with each other, while sister taxa, Pleurophycus gardneri and Pterygophora californica, were positively associated (Fig. 5.4). Moreover, Egregia, the most phylogenetically distinct genus from the family Arthrothamnaceae, was positively associated with some members of three other families (Alariaceae, Agaraceae, Laminariaceae) and negatively associated with a member of one (Agaraceae). There was no clear pattern of phylogenetic clustering among species pairs that were significantly correlated (Fig 5.4).  Despite clear evidence of non-random community sorting, there was no effect of phylogeny on pairwise associations between species. The only significant predictor of pairwise non-random co-occurrence (measured as log response ratio) was distance in PC1 between species pairs (Linear regression: F=5.075, df=69, P=0.02746; Fig 5.5). Phylogenetic distance (Linear regression: F=0.2392, df=69, P=0.6263; Fig 5.5) and PC2 distances (Linear regression: F=0.3037, df=69, P=0.5833; Fig 5.5) did not significantly correlate with the pairwise co-occurrence of species. Use of phylogenetic indices further corroborate the conclusion that communities are not phylogenetically clustered. No communities examined were significantly phylogenetically clustered and most communities trended towards phylogenetic overdispersion relative to randomly simulated communities (Fig 5.6). Although only a few sites were significantly overdispersed (MPD: n = 3, MNTD = 7; Fig 5.6), average phylogenetic SES (MPD) and SES (MNTD) values were significantly different from zero (MPD: t-test: t = 3.917, df = 86, p = 0.00018; MNTD: t-test: t = 9.4708, df = 86, p < 0.0001), suggesting that weak phylogenetic overdispersion is a widespread occurrence across the communities examined here. The few communities that trended towards phylogenetic clustering were composed of a small number of species and the slightly negative SES value could have arisen by chance in them.   120  Fig 5.4. Correlation matrix of species pairs. Colour in each cell indicates whether there was a significant positive or negative correlation between the occurrences of each pair of species, after correcting for false discovery rate (q < 0.05).  Fig 5.5. Species covariation versus phylogenetic and trait dissimilarity. Log response ratio [Log ((observed co-occurrence / expected co-occurrence) + 1)] versus (A) phylogenetic distance between species pairs in millions of years, (B) distance in PC1 for each species pair and, (C) distance in PC2 for each species pair. Dotted lines indicate non-significant trends, while the solid blue line in panel B indicates a significant slope.  121   Fig 5.6. The standard effect size of phylogeny-based indices (MPD and MNTD) of community assembly plotted against the species richness of communities. Data points represent individual communities, and significance is indicated with dot colour. Black dots indicate that communities are significantly structured by phylogeny (i.e. significant overdispersion), while grey dots indicate no significant phylogenetic effect.  There was a significant relationship between the average wave exposure of a species and its value of PC1 (Linear model: F = 6.809, df = 12, P = 0.0228; PGLS model: t = 3.9823, df = 14, P = 0.002; Fig 5.7), but not PC2 (Linear model: F = 0.1225, df=1 and 12, P = 0.732; PGLS model: t = 0.8316, df = 14, P = 0.4219), such that structurally reinforced species tended to be found at more wave exposed sites. The only traits that significantly correlated with the average wave exposure of a species on their own were holdfast mass fraction and tissue extensibility, indicating that kelps growing at more wave exposed sites tended to have larger holdfasts and more extensible blade tissues (Table 5.3). There was a possible, but not significant negative 2 4 6 8 10−3−2−10123Species RichnessSES (MPD)2 4 6 8 10−3−2−10123Species RichnessSES (MNTD) 122 correlation between blade stiffness and average wave exposure, suggesting that species at wave-exposed sites also tended to have more flexible tissues.   Local regressions of species presence and absence along a continuous wave exposure axis further demonstrates how species in each subclade occupy sites that differ in wave exposure (Fig. 5.8). Individual species vary in distribution across the gradient of wave exposure and closely related species (e.g., Nereocystis luetkeana and Macrocystis pyrifera) tend to specialize in different wave exposure regimes. The clear exception here is the sister species pair Pterygophora californica and Pleurophycus gardneri (Fig. 5.4) that have nearly identical distributions across the wave exposure gradient (Fig 5.8). There was a significant effect of wave exposure category on community composition (PERMANOVA: F =13.205, P < 0.001; Fig 5.9), indicating that differences in species distributions across the wave exposure gradient scale up to community level differences in species composition at wave-exposed versus wave-sheltered sites.   Table 5.3. Results of PGLS models testing for correlations between traits and average wave exposure of species (df = 12). Functional Traits t-statistic P-value PC1 3.9283 0.0020** PC2 0.8316 0.4219 HMF 3.8602 0.0023** SMF 0.9203 0.3756 BMA 1.0040 0.3351 DMC 1.5138 0.1560 Strength 0.8776 0.3974 Stiffness -2.1020 0.0573* Extensibility 2.2003 0.0481** *Trending towards significance (P < 0.10) **Significant relationship  (P < 0.05) 123    Fig 5.7. Relationship between wave exposure and principal component 1. Data points represent the average wave exposure that a species was found at (+/- variance) plotted against its value of PC1. −2 −1 0 1 23.54.04.55.0Principal Component 1Wave Exposure1234567891011121413−2 −1 0 1 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 1 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 1 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 1 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 1 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 1 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 1 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 23.54.04.55.0Principal Component 1Wave Exposure−2 −1 0 1 23.54 0.55.0Principal Component 1Wave Exposure−2 −1 0 1 23.54 0.55.0Principal Component 1Wave Exposure−2 −1 0 1 23.54 0.55.0Principal Component 1Wave Exposure141312345678910111213Neoagarum fimbriatum (	n		=	2	)Macrocystis pyrifera(	n		=	33	)Alaria marginata (n	=	32	)Ecklonia arborea (	n		=	11	)Postelsia palmaeformis (	n		=	1	)Lessoniopsis littoralis (	n		=	12	)Saccharina sessilis (	n		=	22	)Nereocystis luetkeana (	n		=	14	)Pleurophycus gardneri (	n		=	5	)Costaria costata (	n		=	18	)Pterygophora californica(	n		=	11	)Saccharina nigripes (	n		=	2	)Egregia menziesii(	n		=	46	)Laminaria setchellii (	n		=	32	) 124   Fig 5.8. Local regression (loess) of species occupancy across a gradient of wave exposure. Columns represent members of two different kelp families (left = Alariaceae, right = Arthrothamnaceae).     125  Fig 5.9. NMDS plot of community presence data for kelp species at 87 sites in Barkley Sound, British Columbia. Sites are coloured by wave exposure category (red = exposed, green = moderate, blue = sheltered) and lines are drawn between all sites and the centroid of the wave exposure category.  5.5 Discussion Contrary to the pervasive assumption that phylogenetic relatedness can serve as a proxy for species ecology (Webb, 2000; Webb et al., 2002; Kraft et al., 2007), I found no significant effect of phylogeny on structural traits or habitat use in northeast Pacific kelps. Principal component analysis resulted in two axes that explain nearly two thirds of the variation in trait values exhibited across the sampled kelp species. Yet, neither of these axes, nor any of the  126 individual traits that make them up, were found to be phylogenetically conserved across species. The first principal component axis (PC1), which describes structural reinforcement of kelp thalli and correlates with the maximum forces that a species can resist (Fig 5.3A), was also significantly related to the average wave exposure of species (Fig. 5.7). This provides evidence that species are filtered along the investigated wave exposure gradient by their ability to tolerate wave forces. In Chapter 3, I demonstrated that weaker species are better at streamlining and reducing drag from water motion. Here, I show that weaker species may be filtered from the most wave-exposed sites, occupying only moderately exposed and sheltered sites. In spite of this evidence for environmental filtering, these results suggest that traits and niche (with respect to wave exposure) are not phylogenetically conserved among closely related species. Thus, these data refute the hypothesis that ecologically relevant traits are similar among closely related species in northeast Pacific kelps. Given this lack of phylogenetic signal on traits, kelps provide an excellent opportunity to investigate trait and niche evolution across an environmental gradient when traits involved in the process are labile through evolutionary time.  Species co-occurrence patterns and community ordination reflect the influence of habitat filtering on community composition. Species that had positively associated traits tended to be those found to specialize in the same wave exposure regime. For example, Lessoniopsis littoralis and Saccharina sessilis, two distantly related species that are similar in PC1 (Fig 5.1B) and specialize in wave swept environments (Fig 5.8), were positively correlated across the community matrix (Fig 5.4). However, species that specialize in different wave exposure regimes tended to be negatively correlated. For example, Neoagarum fimbriatum, a specialist in wave sheltered areas, and Laminaria setchellii, a wave exposed specialist, were negatively correlated (Fig 5.4). Across all species, pairs that were similar in PC1 also tended to positively  127 co-occur and vice versa; there was a negative relationship between PC1 distance and log response ratio of species co-occurrence (Fig 5.5B). This evidence for trait clustering suggests that wave exposure has a significant effect on community assembly through environmental filtering. This suggests that co-occurrence patterns of species pairs (Fig 5.4) resulted from correlated niches, rather than through direct density-dependent interactions such as competition. These patterns of species specialization then scale up to drive significant differences between communities that differ in wave exposure, as determined by PERMANOVA (Fig 5.7). Overall, these results suggest that wave exposure plays a key role in the assembly of communities and has influenced trait evolution across the clade. Phylogenetic community indices (MPD and MNTD) in this system reveal that distantly related species are more likely to coexist than in randomly assembled pseudo-communities, suggesting weak phylogenetic overdispersion of kelp communities. Under the commonly employed approach of assuming phylogenetic signal on ecologically-relevant traits, phylogenetic overdispersion of communities would likely be interpreted as evidence for competitive exclusion (e.g. Webb, 2000; Webb et al., 2002; Cooper et al., 2008). However, given the lack of phylogenetic signal here and the evidence for environmental filtering described above, this interpretation would be flawed in kelps. Traits shown here to influence the ecological niche of a species across this gradient are phylogenetically convergent across the tree rather than conserved. Thus, kelp communities are generally composed of species that are similar in traits associated with mechanical reinforcement (consistent with environmental filtering), but are distantly related.    128 5.5.1 Niche partitioning along a disturbance gradient Character displacement between competing species or populations is known to drive niche partitioning and is thought to be a key component of the process of adaptive radiation (Schluter, 1996, 2000). In this system, I report evidence for environmental filtering while also demonstrating phylogenetic overdispersion of kelp communities (Fig 5.9). Moreover, within each clade, closely related species often showed entirely different trait assemblages (Fig 5.1, Fig 5.2) and distributions across the gradient of wave exposure (Fig 5.8). Ancestral state reconstructions suggess that trait combinations have evolved convergently across the kelp phylogeny (Fig 5.3). Although it is difficult to predict the ancestral trait combinations of each lineage, extreme trait values (both high and low) have appeared across multiple subclades and could not be the result of phylogenetic inertia (Fig 5.3),  This suggests that ecological niches evolved convergently across the kelps, leading to kelp communities that are composed of distantly related species that are each separately adapted to the habitats in which they are found. Thus, closely related species diverged in traits to allow for adaptation to different types of habitats. In this way, close relatives may specialize in different types of environments, maintaining coexistence of species across broad geographic scales (Cavender-Bares et al., 2004b,a; Losos, 2008). These results provide clear evidence that traits are not always phylogenetically conserved, and that phylogenies are not proxies for ecological differences between species, but instead provide insights into the evolutionary forces that likely shaped modern communities (as argued by Gerhold et al., 2015).  Although wave exposure clearly plays important roles in structuring intertidal kelp communities, there are several aspects of ecological niche not captured by my investigations of this gradient alone. In particular, distributions of kelp species differ in tidal heights, with some  129 species specializing in habitats high on the shore and others specializing in low intertidal or subtidal zones (Dayton, 1985; Estes & Steinberg, 1988; Druehl & Elliot, 1996). Moreover, shoreline characteristics and light availability have both been shown to act as significant predictors of species distributions (Bekkby et al., 2009). It is likely that inclusion of tidal height measurements or other environmental characteristics into comparisons of species distributions would demonstrate even more niche partitioning, even among closely related species. For example, in the Atlantic Ocean, sister taxa Laminaria digitata and L. hyberborea are often found at the same sites but at different tidal heights (Robuchon et al., 2017). Structural characteristics and some material properties correlated with the wave exposure at sites where species are most commonly found. Principal component 1, which represents the structural reinforcement of a species, was significantly correlated with average wave exposure, as was its constituent trait holdfast mass fraction (HMF; Table 5.3). This finding is consistent with previous work conducted on morphological variation within species that demonstrated that large holdfasts are selected for at wave exposed sites (e.g. Duggins et al., 2003; Wernberg, 2005; Demes et al., 2013). There was no significant effect of principal component 2 on the average wave exposure at sites where species are found, but there was a significant relationship between extensibility and wave exposure (Table 5.3). Thus, communities at the most wave-exposed sites were composed of species with tissues that were highly extensible, but not particularly strong. Although the adaptive basis of material properties in seaweeds have been widely discussed (e.g. Denny & Gaylord, 2002; Harder et al., 2006; Martone, 2007b; Demes et al., 2013; Janot & Martone, 2016), it is not entirely understood. These results demonstrate that material strength is not a prerequisite for life on wave exposed shores, but that flexible and more extensible tissues may instead be selected for (Koehl & Wainwright, 1977). This lack of a significant relationship  130 between wave exposure and PC2 underlines that multiple strategies can be used to resist dislodgement from waves (Martone, 2007a; Martone et al., 2012; Starko et al., 2015; Starko & Martone, 2016b), and many seaweeds with low tissue strength can still resist enormous forces (Martone, 2007a). Thus, material strength alone may not be an informative measure of how much force a kelp can resist. Instead, increased tissue extensibility and increased flexibility appear to be more indicative of species that specialize in high wave exposure (Table 5.3). This study supports growing evidence that kelps underwent widespread adaptive radiation. In Chapter 2, I demonstrated that kelp diversification may have occurred as a result of ecological opportunity associated with an increasing availability of cold water habitat. In this chapter, I demonstrate that traits and ecological niche are not correlated with phylogeny, supporting the hypothesis that adaptive radiation is widespread throughout the kelps. This is not the first example of adaptive radiation resulting in clades with little to no phylogenetic signal in traits. For example, Hawaiian silverswords cover nearly the entire range of plant traits, despite their recent diversification only about 5 MYA (Ackerly, 2009; Blonder et al., 2016). Silverswords also vary in body plan from shrubs and trees to cushion plants, lianas and rosette plants (Carr et al., 2003). Ecological opportunities caused by heterogeneity in habitat types and wide availability of resources has been used to explain how such striking diversity could have arisen so quickly from a single ancestor (Blonder et al., 2016). For the kelps, cooling of the north Pacific Ocean and extinction of tropical taxa that likely previously occupied these areas may have opened up ecological niche space across which the kelps radiated (Chapter 2). Following initial phylogenetic splits between what are now different kelp families, members of each subclade have evolved to fill available niche space (Fig 5.8) as global climate continued to cool and kelps migrated southward (Chapter 2). As these subclades independently radiated to  131 specialize in different parts of the newly available habitat, distantly related species likely converged in their realized ecological niche, giving rise to the observed overall pattern of phylogenetically overdispersion observed across kelp communities.  5.5.4 Conclusions In this chapter, I demonstrated that morphological traits are highly labile in northeast Pacific kelps and that morphological extremes evolved repeatedly across the kelp phylogeny. I further demonstrated that some morphological traits are associated with species that specialize in different types of wave regimes. Thus, the morphological evolution of kelps was driven by convergent evolution of ecological niches, where closely related species diversified to specialize in different wave exposure regimes. No morphological trait investigated in this study showed a significant phylogenetic signal, indicating that similar trait combinations evolved convergently across species. This has led to a lack of a phylogenetic signal on traits, but also phylogenetic overdispersion of coexisting kelp communities. This pattern of overdispersion may have arisen through a long history of competitive interactions between closely-related kelp species that led to repeated niche partitioning within subclades. These results demonstrate that phylogenetic signals may be eroded in lineages that have radiated under ecological opportunity and support an approach that investigates, rather than assumes, a phylogenetic signal. While assumptions about trait evolution may be appealing when trait data are not available, this study instead highlights the importance of understanding the relationship between species evolutionary histories and their trait combinations.  Only by first clarifying phylogenetic patterns in emergent traits can we understand how local scale patterns of community assembly feedback with macroevolutionary  132 processes of diversification to determine the patterns of species distribution observed in the modern world. 133 6. Conclusions 6.1 Key findings With this thesis, I aimed to explore species characteristics in an evolutionary context to better understand how morphological variation has arisen through evolutionary time and how morphology interacts with physical factors to influence where species can live in the environment. I began by reconstructing the relationships between kelp genera from around the world and performed the most comprehensive phylogenomic analysis of kelps to date (Ch. 2). This allowed me to investigate the evolution of morphological adaptations, the timing of key radiations and the biogeographic dispersal pathways of various lineages (Ch. 2). Throughout successive chapters, I explored how morphological variation is distributed across the phylogeny and how biophysical constraints might have led to trade-offs in growth strategy, trait evolution and ecological niche. I compared the morphological strategies of eight co-occurring kelp species and explored interspecific variation in drag tolerance and drag reduction (i.e., streamlining) strategies (Ch. 3). I showed that patterns of interspecific scaling of organ biomasses are similar between kelps and land plants, despite kelps allocating proportionally more biomass to blades than land plants do to leaves (Ch. 4). I also showed that subtidal kelps tend to have smaller holdfasts per unit blade biomass than intertidal kelps, suggesting that selective pressures on morphology differ between habitats (Ch. 4). Lastly, I used phylogenetic, trait and community data for intertidal kelps to investigate whether traits are phylogenetically conserved or convergent and whether traits or phylogeny could predict where species are found in the environment. I demonstrated that while morphological traits are correlated with species  134 distributions across a gradient of wave exposure, consistent with environmental filtering, traits and niche structure have evolved repeatedly across subclades (Ch. 5).  Three patterns have emerged from this body of work as a whole: (1) environmental factors have played an influential role in the diversification dynamics of kelps and the strategies that they have evolved, (2) biophysical factors have constrained ecological adaptation and (3) adaptations to differing habitats have evolved convergently across various subclades. In this concluding chapter, I explore these emergent patterns in more detail and then discuss some of the limitations of my findings and make suggestions for future work.  6.1.1 Environmental factors have influenced kelp evolution Throughout this dissertation, I have presented evidence that environmental factors (on both broad and fine scales) have influenced the evolutionary dynamics of kelps. At broader scales, temporal patterns of kelp evolution are consistent with a possible role of changing global climate in stimulating kelp radiation (Ch. 2). The complex kelps began diversifying around the time of the Eocene-Oligocene boundary, when the North Pacific began to rapidly cool, consistent with the hypothesis that global cooling opened up new habitat at mid-latitudes, creating an ecological opportunity for the kelps. Although climatic influence on kelp diversification had been hypothesized before (Estes & Steinberg, 1988), it had been debated (Domning, 1989; Estes & Steinberg, 1989; Vermeij et al., 2018) and had not been tested using a formalized analysis of kelp diversification rates through time. Newly available habitat is a common driver of adaptive radiation, since resources in newly invaded habitat are often unexploited (Schluter, 1996, 2000). Thus, adaptive radiation stimulated by newly available temperate habitat may help to explain the wide ecological diversity of kelps observed in modern marine floras.  135 Fine scale environmental drivers have also influenced the morphological diversification of kelps. Subtidal kelps tend to invest less in holdfast biomass than intertidal kelps, suggesting that survival under breaking waves may require increased investment in structural support material relative to other habitats (Ch. 4). Moreover, traits associated with mechanical reinforcement are correlated with the average wave exposure where a species is found and species with similar measures of mechanical reinforcement tend to co-occur (Ch. 5). This suggests that environmental filtering plays an important role in kelp community assembly, perhaps by preventing the establishment of less mechanically supported species at sites with high wave exposure. The niches of closely related species tended to be distinct from one another and closely related species would often differ in the wave exposure regimes where they were found (Ch. 5). Together, this suggests not only that wave exposure and morphology are correlated, but that divergent selection may drive close relatives to occupy different habitats, with wave exposure being an influential environmental axis. Evidence that specialization to habitats that differ in environmental conditions may drive divergent selection has been reported previously among conspecific populations. In particular, previous work has demonstrated how speciation may occur along environmental gradients, including wave exposure. Augyte et al (2017) described a new species of Saccharina, S. angustissima, based on morphology, ecology and molecular distance that is endemic to the western Atlantic and was previously regarded as a forma of S. latissima. Where S. latissima sensu stricto has a broad, ruffled blade and is generally found in low to moderate flow environments, S. angustissima is found at wave exposed sites and has a thick, narrow, strap-like blade (Augyte et al., 2017). Genetic distance between nearby populations of S. angustissima and S. latissima support the hypotheses that populations are not interbreeding and suggests that these  136 species have diverged across a gradient of wave exposure. Although, more recent work by Neiva et al (2018) and single-gene analyses presented in Appendix A suggest that S. angustissima is part of a larger species complex that is regarded only as S. latissima, their results suggest that habitat partitioning among closely related populations may drive reproductive isolation and ultimately speciation. Other examples of very close relatives (or semi-isolated populations of the same species) differing in habitat specialization have been documented elsewhere across the kelp phylogeny (Roberson & Coyer, 2004; Lane et al., 2007; Tellier et al., 2011). These past examples found near the tips of the phylogeny parallel the clade-wide patterns of ecological divergence between close relatives documented in Chapter 5. Thus, collective evidence suggests that niche differentiation across environmental gradients, including wave exposure, contributes to clade-wide morphological diversity.  6.1.2 Physical limitations have constrained morphological evolution I identified two cases where morphological evolution has been constrained by physical limitations. There is a trade-off between streamlining and tolerance strategies that is likely tied to a biomechanical constraint (Ch. 3). Although thick tissues and stipe(s) may allow species to resist forces imposed by waves, this may limit their ability to bend and reconfigure in flow, causing robust species to experience more drag. In contrast, weaker species may have thinner, more flexible tissues and stipes that allow them to fold up and reduce the drag that they experience under breaking waves. This has resulted in a continuum from wave “tolerators” to “streamliners”, where species specialize in tolerating or reducing drag, with some species occupying intermediate strategies but no species strongly specializing in both strategies. A trade-off between drag reduction and drag tolerance has been previously demonstrated in aquatic land  137 plants (Puijalon et al., 2011). Thus, new data presented here on kelps suggest that this trade-off may be widespread and not taxon-specific. Instead, this trade-off may arise from the physical consequences of increasing dislodgement force and structural reinforcement. Increasing the thickness of tissues (especially at the point of bending) can reduce the capacity of species to bend and reconfigure in flow. Thus, physical consequences of geometry may constrain the range of morphologies observed on wave-swept shores. The second example of physical limitations on kelp morphology comes from examinations of interspecific allometry (Ch. 4). Blade and holdfast biomasses scale with negative allometry across kelp species, such that larger species have proportionally larger holdfasts. This finding is likely a consequence of how surface area and volume scale as objects increase in size. As kelps grow larger, they experience more drag and species can compensate by increasing their attachment to the substratum. This pattern is clearly supported by the finding that, for all of the species that I assessed in Chapter 3, tenacity and drag both increased across all species with increasing photosynthetic area. Given that holdfast attachment area and holdfast biomass are likely to scale with negative allometry (e.g. A ~ V2/3 in many objects), larger holdfasts may be less efficient. This may drive larger kelps to invest disproportionately in holdfast biomass to maintain sufficient attachment to the substratum. This is analogous to patterns observed in land plants where both whole plants (Niklas & Enquist, 2001; Enquist & Niklas, 2002b) and leaves (Niklas & Cobb, 2008) grow with patterns of ‘diminishing returns’ due to interactions between size and mechanical or hydraulic demands (Niklas & Spatz, 2004; Niklas et al., 2009; see Ch. 4).    138 6.1.3 Adaptations are convergently distributed across the phylogeny Another emergent pattern is that morphological variation has evolved convergently across the phylogenetic tree of kelps. Seemingly complex morphological characteristics, such as branching and upright stature have evolved repeatedly in various subclades, as kelps radiated in the North Pacific (Ch. 2). This pattern of widespread morphological convergence is paralleled by the results of Chapters 3 and 5, where I found no evidence of phylogenetic signal on kelp strategies and ecological niches. The streamliner-tolerator trade-off discussed above was not related to phylogeny, and phylogenetic comparative methods suggested that extremes of each strategy have evolved multiple times, even across the relatively small subset of species explored in that study (Ch. 3). Quantitative traits and ecological niche also do not correlate with phylogeny, and closely related species are just as likely to differ in this context as distantly related species (Ch. 5). Again, extremes of traits and of ecological niche have evolved repeatedly across the tree. This has led to an overall pattern of kelp communities that are phylogenetically overdispersed but similar in traits (Ch. 5).  The widespread nature of convergent evolution observed across the kelps provides further evidence that adaptive radiation has been a key process in the diversification of kelps. Closely related species may compete strongly (Cahill et al., 2008) and this could lead to divergent selection and large ecological differences among close relatives (Schluter, 2000; Losos, 2008; Ackerly, 2009). Convergence in niche might then arise if distantly related species are more capable of coexistence, and species evolve similar niches to distantly related species but different niches from closely related species. This partitioning of habitats could help to maintain coexistence of species at broad geographic scales (Cavender-Bares et al., 2004b,a).   139 6.2 Limitations and future directions 6.2.1 Intraspecific variation Throughout this dissertation, I quantified traits and developmental strategies using species averages derived from samples of individual populations. Although this approach allowed for novel insights into kelp evolution at the scale at which I was interested, intraspecific variation in trait values was not adequately explored. Phenotypic plasticity and local adaptation are known to drive morphological variation across populations of the same kelp species (e.g. Gerard, 1988; Johnson & Koehl, 1994; Duggins et al., 2003; Koehl et al., 2008b). While interspecific variation is likely to strongly exceed the variation exhibited by a single species, populations of the same species do often differ in morphological traits and strategies in ways that are important for their ecological niches (Gerard, 1988; Johnson & Koehl, 1994; Duggins et al., 2003; Roberson & Coyer, 2004; Koehl et al., 2008b). Duggins et al (2003) found that populations of generalist species growing under high flow environments were more mechanically reinforced than populations of the same species found at sites with lower flow rates, suggesting that high flow populations are better adapted to high energy environments than populations growing under slow flow. I hypothesize that incorporating this intraspecific trait variation into analyses may allow for improved predictions of how species are distributed across environmental gradients in nature. For example, some of the species shown to be ‘generalists’ across the gradient of wave exposure (Ch. 5) are also those that are highly plastic or have wide genetically-determined intraspecific variation. For example, Nereocystis luetkeana has been shown to exhibit phenotypic plasticity across gradients of water motion (Johnson & Koehl, 1994; Koehl et al., 2008b) and was also found across a wide range of wave exposures in the Barkley Sound community dataset. Similarly, Alaria marginata was found to be relatively cosmopolitan  140 and was found at a wide range of sites. There is strong sub-species molecular structure in Alaria marginata with different ‘morphs’ found in different wave exposure environments. For example, the “Alaria nana” morph is found predominately at sites with high wave exposure. The different morphs of Alaria marginata are likely to be genetically determined, with morphs generally differing from one another in sequence data (Lane et al., 2007). However, they are known to interbreed and are therefore considered a single species (Lane et al., 2007). Given the new phylogenetic inferences reported in Chapter 2, future work should rigorously examine the evolutionary distribution of intraspecific variation and explore its limits across the phylogeny.  6.2.2 Traits and productivity In order to better understand the link between traits and ecological niche, and to scale up differences in traits to community dynamics and productivity, future work should explore the relationship between morphology and growth rate. In this dissertation, I documented that species exhibit a wide range of traits and that organ investment depends on both the size of a species and on the type of habitat in which it is found (Ch. 4). Blade mass per area has been found to correlate negatively with productivity in kelp species from Asia (Sakanishi et al., 2017), suggesting that traits influence productivity. Moreover, previous work on Macrocystis pyrifera has shown that holdfasts negatively affect the carbon budget of kelps; despite being weakly pigmented, holdfasts use more energy through respiration than they fix through photosynthesis (Arnold & Manley, 1985). These two studies, although limited in scope, suggest that there are consequences of morphology on photosynthesis and carbon budgeting. An improved understanding of how morphological traits influence productivity and growth rate of kelps could  141 improve our ability to model ecosystem productivity and understand potential trade-offs between mechanical reinforcement and other ecological attributes, such as competitive ability.  Testing for correlations between morphological traits and productivity could be accomplished in a variety of ways. For example, productivity could be directly measured in the lab (as O2 production or CO2 consumption) for kelp species that differ in morphological traits, such as biomass allocation and tissue properties. Productivity could also be inferred in situ from growth rate data. Total growth (in biomass) could be measured in a common garden for species that differ in morphology. In either case, the influence of morphology on productivity could be tested using phylogenetically controlled models (such as PGLS).  6.2.3 Niche modeling In Chapter 5, I used a simple model of ecological niche by investigating the presence and absence of species across one environmental axis: wave exposure. While this approach confirmed that wave exposure is an important environmental gradient associated with niche-based processes, and captured clear differences between how species are distributed across fine scales, there are other environmental variables that are known to influence where species are found in space. In addition to differences in the ability of species to resist waves, species differ in their light requirements (Fortes & Lüning, 1980; Miller et al., 2012), temperature tolerances (tom Dieck, 1993) and desiccation tolerances (Widdowson, 1959). Future work should look at the roles that multiple, interacting environmental variables play in determining the distributions of species on fine scales.  More detailed niche modelling could be accomplished by collecting field data on kelp distributions, such as the dataset by Druehl and Elliot (1996), but measuring additional  142 environmental parameters at each site. For example, sites that differ in aspect and wave exposure may also differ in temperature (Harley, 2003; Harley & Helmuth, 2003; Fitzhenry et al., 2004), which could be measured using digital temperature loggers. Moreover, the upper and lower elevational limits of kelp species could be measured at each site and interactions between tidal height, wave exposure and temperature could be investigated. The presence and absence of kelps could be related to environmental parameters using multivariate approaches, such as newly developed methods to explore and compare n-dimensional hypervolumes (Blonder et al., 2014). This approach would also allow for the putative identification of sites that are habitable to a particular species but are uninhabited at the time of sampling, an inference that could be confirmed using field transplantation. Determining which sites are habitable to a particular species but uninhabited at any point in time could lend insight into processes important to community assembly, such as competitive exclusion and dispersal limitation. Patterns of local niche structure could also be compared across latitudes to determine how broad scale patterns in ocean temperature and upwelling interact with local environmental conditions to influence the distributions of species in space.   6.3 Final thoughts The kelps offer a remarkable example of adaptive radiation and a fascinating playground of morphological diversity. In this thesis, I reconstructed the evolutionary relationships between kelp species and examined the distribution of morphological variation across the phylogeny. In each subclade, kelp species exhibit a wide array of morphological adaptations and have diversified to fill a range of ecological niches. Morphological strategies to resist wave action exist along a continuum from tolerating forces (via attachment) to minimizing them (via  143 streamlining) and extremes of these strategies have evolved repeatedly. There are size-dependent and habitat-dependent patterns of biomass allocation across kelp species, demonstrating how size and environment both influence the morphology of species. Moreover, these size-dependent patterns parallel those observed in land plants, suggesting that shared selective pressures may have shaped the morphological strategies of both groups. Closely related species were found to often differ in morphological traits and ecological distributions, further supporting the hypothesis that ecological speciation has been an important driver of kelp diversification, and calling into question the generality of assumed correlations between phylogeny and ecological traits. Overall, kelps offer an excellent study system for understanding the factors that drive and limit the morphological and ecological diversification of marine species. Moreover, data presented in this dissertation not only improves our understanding of kelp evolution, but also lends insights into the deep history of species that rely on kelps, and into the temporal patterns that led to the formation of modern temperate ecosystems.  144 Bibliography Abbott	I,	Hollenberg	G.	1976.	Abbott,	I.	&	Hollenberg,	G.	1976.	Marine	Algae	of	California.	Stanford	University	Press.	Stanford,	Calfornia.	800	pp.	Stanford	University	Press.	Stanford,	California.	Ackerly	D.	2009.	Conservatism	and	diversification	of	plant	functional	traits:	evolutionary	rates	versus	phylogenetic	signal.	Proceedings	of	the	National	Academy	of	Sciences	106:	19699–19706.	Aguilar-Medrano	R,	Reyes-Bonilla	H,	Polly	PD.	2015.	Adaptive	radiation	of	damselfishes	(Perciformes,	Pomacentridae)	in	the	eastern	Pacific.	Marine	Biology	162:	2291–2303.	Altschul	SF,	Gish	W,	Miller	W,	Myers	EW,	Lipman	DJ.	1990.	Basic	local	alignment	search	tool.	Journal	of	Molecular	Biology	215:	403–410.	Anten	NPR,	Sterck	FJ.	2012.	Terrestrial	vs	aquatic	plants:	how	general	is	the	drag	tolerance-avoidance	trade-off?:	Commentary.	New	Phytologist	193:	6–8.	Archibald	JM.	2009.	The	puzzle	of	plastid	evolution.	Current	Biology	19:	R81–R88.	Arizona	Software.	2010.	GraphClick	3.0.2.	Arnold	KE,	Manley	SL.	1985.	Carbon	allocation	in	Macrocystis	pyrifera	(Phaeophyta):	intrinsic	variability	in	photosynthesis	and	respiration.	Journal	of	Phycology	21:	154–167.	Assis	J,	Araújo,	Miguel	B.,	Serrão,	Ester	A.	2017.	Projected	climate	changes	threaten	ancient	refugia	of	kelp	forests	in	the	North	Atlantic.	Global	Change	Biology	24:	e55–e66.	Augyte	S,	Lewis	L,	Lin	S,	Neefus	CD,	Yarish	C.	2017.	Speciation	in	the	exposed	intertidal	zone:	the	case	of	Saccharina	angustissima	comb.	nov.	&	stat.	nov.	(Laminariales,	Phaeophyceae).	Phycologia	57:	100–112.	Baack	EJ,	Rieseberg	LH.	2007.	A	genomic	view	of	introgression	and	hybrid	speciation.	Current	Opinion	in	Genetics	&	Development	17:	513–518.	Ballhorn	DJ,	Pietrowski	A,	Lieberei	R.	2010.	Direct	trade-off	between	cyanogenesis	and	resistance	to	a	fungal	pathogen	in	lima	bean	(Phaseolus	lunatus	L.).	Journal	of	Ecology	98:	226–236.	Bartsch	I,	Wiencke	C,	Bischof	K,	Buchholz	CM,	Buck	BH,	Eggert	A,	Feuerpfeil	P,	Hanelt	D,	Jacobsen	S,	Karez	R,	et	al.	2008.	The	genus	Laminaria	sensu	lato :	recent	insights	and	developments.	European	Journal	of	Phycology	43:	1–86.	 145 Bass	J,	Dabney	A,	Robinson	D.	2018.	qvalue:	Q-value	estimation	for	false	discovery	rate	control.	R	package	version	2.14.0,	http://github.com/jdstorey/qvalue.	Baucom	RS,	Mauricio	R.	2008.	Constraints	on	the	evolution	of	tolerance	to	herbicide	in	the	common	morning	glory:	resistance	and	tolerance	are	mutually	exclusive.	Evolution;	International	Journal	of	Organic	Evolution	62:	2842–2854.	Bekkby	T,	Rinde	E,	Erikstad	L,	Bakkestuen	V.	2009.	Spatial	predictive	distribution	modelling	of	the	kelp	species	Laminaria	hyperborea.	ICES	Journal	of	Marine	Science	66:	2106–2115.	Benjamini	Y,	Hochberg	Y.	1995.	Controlling	the	False	Discovery	Rate:	A	Practical	and	Powerful	Approach	to	Multiple	Testing.	Journal	of	the	Royal	Statistical	Society.	Series	B	(Methodological)	57:	289–300.	Bennett	S,	Wernberg	T,	Connell	SD,	Hobday	AJ,	Johnson	CR,	Poloczanska	ES.	2016.	The	‘Great	Southern	Reef’:	social,	ecological	and	economic	value	of	Australia’s	neglected	kelp	forests.	Marine	and	Freshwater	Research	67:	47.	Best	RJ,	Caulk	NC,	Stachowicz	JJ.	2013.	Trait	vs.	phylogenetic	diversity	as	predictors	of	competition	and	community	composition	in	herbivorous	marine	amphipods.	Ecology	Letters	16:	72–80.	Best	RJ,	Stachowicz	JJ.	2013.	Phylogeny	as	a	proxy	for	ecology	in	seagrass	amphipods:	which	traits	are	most	conserved?	PLOS	ONE	8:	e57550.	de	Bettignies	T,	Wernberg	T,	Lavery	PS.	2013.	Size,	not	morphology,	determines	hydrodynamic	performance	of	a	kelp	during	peak	flow.	Marine	Biology	160:	843–851.	Blanchette	CA,	Miner	BG,	Gaines	SD.	2002.	Geographic	variability	in	form,	size	and	survival	of	Egregia	menziesii	around	Point	Conception,	California.	Marine	Ecology	Progress	Series	239:	69–82.	Blomberg	SP,	Garland	T,	Ives	AR.	2003.	Testing	for	phylogenetic	signal	in	comparative	data:	behavioral	traits	are	more	labile.	Evolution	57:	717–745.	Blonder	B,	Baldwin	BG,	Enquist	BJ,	Robichaux	RH.	2016.	Variation	and	macroevolution	in	leaf	functional	traits	in	the	Hawaiian	silversword	alliance	(Asteraceae).	Journal	of	Ecology	104:	219–228.	Blonder	B,	Lamanna	C,	Violle	C,	Enquist	BJ.	2014.	The	n-dimensional	hypervolume.	Global	Ecology	and	Biogeography	23:	595–609.	Böhning-Gaese	K,	Oberrath	R.	1999.	Phylogenetic	effects	on	morphological,	life-history,	behavioural	and	ecological	traits	of	birds.	Evolutionary	Ecology	Research	1:	347–364.	 146 Boizard	SMDSV.	2007.	Boizard,	S.M.D.S.V.	2007.	The	ecology	and	anchorage	mechanics	of	kelp	holdfast.	PhD	dissertation.	The	University	of	British	Columbia,	Vancouver.	193	pp.	Boller	ML,	Carrington	E.	2006.	The	hydrodynamic	effects	of	shape	and	size	change	during	reconfiguration	of	a	flexible	macroalga.	Journal	of	Experimental	Biology	209:	1894–1903.	Boller	ML,	Carrington	E.	2007.	Interspecific	comparison	of	hydrodynamic	performance	and	structural	properties	among	intertidal	macroalgae.	Journal	of	Experimental	Biology	210:	1874–1884.	Bolton	JJ.	2010.	The	biogeography	of	kelps	(Laminariales,	Phaeophyceae):	a	global	analysis	with	new	insights	from	recent	advances	in	molecular	phylogenetics.	Helgoland	Marine	Research	64:	263–279.	Boo	SM,	Lee	WJ,	Yoon	HS,	Kato	A,	Kawai	H.	1999.	Molecular	phylogeny	of	Laminariales	(Phaeophyceae)	inferred	from	small	subunit	ribosomal	DNA	sequences.	Phycological	Research	47:	109–114.	Braje	TJ,	Dillehay	TD,	Erlandson	JM,	Klein	RG,	Rick	TC.	2017.	Finding	the	first	Americans.	Science	358:	592–594.	Brown	JH,	Marquet	PA,	Taper	ML.	1993.	Evolution	of	body	size:	consequences	of	an	energetic	definition	of	fitness.	The	American	Naturalist	142:	573–584.	Buck	BH,	Buchholz	CM.	2005.	Response	of	offshore	cultivated	Laminaria	saccharina	to	hydrodynamic	forcing	in	the	North	Sea.	Aquaculture	3–4:	674–691.	Burns	JH,	Strauss	SY.	2011.	More	closely	related	species	are	more	ecologically	similar	in	an	experimental	test.	Proceedings	of	the	National	Academy	of	Sciences	108:	5302–5307.	Butler	DW,	Gleason	SM,	Davidson	I,	Onoda	Y,	Westoby	M.	2012.	Safety	and	streamlining	of	woody	shoots	in	wind:	an	empirical	study	across	39	species	in	tropical	Australia.	New	Phytologist	193:	137–149.	Cahill	JF,	Kembel	SW,	Lamb	EG,	Keddy	PA.	2008.	Does	phylogenetic	relatedness	influence	the	strength	of	competition	among	vascular	plants?	Perspectives	in	Plant	Ecology,	Evolution	and	Systematics	10:	41–50.	Cánovas	FG,	Mota	CF,	Serrão	EA,	Pearson	GA.	2011.	Driving	south:	a	multi-gene	phylogeny	of	the	brown	algal	family	Fucaceae	reveals	relationships	and	recent	drivers	of	a	marine	radiation.	BMC	Evolutionary	Biology	11:	371.	Carr	GD,	Baldwin	BG,	Carlquist	SJ.	2003.	Tarweeds	&	silverswords:	evolution	of	the	Madiinae	(Asteraceae).	St.	Louis:	Missouri	Botanical	Garden	Press.	 147 Carrington	E.	1990.	Drag	and	dislodgment	of	an	intertidal	macroalga:	consequences	of	morphological	variation	in	Mastocarpus	papillatus	Kützing.	Journal	of	Experimental	Marine	Biology	and	Ecology	139:	185–200.	Carrington	E.	2013.	High-endurance	algae.	Nature	503:	345–346.	Cavalheri	H,	Both	C,	Martins	M.	2015.	The	interplay	between	environmental	filtering	and	spatial	processes	in	structuring	communities:	the	case	of	Neotropical	snake	communities.	PLOS	ONE	10:	e0127959.	Cavalier-Smith	T.	1998.	A	revised	six-kingdom	system	of	life.	Biological	Reviews	73:	203–266.	Cavalier-Smith	T.	2018.	Kingdom	Chromista	and	its	eight	phyla:	a	new	synthesis	emphasising	periplastid	protein	targeting,	cytoskeletal	and	periplastid	evolution,	and	ancient	divergences.	Protoplasma	255:	297–357.	Cavender-Bares	J,	Ackerly	DD,	Baum	DA,	Bazzaz	FA.	2004a.	Phylogenetic	overdispersion	in	Floridian	oak	communities.	The	American	Naturalist	163:	823–843.	Cavender-Bares	J,	Kitajima	K,	Bazzaz	FA.	2004b.	Multiple	trait	associations	in	relation	to	habitat	differentiation	among	17	Floridian	oak	species.	Ecological	Monographs	74:	635–662.	Cavender-Bares	J,	Kozak	KH,	Fine	PVA,	Kembel	SW.	2009.	The	merging	of	community	ecology	and	phylogenetic	biology.	Ecology	Letters	12:	693–715.	Charrier	B,	Le	Bail	A,	de	Reviers	B.	2012.	Plant	proteus:	brown	algal	morphological	plasticity	and	underlying	developmental	mechanisms.	Trends	in	Plant	Science	17:	468–477.	Cheng	D-L,	Wang	G-X,	Li	T,	Tang	Q-L,	Gong	C-M.	2007.	Relationships	among	the	stem,	aboveground	and	total	biomass	across	Chinese	forests.	Journal	of	Integrative	Plant	Biology	49:	1573–1579.	Cho	GY,	Klochkova	NG,	Krupnova	TN,	Boo	SM.	2006.	The	reclassification	of	Lessonia	laminarioides	(Laminariales,	Phaeophyceae):	Pseudolessonia	gen.	nov.	Journal	of	Phycology	42:	1289–1299.	Cock	JM,	Sterck	L,	Rouzé	P,	Scornet	D,	Allen	AE,	Amoutzias	G,	Anthouard	V,	Artiguenave	F,	Aury	J-M,	Badger	JH,	et	al.	2010.	The	Ectocarpus	genome	and	the	independent	evolution	of	multicellularity	in	brown	algae.	Nature	465:	617–621.	Cooper	N,	Rodríguez	J,	Purvis	A.	2008.	A	common	tendency	for	phylogenetic	overdispersion	in	mammalian	assemblages.	Proceedings	of	the	Royal	Society	of	London	B:	Biological	Sciences	275:	2031–2037.	 148 Cowman	PF,	Bellwood	DR.	2013.	Vicariance	across	major	marine	biogeographic	barriers:	temporal	concordance	and	the	relative	intensity	of	hard	versus	soft	barriers.	Proceedings	of	the	Royal	Society	B:	Biological	Sciences	280.	Cramer	BS,	Toggweiler	J.	R.,	Wright	J.	D.,	Katz	M.	E.,	Miller	K.	G.	2009.	Ocean	overturning	since	the	Late	Cretaceous:	inferences	from	a	new	benthic	foraminiferal	isotope	compilation.	Paleoceanography	24.	Crisp	MD,	Cook	LG.	2005.	Do	early	branching	lineages	signify	ancestral	traits?	Trends	in	Ecology	&	Evolution	20:	122–128.	Dayton	PK.	1973.	Dispersion,	dispersal,	and	persistence	of	the	annual	intertidal	alga,	Postelsia	palmaeformis	Ruprecht.	Ecology	54:	433–438.	Dayton	PK.	1985.	Ecology	of	kelp	communities.	Annual	Review	of	Ecology	and	Systematics	16:	215–245.	Demes	KW,	Carrington	E,	Gosline	J,	Martone	PT.	2011.	Variation	in	anatomical	and	material	properties	explains	differences	in	hydrodynamic	performances	of	foliose	red	macroalgae	(Rhodophyta).	Journal	of	Phycology	47:	1360–1367.	Demes	KW,	Pruitt	JN,	Harley	CDG,	Carrington	E.	2013.	Survival	of	the	weakest:	increased	frond	mechanical	strength	in	a	wave-swept	kelp	inhibits	self-pruning	and	increases	whole-plant	mortality.	Functional	Ecology	27:	439–445.	Denny	M.	1995.	Predicting	Physical	Disturbance:	Mechanistic	Approaches	to	the	Study	of	Survivorship	on	Wave-Swept	Shores.	Ecological	Monographs	65:	371–418.	Denny	MW.	2006.	Ocean	waves,	nearshore	ecology,	and	natural	selection.	Aquatic	Ecology	40:	439–461.	Denny	M,	Gaylord	B.	2002.	The	mechanics	of	wave-swept	algae.	Journal	of	Experimental	Biology	205:	1355–1362.	Denny	M,	Gaylord	B,	Cowen	B.	1997.	Flow	and	flexibility.	II.	The	roles	of	size	and	shape	in	determining	wave	forces	on	the	bull	kelp	Nereocystis	luetkeana.	Journal	of	Experimental	Biology	200:	3165–3183.	Denny	MW,	Miller	LP,	Stokes	MD,	Hunt	LJH,	Helmuth	BST.	2003a.	Extreme	water	velocities:	Topographical	amplification	of	wave-induced	flow	in	the	surf	zone	of	rocky	shores.	Limnology	and	Oceanography	48:	1–8.	Denny	MW,	Miller	LP,	Stokes	MD,	Hunt	LJH,	Helmuth	BST.	2003b.	Extreme	water	velocities:	Topographical	amplification	of	wave-induced	flow	in	the	surf	zone	of	rocky	shores.	Limnology	and	Oceanography	48:	1–8.	 149 DeWreede	RE.	1986.	Demographic	characteristics	of	Pterygophora	californica	(Laminariales,	Phaeophyta).	Phycologia	25:	11–17.	tom	Dieck	I.	1993.	Temperature	tolerance	and	survival	in	darkness	of	kelp	gametophytes	(Laminariales,	Phaeophyta):	ecological	and	biogeographical	implications.	Marine	Ecology	Progress	Series	100:	253–264.	Domning	DP.	1976.	An	ecological	model	for	Late	Tertiary	Sirenian	evolution	in	the	North	Pacific	Ocean.	Systematic	Biology	25:	352–362.	Domning	DP.	1989.	Kelp	evolution:	a	comment.	Paleobiology	15:	53–56.	Doyle	JJ.	1987.	A	rapid	DNA	isolation	procedure	for	small	quantities	of	fresh	leaf	tissue.	Phytochem	Bull	Bot	Soc	Am	19:	11–15.	Drobnitch	ST,	Jensen	KH,	Prentice	P,	Pittermann	J.	2015.	Convergent	evolution	of	vascular	optimization	in	kelp	(Laminariales).	Proc.	R.	Soc.	B	282:	20151667.	Druehl	L,	Clarkston	B.	2016.	Pacific	Seaweeds.	Harbour	Publishing.	Druehl	LD,	Elliot	CTJ.	1996.	Parks	Canada	Barkley	Sound	Kelp	Distribution.	Parks	Canada	Contract	#K3489-3-0006.	Drummond	AJ,	Suchard	MA,	Xie	D,	Rambaut	A.	2012.	Bayesian	phylogenetics	with	BEAUti	and	the	BEAST	1.7.	Molecular	Biology	and	Evolution	29:	1969–1973.	Dubois	A,	Iken	K.	2012.	Seasonal	variation	in	kelp	phlorotannins	in	relation	to	grazer	abundance	and	environmental	variables	in	the	Alaskan	sublittoral	zone.	Duggins	DO,	Eckman	JE,	Siddon	CE,	Klinger	T.	2003.	Population,	morphometric	and	biomechanical	studies	of	three	understory	kelps	along	a	hydrodynamic	gradient.	Marine	Ecology	Progress	Series	265:	57–76.	Duggins	DO,	Simenstad	CA,	Estes	JA.	1989.	Magnification	of	secondary	production	by	kelp	detritus	in	coastal	marine	ecosystems.	Science	245:	170–173.	Ennos	AR.	1997.	Wind	as	an	ecological	factor.	Trends	in	Ecology	&	Evolution	12:	108–111.	Enquist	BJ.	2002.	Universal	scaling	in	tree	and	vascular	plant	allometry:	toward	a	general	quantitative	theory	linking	plant	form	and	function	from	cells	to	ecosystems.	Tree	Physiology	22:	1045–1064.	Enquist	BJ,	Allen	AP,	Brown	JH,	Gillooly	JF,	Kerkhoff	AJ,	Niklas	KJ,	Price	CA,	West	GB.	2007.	Biological	scaling:	does	the	exception	prove	the	rule?	Nature	445:	E9-10;	discussion	E10-11.	 150 Enquist	BJ,	Niklas	KJ.	2002a.	Response	to	Sack	et	al.	Global	allocation	rules	for	patterns	in	partioning.	Science	296:	1923a.	Enquist	BJ,	Niklas	KJ.	2002b.	Global	allocation	rules	for	patterns	of	biomass	partitioning	in	seed	plants.	Science	295:	1517–1520.	Enquist	BJ,	Norberg	J,	Bonser	SP,	Violle	C,	Webb	CT,	Henderson	A,	Sloat	LL,	Savage	VM.	2015.	Scaling	from	traits	to	ecosystems:	Developing	a	general	Trait	Driver	Theory	via	integrating	trait-based	and	metabolic	scaling	theories.	arXiv:1502.06629	[q-bio].	Erlandson	JM,	Graham	MH,	Bourque	BJ,	Corbett	D,	Estes	JA,	Steneck	RS.	2007.	The	Kelp	Highway	Hypothesis:	marine	ecology,	the	Coastal	Migration	Theory,	and	the	peopling	of	the	Americas.	The	Journal	of	Island	and	Coastal	Archaeology	2:	161–174.	Erwin	DH.	1998.	The	end	and	the	beginning:	recoveries	from	mass	extinctions.	Trends	in	Ecology	&	Evolution	13:	344–349.	Estes	JA,	Burdin	A,	Doak	DF.	2016.	Sea	otters,	kelp	forests,	and	the	extinction	of	Steller’s	sea	cow.	Proceedings	of	the	National	Academy	of	Sciences	113:	880–885.	Estes	JA,	Duggins	DO.	1995a.	Sea	otters	and	kelp	forests	in	Alaska:	generality	and	variation	in	a	community	ecological	paradigm.	Ecological	Monographs	65:	75–100.	Estes	JA,	Duggins	DO.	1995b.	Sea	otters	and	kelp	forests	in	Alaska:	generality	and	variation	in	a	community	ecological	paradigm.	Ecological	Monographs	65:	75–100.	Estes	JA,	Steinberg	PD.	1988.	Predation,	herbivory,	and	kelp	evolution.	Paleobiology	14:	19–36.	Estes	JA,	Steinberg	PD.	1989.	Response	to	Domning.	Paleobiology	15:	57–60.	Estes	JA,	Terborgh	J,	Brashares	JS,	Power	ME,	Berger	J,	Bond	WJ,	Carpenter	SR,	Essington	TE,	Holt	RD,	Jackson	JBC,	et	al.	2011.	Trophic	downgrading	of	planet	Earth.	Science	333:	301–306.	Falster	DS,	Westoby	M.	2003.	Plant	height	and	evolutionary	games.	Trends	in	Ecology	&	Evolution	18:	337–343.	Felsenstein	J.	1985.	Phylogenies	and	the	comparative	Method.	The	American	Naturalist	125:	1–15.	Filbee-Dexter	K,	Wernberg	T.	2018.	Rise	of	turfs:	a	new	battlefront	for	globally	declining	kelp	forests.	BioScience	68:	64–76.	Fineblum	WL,	Rausher	MD.	1995.	Tradeoff	between	resistance	and	tolerance	to	herbivore	damage	in	a	morning	glory.	Nature	377:	517–520.	 151 Fitzhenry	T,	Halpin	PM,	Helmuth	B.	2004.	Testing	the	effects	of	wave	exposure,	site,	and	behavior	on	intertidal	mussel	body	temperatures:	applications	and	limits	of	temperature	logger	design.	Marine	Biology	145:	339–349.	Fornoni	J,	Valverde	*	Pedro	L.,	Núñez-Farfán	J.	2003.	Quantitative	genetics	of	plant	tolerance	and	resistance	against	natural	enemies	of	two	natural	populations	of	Datura	stramonium.	Evolutionary	Ecology	Research	5:	1049–1065.	Fortes	MD,	Lüning	K.	1980.	Growth	rates	of	North	Sea	macroalgae	in	relation	to	temperature,	irradiance	and	photoperiod.	Helgoländer	Meeresuntersuchungen	34:	15–29.	Fowler-Walker	MJ,	Wernberg	T,	Connell	SD.	2006.	Differences	in	kelp	morphology	between	wave	sheltered	and	exposed	localities:	morphologically	plastic	or	fixed	traits?	Marine	Biology	148:	755–767.	Fredriksen	S.	2003.	Food	web	studies	in	a	Norwegian	kelp	forest	based	on	stable	isotope	(δ13C	and	δ15N)	analysis.	Marine	Ecology	Progress	Series	260:	71–81.	Friedland	MT,	Denny	MW.	1995.	Surviving	hydrodynamic	forces	in	a	wave-swept	environment:	Consequences	of	morphology	in	the	feather	boa	kelp,	Egregia	menziesii	(Turner).	Journal	of	Experimental	Marine	Biology	and	Ecology	190:	109–133.	Frimpong	EA,	Angermeier	PL.	2009.	Fish	Traits:	a	database	of	ecological	and	life-history	traits	of	freshwater	fishes	of	the	United	States.	Fisheries	34:	487–495.	Frimpong	EA,	Angermeier	PL.	2010.	Trait-based	approaches	in	the	analysis	of	stream	fish	communities.	In:	Community	ecology	of	stream	fishes:	concepts,	approaches,	and	techniques.	American	Fisheries	Society,	Symposium.	109–136.	Fritsch	FE.	1935.	The	structure	and	reproduction	of	algae.	Cambridge:	Cambridge	University	Press.	Gaylord	B,	Blanchette	CA,	Denny	MW.	1994.	Mechanical	consequences	of	size	in	wave-swept	algae.	Ecological	Monographs	64:	287–313.	Gaylord	B,	Denny	M.	1997.	Flow	and	flexibility.	I.	Effects	of	size,	shape	and	stiffness	in	determining	wave	forces	on	the	stipitate	kelps	Eisenia	arborea	and	Pterygophora	californica.	Journal	of	Experimental	Biology	200:	3141–3164.	Gaylord	B,	Denny	MW,	Koehl	M	a.	R.	2008.	Flow	forces	on	seaweeds:	field	evidence	for	roles	of	wave	impingement	and	organism	inertia.	The	Biological	Bulletin	215:	295–308.	Gerard	VA.	1988.	Ecotypic	differentiation	in	light-related	traits	of	the	kelp	Laminaria	saccharina.	Marine	Biology	97:	25–36.	 152 Gerard	VA,	Mann	KH.	1979.	Growth	and	production	of	Laminaria	longicruris	(Phaeophyta)	populations	exposed	to	different	intensities	of	water	movement.	Journal	of	Phycology	15:	33–41.	Gerhold	P,	Cahill	JF,	Winter	M,	Bartish	IV,	Prinzing	A.	2015.	Phylogenetic	patterns	are	not	proxies	of	community	assembly	mechanisms	(they	are	far	better).	Functional	Ecology	29:	600–614.	Givnish	TJ.	2015.	Adaptive	radiation	versus	‘radiation’	and	‘explosive	diversification’:	why	conceptual	distinctions	are	fundamental	to	understanding	evolution.	New	Phytologist	207:	297–303.	Goldner	A,	Herold	N,	Huber	M.	2014.	Antarctic	glaciation	caused	ocean	circulation	changes	at	the	Eocene–Oligocene	transition.	Nature	511:	574–577.	Gould	SJ.	1966.	Allometry	and	size	in	ontogeny	and	phylogeny.	Biological	Reviews	of	the	Cambridge	Philosophical	Society	41:	587–640.	Graham	MH.	2004.	Effects	of	local	deforestation	on	the	diversity	and	structure	of	southern	California	giant	kelp	forest	food	webs.	Ecosystems	7:	341–357.	Graham	JE,	Graham	LE,	Wilcox	LW.	2008.	Algae.	Benjamin	Cummings.	Graham	MH,	Kinlan	BP,	Druehl	LD,	Garske	LE,	Banks	S.	2007.	Deep-water	kelp	refugia	as	potential	hotspots	of	tropical	marine	diversity	and	productivity.	Proceedings	of	the	National	Academy	of	Sciences	104:	16576–16580.	Griffith	DM,	Veech	JA,	Marsh	CJ.	2016.	cooccur:	probabilistic	species	co-occurrence	analysis	in	R.	Journal	of	Statistical	Software	69:	1–17.	Guiry	MD,	Guiry	GM.	AlgaeBase.	World-wide	electronic	publication,	National	University	of	Ireland,	Galway.	http://www.algaebase.org.	Hanken	J.	1993.	Miniaturization	of	body	size:	organismal	consequences	and	evolutionary	significance.	Annual	Review	of	Ecology	and	Systematics	24:	501–519.	Harder	DL,	Hurd	CL,	Speck	T.	2006.	Comparison	of	mechanical	properties	of	four	large,	wave-exposed	seaweeds.	American	Journal	of	Botany	93:	1426–1432.	Hardin	G.	1960.	The	competitive	exclusion	principle.	Science	131:	1292–1297.	Harley	CDG.	2003.	Abiotic	stress	and	herbivory	interact	to	set	range	limits	across	a	two-dimensional	stress	gradient.	Ecology	84:	1477–1488.	 153 Harley	CDG,	Helmuth	BST.	2003.	Local-	and	regional-scale	effects	of	wave	exposure,	thermal	stress,	and	absolute	versus	effective	shore	level	on	patterns	of	intertidal	zonation.	Limnology	and	Oceanography	48:	1498–1508.	Henley	WJ,	Dunton	KH.	1997.	Effects	of	nitrogen	supply	and	continuous	darkness	on	growth	and	photosynthesis	of	the	arctic	kelp	Laminaria	solidungula.	Limnology	and	Oceanography	42:	209–216.	Henry	HAL,	Aarssen	LW.	1997.	On	the	relationship	between	shade	tolerance	and	shade	avoidance	strategies	in	woodland	plants.	Oikos	80:	575–582.	Hoek	C,	Hoeck	HV	den,	Mann	D,	Jahns	HM.	1995.	Algae:	An	Introduction	to	Phycology.	Cambridge	University	Press.	Hughes	BB.	2010.	Variable	effects	of	a	kelp	foundation	species	on	rocky	intertidal	diversity	and	species	interactions	in	central	California.	Journal	of	Experimental	Marine	Biology	and	Ecology	393:	90–99.	Hunter	MD,	Price	PW.	1992.	Playing	chutes	and	ladders:	heterogeneity	and	the	relative	roles	of	bottom-up	and	top-down	forces	in	natural	communities.	Ecology	73:	724–732.	Hurd	CL.	2000.	Water	motion,	marine	macroalgal	physiology,	and	production.	Journal	of	Phycology	36:	453–472.	Hurd	CL.	2017.	Shaken	and	stirred:	the	fundamental	role	of	water	motion	in	resource	acquisition	and	seaweed	productivity.	Perspectives	in	Phycology:	73–81.	Hurd	CL,	Harrison	PJ,	Bischof	K,	Lobban	CS.	2014.	Seaweed	ecology	and	physiology.	Cambridge	University	Press.	Huxley	J.	1932.	Problems	of	relative	growth.	New	York,	NY,	USA:	Dial	Press.	Iason	GR,	Villalba	JJ.	2006.	Behavioral	strategies	of	mammal	herbivores	against	plant	secondary	metabolites:	the	avoidance–tolerance	continuum.	Journal	of	Chemical	Ecology	32:	1115–1132.	Ingram	T.	2011.	Speciation	along	a	depth	gradient	in	a	marine	adaptive	radiation.	Proceedings	of	the	Royal	Society	B:	Biological	Sciences	278:	613–618.	Ishii	R,	Crawley	MJ.	2011.	Herbivore-induced	coexistence	of	competing	plant	species.	Journal	of	Theoretical	Biology	268:	50–61.	Ivany	LC,	Patterson	WP,	Lohmann	KC.	2000.	Cooler	winters	as	a	possible	cause	of	mass	extinctions	at	the	Eocene-Oligocene	boundary.	Nature	407:	887–890.	 154 Jackson	C,	Salomaki	ED,	Lane	CE,	Saunders	GW.	2017.	Kelp	transcriptomes	provide	robust	support	for	interfamilial	relationships	and	revision	of	the	little	known	Arthrothamnaceae	(Laminariales).	Journal	of	Phycology	53:	1–6.	Janot	K,	Martone	PT.	2016.	Convergence	of	joint	mechanics	in	independently	evolving,	articulated	coralline	algae.	Journal	of	Experimental	Biology	219:	383–391.	Jensen	MM,	Denny	MW.	2015.	Experimental	determination	of	the	hydrodynamic	forces	responsible	for	wave	impact	events.	Journal	of	Experimental	Marine	Biology	and	Ecology	469:	123–130.	Johnson	A,	Koehl	M.	1994.	Maintenance	of	dynamic	strain	similarity	and	environmental	stress	factor	in	different	flow	habitats:	thallus	allometry	and	material	properties	of	a	giant	kelp.	The	Journal	of	Experimental	Biology	195:	381–410.	Joly	S,	McLenachan	PA,	Lockhart	PJ.	2009.	A	statistical	approach	for	distinguishing	hybridization	and	incomplete	lineage	sorting.	The	American	Naturalist	174:	E54–E70.	Kain	JM,	Norton	TA,	Montegut	AE.	1987.	Growth	of	blades	of	Nereocystis	luetkeana	(Phaeophyta)	in	Darkness.	Journal	of	Phycology	23:	464–469.	Katoh	K,	Misawa	K,	Kuma	K,	Miyata	T.	2002.	MAFFT:	a	novel	method	for	rapid	multiple	sequence	alignment	based	on	fast	Fourier	transform.	Nucleic	Acids	Research	30:	3059–3066.	Kawai	H,	Hanyuda	T,	Ridgway	LM,	Holser	K.	2013.	Ancestral	reproductive	structure	in	basal	kelp	Aureophycus	aleuticus.	Scientific	Reports	3:	2491.	Kawai	H,	Hanyuda	Takeaki,	Gao	Xu,	Terauchi	Makoto,	Miyata	Masahiko,	Lindstrom	Sandra	C.,	Klochkova	Nina	G.,	Miller	Kathy	Ann,	Lane	C.	2017.	Taxonomic	revision	of	the	Agaraceae	with	a	description	of	Neoagarum	gen.	nov.	and	reinstatement	of	Thalassiophyllum.	Journal	of	Phycology	53:	261–270.	Kawai	H,	Hanyuda	T,	Uwai	S.	2016.	Evolution	and	biogeography	of	Laminarialean	kelps.	In:	Hu	Z-M,	Fraser	C,	eds.	Seaweed	Phylogeography.	Dordrecht:	Springer	Netherlands,	227–249.	Kawamata	S.	2001.	Adaptive	mechanical	tolerance	and	dislodgement	velocity	of	the	kelp	Laminaria	japonica	in	wave-induced	water	motion.	Marine	Ecology	Progress	Series	211:	89–104.	Kearse	M,	Moir	R,	Wilson	A,	Stones-Havas	S,	Cheung	M,	Sturrock	S,	Buxton	S,	Cooper	A,	Markowitz	S,	Duran	C,	et	al.	2012.	Geneious	Basic:	An	integrated	and	extendable	desktop	software	platform	for	the	organization	and	analysis	of	sequence	data.	Bioinformatics	28:	1647–1649.	Keeling	PJ.	2004.	Diversity	and	evolutionary	history	of	plastids	and	their	hosts.	American	Journal	of	Botany	91:	1481–1493.	 155 Keeling	PJ.	2010.	The	endosymbiotic	origin,	diversification	and	fate	of	plastids.	Philosophical	Transactions	of	the	Royal	Society	B:	Biological	Sciences	365:	729–748.	Kembel	SW,	Cowan	PD,	Helmus	MR,	Cornwell	WK,	Morlon	H,	Ackerly	DD,	Blomberg	SP,	Webb	CO.	2010.	Picante:	R	tools	for	integrating	phylogenies	and	ecology.	Bioinformatics	(Oxford,	England)	26:	1463–1464.	Kishino	H,	Hasegawa	M.	1989.	Evaluation	of	the	maximum	likelihood	estimate	of	the	evolutionary	tree	topologies	from	DNA	sequence	data,	and	the	branching	order	in	Hominoidea.	Journal	of	Molecular	Evolution	29:	170–179.	Kitzes	JA,	Denny	MW.	2005.	Red	algae	respond	to	waves:	morphological	and	mechanical	variation	in	<i>Mastocarpus	papillatus</I>	along	a	gradient	of	force.	The	Biological	Bulletin	208:	114–119.	Kleiber	M.	1932.	Body	size	and	metabolism.	Hilgardia:	315–351.	Klinger	T,	DeWreede	RE.	1988.	Stipe	rings,	age,	and	size	in	populations	of	Laminaria	setchellii	Silva	(Laminariales,	Phaeophyta)	in	British	Columbia,	Canada.	Phycologia	27:	234–240.	Klochkova	TA,	Klochkova	NG,	Yotsukura	N,	Kim	GH.	2017.	Morphological,	molecular,	and	chromosomal	identification	of	dwarf	haploid	parthenosporophytes	of	Tauya	basicrassa	(Phaeophyceae,	Laminariales)	from	the	Sea	of	Okhotsk.	ALGAE	32:	15–28.	Kober	KM,	Bernardi	G.	2013.	Phylogenomics	of	strongylocentrotid	sea	urchins.	BMC	Evolutionary	Biology	13:	88.	Koehl	M	a.	R.	1982.	Mechanical	design	of	spicule-reinforced	connective	tissue:	stiffness.	Journal	of	Experimental	Biology	98:	239–267.	Koehl	M	a.	R.	1984.	How	do	benthic	organisms	withstand	moving	water?	Integrative	and	Comparative	Biology	24:	57–70.	Koehl	MA.	1999.	Ecological	biomechanics	of	benthic	organisms:	life	history,	mechanical	design	and	temporal	patterns	of	mechanical	stress.	Journal	of	Experimental	Biology	202:	3469–3476.	Koehl	M	a.	R,	Silk	WK,	Liang	H,	Mahadevan	L.	2008a.	How	kelp	produce	blade	shapes	suited	to	different	flow	regimes:	a	new	wrinkle.	Integrative	and	Comparative	Biology	48:	834–851.	Koehl	MAR,	Silk	WK,	Liang	H,	Mahadevan	L.	2008b.	How	kelp	produce	blade	shapes	suited	to	different	flow	regimes:	a	new	wrinkle.	Integrative	and	Comparative	Biology	48:	834–851.	Koehl	MAR,	Wainwright	SA.	1977.	Mechanical	adaptations	of	a	giant	kelp.	Limnology	and	Oceanography	22:	1067–1071.	 156 Koontz	TL,	Petroff	A,	West	GB,	Brown	JH.	2009.	Scaling	relations	for	a	functionally	two-dimensional	plant:	Chamaesyce	setiloba	(Euphorbiaceae).	American	Journal	of	Botany	96:	877–884.	Kraft	NJB,	Adler	PB,	Godoy	O,	James	EC,	Fuller	S,	Levine	JM.	2014.	Community	assembly,	coexistence	and	the	environmental	filtering	metaphor.	Functional	Ecology	29:	592–599.	Kraft	NJ,	Comita	LS,	Chase	JM,	Sanders	NJ,	Swenson	NG,	Crist	TO,	Stegen	JC,	Vellend	M,	Boyle	B,	Anderson	MJ.	2011.	Disentangling	the	drivers	of	β	diversity	along	latitudinal	and	elevational	gradients.	Science	333:	1755–1758.	Kraft	NJB,	Cornwell	WK,	Webb	CO,	Ackerly	DD.	2007.	Trait	evolution,	community	assembly,	and	the	phylogenetic	structure	of	ecological	communities.	The	American	Naturalist	170:	271–283.	Kraft	NJB,	Valencia	R,	Ackerly	DD.	2008.	Functional	traits	and	niche-based	tree	community	assembly	in	an	Amazonian	forest.	Science	322:	580–582.	Kremen	C.	2005.	Managing	ecosystem	services:	what	do	we	need	to	know	about	their	ecology?	Ecology	letters	8:	468–479.	LaBarbera	M.	1989.	Analyzing	body	size	as	a	factor	in	ecology	and	evolution.	Annual	Review	of	Ecology	and	Systematics	20:	97–117.	Lago-Lestón	A,	Mota	C,	Kautsky	L,	Pearson	GA.	2010.	Functional	divergence	in	heat	shock	response	following	rapid	speciation	of	Fucus	spp.	in	the	Baltic	Sea.	Marine	Biology	157:	683–688.	Landis	MJ,	Matzke	NJ,	Moore	BR,	Huelsenbeck	JP.	2013.	Bayesian	analysis	of	biogeography	when	the	number	of	areas	is	large.	Systematic	biology	62:	789–804.	Lane	CE,	Lindstrom	SC,	Saunders	GW.	2007.	A	molecular	assessment	of	northeast	Pacific	Alaria	species	(Laminariales,	Phaeophyceae)	with	reference	to	the	utility	of	DNA	barcoding.	Molecular	Phylogenetics	and	Evolution	44:	634–648.	Lane	CE,	Mayes	C,	Druehl	LD,	Saunders	GW.	2006.	A	multi-gene	molecular	investigation	of	the	kelp	(Laminariales,	Phaeophyceae)	supports	substantial	taxonomic	re-organization.	Journal	of	Phycology	42:	493–512.	Lanfear	R,	Frandsen	PB,	Wright	AM,	Senfeld	T,	Calcott	B.	2017.	PartitionFinder	2:	New	methods	for	selecting	partitioned	models	of	evolution	for	molecular	and	morphological	phylogenetic	analyses.	Molecular	Biology	and	Evolution	34:	772–773.	Legendre	P.	2014.	lmodel2:	Model	II	Regression.	R	package	version	1.7-2.	Available	at:	ht	tp://CRAN.	R-project.	org/package=	lmodel2	(accessed	2	March	2015).	 157 Leigh	EG,	Paine	RT,	Quinn	JF,	Suchanek	TH.	1987.	Wave	energy	and	intertidal	productivity.	Proceedings	of	the	National	Academy	of	Sciences	84:	1314–1318.	Levin	SA.	1970.	Community	equilibria	and	stability,	and	an	extension	of	the	competitive	exclusion	principle.	The	American	Naturalist	104:	413–423.	Liggan	LM,	Martone	PT.	2018.	Under	pressure:	biomechanical	limitations	of	developing	pneumatocysts	in	the	bull	kelp	(Nereocystis	luetkeana,	Phaeophyceae).	Journal	of	Phycology	0.	Lindberg	DR,	Marincovich	L.	1988.	New	species	of	limpets	from	the	Neogene	of	Alaska	(Patellogastropoda:	Mollusca).	ARCTIC	41.	Litchman	E,	Klausmeier	CA.	2008.	Trait-based	community	ecology	of	phytoplankton.	Annual	Review	of	Ecology,	Evolution,	and	Systematics	39:	615–639.	López-Cristoffanini	C,	Tellier	F,	Otaíza	R,	Correa	JA,	Contreras-Porcia	L.	2013.	Tolerance	to	air	exposure:	a	feature	driving	the	latitudinal	distribution	of	two	sibling	kelp	species.	Botanica	Marina	56:	431–440.	Losos	JB.	1995.	Community	evolution	in	Greater	Antillean	Anolis	lizards:	phylogenetic	patterns	and	experimental	tests.	Phil.	Trans.	R.	Soc.	Lond.	B	349:	69–75.	Losos	JB.	2008.	Phylogenetic	niche	conservatism,	phylogenetic	signal	and	the	relationship	between	phylogenetic	relatedness	and	ecological	similarity	among	species.	Ecology	Letters	11:	995–1003.	Lüning	K.	1991.	Circannual	growth	rhythm	in	a	brown	alga,	Pterygophora	californica.	Botanica	Acta	104:	157–162.	Lüning	K,	tom	Dieck	I.	1990.	The	distribution	and	evolution	of	the	Laminariales:	North	Pacific—Atlantic	relationships.	In:	Evolutionary	biogeography	of	the	marine	algae	of	the	North	Atlantic.	Springer,	187–204.	Lüning	K,	Freshwater	W.	1988.	Temperature	tolerance	of	Northeast	Pacific	marine	algae.	Journal	of	Phycology	24:	310–315.	Macarthur	R,	Levins	R.	1967.	The	limiting	similarity,	convergence,	and	divergence	of	coexisting	species.	The	American	Naturalist	101:	377–385.	Mach	KJ,	Hale	BB,	Denny	MW,	Nelson	DV.	2007.	Death	by	small	forces:	a	fracture	and	fatigue	analysis	of	wave-swept	macroalgae.	The	Journal	of	Experimental	Biology	210:	2231–2243.	Mach	KJ,	Tepler	SK,	Staaf	AV,	Bohnhoff	JC,	Denny	MW.	2011.	Failure	by	fatigue	in	the	field:	a	model	of	fatigue	breakage	for	the	macroalga	Mazzaella,	with	validation.	The	Journal	of	Experimental	Biology	214:	1571–1585.	 158 Maddison	WP,	Knowles	LL,	Collins	T.	2006.	Inferring	phylogeny	despite	incomplete	lineage	sorting.	Systematic	Biology	55:	21–30.	Mahler	DL,	Revell	LJ,	Glor	RE,	Losos	JB.	2010.	Ecological	opportunity	and	the	rate	of	morphological	evolution	in	the	diversification	of	greater	antillean	Anoles.	Evolution	64:	2731–2745.	Martin	P,	Zuccarello	GC.	2012.	Molecular	phylogeny	and	timing	of	radiation	in	Lessonia	(Phaeophyceae,	Laminariales).	Phycological	Research	60:	276–287.	Martone	PT.	2007a.	Kelp	versus	coralline:	cellular	basis	for	mechanical	strength	in	the	wave-swept	seaweed	Calliarthron	(Corallinaceae,	Rhodophyta).	Journal	of	Phycology	43:	882–891.	Martone	PT.	2007b.	Kelp	versus	coralline:	cellular	basis	for	mechanical	strength	in	the	wave-swept	seaweed	Calliarthron	(Corallinaceae,	Rhodophyta).	Journal	of	Phycology	43:	882–891.	Martone	PT,	Denny	MW.	2008.	To	break	a	coralline:	mechanical	constraints	on	the	size	and	survival	of	a	wave-swept	seaweed.	Journal	of	Experimental	Biology	211:	3433–3441.	Martone	PT,	Kost	L,	Boller	M.	2012.	Drag	reduction	in	wave-swept	macroalgae:	alternative	strategies	and	new	predictions.	American	Journal	of	Botany	99:	806–815.	Matzke	NJ.	2013.	BioGeoBEARS:	BioGeography	with	Bayesian	(and	likelihood)	evolutionary	analysis	in	R	Scripts.	R	package,	version	0.2	1:	2013.	McDevit	DC,	Saunders	GW.	2010.	A	DNA	barcode	examination	of	the	Laminariaceae	(Phaeophyceae)	in	Canada	reveals	novel	biogeographical	and	evolutionary	insights.	Phycologia;	Lawrence	49:	235–248.	McFadden	GI.	2001.	Primary	and	secondary	endosymbiosis	and	the	origin	of	plastids.	Journal	of	Phycology	37:	951–959.	Menge	BA,	Sutherland	JP.	1987.	Community	regulation:	variation	in	disturbance,	competition,	and	predation	in	relation	to	environmental	stress	and	recruitment.	The	American	Naturalist	130:	730–757.	Messier	J,	McGill	BJ,	Lechowicz	MJ.	2010.	How	do	traits	vary	across	ecological	scales?	A	case	for	trait-based	ecology.	Ecology	Letters	13:	838–848.	Michel	G,	Tonon	T,	Scornet	D,	Cock	JM,	Kloareg	B.	2010.	The	cell	wall	polysaccharide	metabolism	of	the	brown	alga	Ectocarpus	siliculosus.	Insights	into	the	evolution	of	extracellular	matrix	polysaccharides	in	Eukaryotes.	New	Phytologist	188:	82–97.	Miller	RJ,	Harrer	S,	Reed	DC.	2012.	Addition	of	species	abundance	and	performance	predicts	community	primary	production	of	macroalgae.	Oecologia	168:	797–806.	 159 Miller	RJ,	Lafferty	KD,	Lamy	T,	Kui	L,	Rassweiler	A,	Reed	DC.	2018.	Giant	kelp,	Macrocystis	pyrifera,	increases	faunal	diversity	through	physical	engineering.	Proc.	R.	Soc.	B	285:	20172571.	Miller	MA,	Pfeiffer	W,	Schwartz	T.	2011.	The	CIPRES	Science	Gateway:	A	community	resource	for	phylogenetic	analyses.	In:	TG	’11.	Proceedings	of	the	2011	TeraGrid	Conference:	Extreme	Digital	Discovery.	New	York,	NY,	USA:	ACM,	41:1–41:8.	Milligan	K,	DeWreede	R.	2000.	Variations	in	holdfast	attachment	mechanics	with	developmental	stage,	substratum-type,	season,	and	wave-exposure	for	the	intertidal	kelp	species	Hedophyllum	sessile	(C.	Agardh)	Setchell.	Journal	of	Experimental	Marine	Biology	and	Ecology	254:	189–209.	Molloy	FJ,	Bolton	JJ.	2009.	The	effects	of	wave	exposure	and	depth	on	the	morphology	of	inshore	populations	of	the	Namibian	kelp,	Laminaria	schinzii	Foslie.	Botanica	Marina	39:	525–532.	Mori	S,	Yamaji	K,	Ishida	A,	Prokushkin	SG,	Masyagina	OV,	Hagihara	A,	Hoque	ATMR,	Suwa	R,	Osawa	A,	Nishizono	T,	et	al.	2010.	Mixed-power	scaling	of	whole-plant	respiration	from	seedlings	to	giant	trees.	Proceedings	of	the	National	Academy	of	Sciences	107:	1447–1451.	Neiva	J,	Paulino	C,	Nielsen	MM,	Krause-Jensen	D,	Saunders	GW,	Assis	J,	Bárbara	I,	Tamigneaux	É,	Gouveia	L,	Aires	T,	et	al.	2018.	Glacial	vicariance	drives	phylogeographic	diversification	in	the	amphi-boreal	kelp	Saccharina	latissima.	Scientific	Reports	8:	1112.	Neufeld	CJ,	Starko	S,	Burns	K.	2017.	Disturbance	and	diversity	in	a	continental	archipelago:	A	mechanistic	framework	linking	area,	height,	and	exposure.	Ecosphere	8:	e01957.	Nielsen	KJ,	Blanchette	CA,	Menge	BA,	Lubchenco	J.	2006.	Physiological	snapshots	reflect	ecological	performance	of	the	sea	palm,	Postelsia	palmaeformis	(Phaeophyceae)	across	intertidal	elevation	and	exposure	gradients.	Journal	of	Phycology	42:	548–559.	Niklas	KJ.	1993.	Influence	of	tissue	density-specific	mechanical	properties	on	the	scaling	of	plant	height.	Annals	of	Botany	72:	173–179.	Niklas	KJ.	1994.	Plant	allometry:	the	scaling	of	form	and	process.	University	of	Chicago	Press.	Niklas	KJ.	2003.	Reexamination	of	a	canonical	model	for	plant	organ	biomass	partitioning.	American	Journal	of	Botany	90:	250–254.	Niklas	KJ.	2004.	Plant	allometry:	is	there	a	grand	unifying	theory?	Biological	Reviews	of	the	Cambridge	Philosophical	Society	79:	871–889.	Niklas	KJ.	2006.	A	phyletic	perspective	on	the	allometry	of	plant	biomass-partitioning	patterns	and	functionally	equivalent	organ-categories.	New	Phytologist	171:	27–40.	 160 Niklas	KJ,	Cobb	ED.	2008.	Evidence	for	‘diminishing	returns’	from	the	scaling	of	stem	diameter	and	specific	leaf	area.	American	Journal	of	Botany	95:	549–557.	Niklas	KJ,	Cobb	ED,	Spatz	H-C.	2009.	Predicting	the	allometry	of	leaf	surface	area	and	dry	mass.	American	Journal	of	Botany	96:	531–536.	Niklas	KJ,	Enquist	BJ.	2001.	Invariant	scaling	relationships	for	interspecific	plant	biomass	production	rates	and	body	size.	Proceedings	of	the	National	Academy	of	Sciences	98:	2922–2927.	Niklas	KJ,	Enquist	BJ.	2002.	On	the	vegetative	biomass	partitioning	of	seed	plant	leaves,	stems,	and	roots.	The	American	Naturalist	159:	482–497.	Niklas	KJ,	Enquist	BJ.	2004.	Biomass	allocation	and	growth	data	of	seeded	plants.	ORNL	DAAC.	Niklas	KJ,	Spatz	H-C.	2004.	Growth	and	hydraulic	(not	mechanical)	constraints	govern	the	scaling	of	tree	height	and	mass.	Proceedings	of	the	National	Academy	of	Sciences	101:	15661–15663.	Norton	TA.	1991.	Conflicting	constraints	on	the	form	of	intertidal	algae.	British	Phycological	Journal	26:	203–218.	Oliveira	L,	Walker	DC,	Bisalputra	T.	1980.	Ultrastructural,	cytochemical,	and	enzymatic	studies	on	the	adhesive	“plaques”	of	the	brown	algae	Laminaria	saccharina	(L.)	Lamour.	and	Nereocystis	luetkeana	(Nert.)	Post.	et	Rupr.	Protoplasma	104:	1–15.	Oudot-Le	Secq	M-P,	Kloareg	B,	Loiseaux-De	Goër	S.	2002.	The	mitochondrial	genome	of	the	brown	alga	Laminaria	digitata:	a	comparative	analysis.	European	Journal	of	Phycology	37:	163–172.	Pagel	M.	1999.	Inferring	the	historical	patterns	of	biological	evolution.	Nature	401:	877–884.	Palumbi	SR.	1994.	Genetic	divergence,	reproductive	isolation,	and	marine	speciation.	Annual	Review	of	Ecology	and	Systematics	25:	547–572.	Paradis	E,	Claude	J,	Strimmer	K.	2004.	APE:	Analyses	of	phylogenetics	and	evolution	in	R	language.	Bioinformatics	20:	289–290.	Parker	B,	Dawson	E.	1965.	Non-calcareous	marine	algae	from	California	Miocene	marine	deposits.	Nova	Hedwigia	10:	273–295.	Peters	RH,	Peters	RH.	1986.	The	Ecological	Implications	of	Body	Size.	Cambridge	University	Press.	Pfister	CA.	1992.	Costs	of	reproduction	in	an	intertidal	kelp:	patterns	of	allocation	and	life	history	consequences.	Ecology	73:	1586–1596.	 161 Pfister	CA,	Berry	HD,	Mumford	T.	2017.	The	dynamics	of	kelp	forests	in	the	Northeast	Pacific	Ocean	and	the	relationship	with	environmental	drivers.	Journal	of	Ecology	106:	1520–1533.	Phillimore	AB,	Price	TD.	2008.	Density-dependent	cladogenesis	in	birds.	PLOS	Biology	6:	e71.	Poff	NL.	1997.	Landscape	filters	and	species	traits:	towards	mechanistic	understanding	and	prediction	in	stream	ecology.	Journal	of	the	North	American	Benthological	Society	16:	391–409.	Poorter	H,	Jagodzinski	AM,	Ruiz-Peinado	R,	Kuyah	S,	Luo	Y,	Oleksyn	J,	Usoltsev	VA,	Buckley	TN,	Reich	PB,	Sack	L.	2015.	How	does	biomass	distribution	change	with	size	and	differ	among	species?	An	analysis	for	1200	plant	species	from	five	continents.	New	Phytologist	208:	736–749.	Poorter	H,	Niklas	KJ,	Reich	PB,	Oleksyn	J,	Poot	P,	Mommer	L.	2012.	Biomass	allocation	to	leaves,	stems	and	roots:	meta-analyses	of	interspecific	variation	and	environmental	control:	Tansley	review.	New	Phytologist	193:	30–50.	Power	ME.	1992.	Top-down	and	bottom-up	forces	in	food	webs:	do	plants	have	primacy.	Ecology	73:	733–746.	Price	CA,	Enquist	BJ.	2006.	Scaling	of	mass	and	morphology	in	plants	with	minimal	branching:	an	extension	of	the	WBE	model.	Functional	Ecology	20:	11–20.	Price	CA,	Enquist	BJ,	Savage	VM.	2007.	A	general	model	for	allometric	covariation	in	botanical	form	and	function.	Proceedings	of	the	National	Academy	of	Sciences	104:	13204–13209.	Puijalon	S,	Bouma	TJ,	Douady	CJ,	van	Groenendael	J,	Anten	NPR,	Martel	E,	Bornette	G.	2011.	Plant	resistance	to	mechanical	stress:	evidence	of	an	avoidance-tolerance	trade-off.	New	Phytologist	191:	1141–1149.	R	Development	Core	Team.	2008.	R:	A	language	and	environment	for	statistical	computing.	Vienna,	Austria:	R	Foundation	for	Statistical	Computing.	Rabosky	DL.	2013.	Diversity-dependence,	ecological	speciation,	and	the	role	of	competition	in	macroevolution.	Annual	Review	of	Ecology,	Evolution,	and	Systematics	44:	481–502.	Rabosky	DL.	2014.	Automatic	detection	of	key	innovations,	rate	shifts,	and	diversity-dependence	on	phylogenetic	trees.	PLOS	ONE	9:	e89543.	Rai	HS,	O’Brien	HE,	Reeves	PA,	Olmstead	RG,	Graham	SW.	2003.	Inference	of	higher-order	relationships	in	the	cycads	from	a	large	chloroplast	data	set.	Molecular	phylogenetics	and	evolution	29:	350–359.	Rambaut	A,	Drummond	AJ.	Tracer	v1.4,	Available	from	http://beast.bio.ed.uk/Tracer.	Reich	PB,	Oleksyn	J.	2004.	Global	patterns	of	plant	leaf	N	and	P	in	relation	to	temperature	and	latitude.	Proceedings	of	the	National	Academy	of	Sciences	101:	11001–11006.	 162 Repenning	CA.	1976.	Adaptive	evolution	of	sea	lions	and	walruses.	Systematic	Zoology	25:	375–390.	Revell	LJ.	2012.	phytools:	an	R	package	for	phylogenetic	comparative	biology	(and	other	things).	Methods	in	Ecology	and	Evolution	3:	217–223.	Roberson	L,	Coyer	J.	2004.	Variation	in	blade	morphology	of	the	kelp	Eisenia	arborea:	incipient	speciation	due	to	local	water	motion?	Marine	Ecology	Progress	Series	282:	115–128.	Robuchon	M,	Valero	M,	Thiébaut	E,	Gall	LL.	2017.	Multi-scale	drivers	of	community	diversity	and	composition	across	tidal	heights:	an	example	on	temperate	seaweed	communities.	Journal	of	Ecology	105:	1791–1805.	Roderick	ML,	Cochrane	MJ.	2002.	On	the	conservative	nature	of	the	leaf	mass–area	relationship.	Annals	of	Botany	89:	537–542.	Ronquist	F,	Teslenko	M,	van	der	Mark	P,	Ayres	DL,	Darling	A,	Höhna	S,	Larget	B,	Liu	L,	Suchard	MA,	Huelsenbeck	JP.	2012.	MrBayes	3.2:	Efficient	Bayesian	phylogenetic	inference	and	model	choice	across	a	large	model	space.	Systematic	Biology	61:	539–542.	Rothman	MD,	Lydiane	M,	Anderson	RJ,	Bolton	JJ,	Verbruggen	H.	2017.	A	phylogeographic	investigation	of	the	kelp	genus	Laminaria	(Laminariales,	Phaeophyceae),	with	emphasis	on	the	South	Atlantic	Ocean.	Journal	of	Phycology	53:	778–789.	Sack	L,	Marañón	T,	Grubb	PJ.	2002.	Global	allocation	rules	for	patterns	of	biomass	partitioning.	Science	296:	1923–1923.	Sakanishi	Y,	Kasai	H,	Tanaka	J.	2017.	Trade-off	relationship	between	productivity	and	thallus	toughness	in	Laminariales	(Phaeophyceae).	Phycological	Research	65:	103–110.	Saunders	GW,	Druehl	LD.	1992.	Nucleotide	sequences	of	the	small-subunit	ribosomal	RNA	genes	from	selected	Laminariales	(Phaeophyta):	implications	for	kelp	evolution.	Journal	of	Phycology	28:	544–549.	Saunders	GW,	McDevit	DC.	2012.	Methods	for	DNA	barcoding	photosynthetic	protists	emphasizing	the	macroalgae	and	diatoms.	In:	DNA	barcodes.	Springer,	207–222.	Saunders	GW,	McDevit	DC.	2014.	A	DNA	barcode	survey	of	Haida	Gwaii	kelp	(Laminariales,	Phaeophyceae)	reveals	novel	ecological	and	distributional	observations	and	Saccharina	druehlii	sp.	nov.	Botany	92:	821–826.	Savage	VM,	Bentley	LP,	Enquist	BJ,	Sperry	JS,	Smith	DD,	Reich	PB,	Allmen	EI	von.	2010.	Hydraulic	trade-offs	and	space	filling	enable	better	predictions	of	vascular	structure	and	function	in	plants.	Proceedings	of	the	National	Academy	of	Sciences	107:	22722–22727.	 163 Savage	VM,	Gillooly	JF,	Woodruff	WH,	West	GB,	Allen	AP,	Enquist	BJ,	Brown	JH.	2004.	The	predominance	of	quarter-power	scaling	in	biology.	Functional	Ecology	18:	257–282.	Schaffelke	B,	Lüning	K.	1994.	A	circannual	rhythm	controls	seasonal	growth	in	the	kelps	Laminaria	hyperborea	and	L.	digitata	from	Helgoland	(North	Sea).	European	Journal	of	Phycology	29:	49–56.	Schluter	D.	1996.	Ecological	causes	of	adaptive	radiation.	The	American	Naturalist	148:	S40–S64.	Schluter	D.	2000.	The	Ecology	of	Adaptive	Radiation.	OUP	Oxford.	Shimodaira	H,	Hasegawa	M.	1999.	Multiple	comparisons	of	log-likelihoods	with	applications	to	phylogenetic	Inference.	Molecular	Biology	and	Evolution	16:	1114–1114.	Shimodaira	H,	Hasegawa	M.	2001.	CONSEL:	for	assessing	the	confidence	of	phylogenetic	tree	selection.	Bioinformatics	17:	1246–1247.	Silberfeld	T,	Leigh	JW,	Verbruggen	H,	Cruaud	C,	de	Reviers	B,	Rousseau	F.	2010.	A	multi-locus	time-calibrated	phylogeny	of	the	brown	algae	(Heterokonta,	Ochrophyta,	Phaeophyceae):	investigating	the	evolutionary	nature	of	the	‘brown	algal	crown	radiation’.	Molecular	Phylogenetics	and	Evolution	56:	659–674.	Silberfeld	T,	Rousseau	F,	Reviers	B	de.	2014.	An	Updated	classification	of	brown	algae	(Ochrophyta,	Phaeophyceae).	Cryptogamie,	Algologie	35:	117–156.	Sillett	SC,	Van	Pelt	R,	Koch	GW,	Ambrose	AR,	Carroll	AL,	Antoine	ME,	Mifsud	BM.	2010.	Increasing	wood	production	through	old	age	in	tall	trees.	Forest	Ecology	and	Management	259:	976–994.	Sillett	SC,	Van	Pelt	R,	Kramer	RD,	Carroll	AL,	Koch	GW.	2015.	Biomass	and	growth	potential	of	Eucalyptus	regnans	up	to	100m	tall.	Forest	Ecology	and	Management	348:	78–91.	Silvertown	J,	McConway	K,	Gowing	D,	Dodd	M,	Fay	MF,	Joseph	JA,	Dolphin	K.	2006.	Absence	of	phylogenetic	signal	in	the	niche	structure	of	meadow	plant	communities.	Proceedings	of	the	Royal	Society	B:	Biological	Sciences	273:	39–44.	Smale	DA,	Burrows	MT,	Moore	P,	O’Connor	N,	Hawkins	SJ.	2013.	Threats	and	knowledge	gaps	for	ecosystem	services	provided	by	kelp	forests:	a	northeast	Atlantic	perspective.	Ecology	and	evolution	3:	4016–4038.	Stadler	T,	Rabosky	DL,	Ricklefs	RE,	Bokma	F.	2014.	On	age	and	species	richness	of	higher	taxa.	The	American	Naturalist	184:	447–455.	Stamatakis	A.	2014.	RAxML	version	8:	a	tool	for	phylogenetic	analysis	and	post-analysis	of	large	phylogenies.	Bioinformatics	30:	1312–1313.	 164 Starko	S,	Boo	GH,	Martone	PT,	Lindstrom	SC.	2018a.	A	molecular	investigation	of	Saccharina	sessilis	from	the	Aleutian	Islands	reveals	a	species	complex,	necessitating	the	new	combination	Saccharina	subsessilis.	ALGAE	33:	157–166.	Starko	S,	Mansfield	SD,	Martone	P.	2018b.	Cell	wall	chemistry	and	tissue	structure	underlie	shifts	in	material	properties	of	a	perennial	kelp.	European	Journal	of	Phycology	53:	307–317.	Starko	S,	Martone	P.	2016a.	An	empirical	test	of	‘universal’	biomass	scaling	relationships	in	kelps:	Evidence	of	convergence	with	seed	plants.	New	Phytologist	212:	719–729.	Starko	S,	Martone	P.	2016b.	Evidence	of	an	evolutionary-developmental	trade-off	between	drag	avoidance	and	tolerance	strategies	in	wave-swept	intertidal	kelps	(Laminariales,	Phaeophyceae).	Journal	of	Phycology	52:	54–63.	Starko	S,	Z	Claman	B,	Martone	P.	2015.	Biomechanical	consequences	of	branching	in	flexible	wave-swept	macroalgae.	New	Phytologist	206:	133–140.	Steinberg	PD.	1985.	Feeding	preferences	of	Tegula	funebralis	and	chemical	defenses	of	marine	brown	algae.	Ecological	Monographs	55:	333–349.	Steinberg	PD,	Estes	JA,	Winter	FC.	1995.	Evolutionary	consequences	of	food	chain	length	in	kelp	forest	communities.	Proceedings	of	the	National	Academy	of	Sciences	92:	8145–8148.	Steneck	RS,	Graham	MH,	Bourque	BJ,	Corbett	D,	Erlandson	JM,	Estes	JA,	Tegner	MJ.	2002.	Kelp	forest	ecosystems:	biodiversity,	stability,	resilience	and	future.	Environmental	conservation	29:	436–459.	Stephens	TA,	Hepburn	CD.	2016.	A	kelp	with	integrity:	Macrocystis	pyrifera	prioritises	tissue	maintenance	in	response	to	nitrogen	fertilisation.	Oecologia	182:	71–84.	Storey	KB,	Storey	JM.	1986.	Freeze	tolerance	and	intolerance	as	strategies	of	winter	survival	in	terrestrially-hibernating	amphibians.	Comparative	biochemistry	and	physiology.	A,	Comparative	physiology	83:	613–617.	Swenson	NG,	Enquist	BJ.	2007.	Ecological	and	evolutionary	determinants	of	a	key	plant	functional	trait:	wood	density	and	its	community-wide	variation	across	latitude	and	elevation.	American	Journal	of	Botany	94:	451–459.	Teagle	H,	Hawkins	SJ,	Moore	PJ,	Smale	DA.	2017.	The	role	of	kelp	species	as	biogenic	habitat	formers	in	coastal	marine	ecosystems.	Journal	of	Experimental	Marine	Biology	and	Ecology	492:	81–98.	Tellier	F,	Tapia	J,	Faugeron	S,	Destombe	C,	Valero	M.	2011.	The	Lessonia	nigrescens	species	complex	(Laminariales,	Phaeophyceae)	shows	strict	parapatry	and	complete	reproductive	isolation	in	a	secondary	contact	zone.	Journal	of	Phycology	47:	894–903.	 165 Thompson	PL,	Davies	TJ,	Gonzalez	A.	2015.	Ecosystem	functions	across	trophic	levels	are	linked	to	functional	and	phylogenetic	diversity.	PLoS	ONE	10:	e0117595.	Thomsen	MS,	Wernberg	T,	Kendrick	GA.	2005a.	The	effect	of	thallus	size,	life	stage,	aggregation,	wave	exposure	and	substratum	conditions	on	the	forces	required	to	break	or	dislodge	the	small	kelp	Ecklonia	radiata.	Botanica	Marina	47:	454–460.	Thomsen	MS,	Wernberg	T,	Kendrick	GA.	2005b.	The	effect	of	thallus	size,	life	stage,	aggregation,	wave	exposure	and	substratum	conditions	on	the	forces	required	to	break	or	dislodge	the	small	kelp	Ecklonia	radiata.	Botanica	Marina	47:	454–460.	Touchette	BW,	Iannacone	LR,	Turner	GE,	Frank	AR.	2007.	Drought	tolerance	versus	drought	avoidance:	A	comparison	of	plant-water	relations	in	herbaceous	wetland	plants	subjected	to	water	withdrawal	and	repletion.	Wetlands	27:	656–667.	Ulrich	W,	Sewerniak	P,	Puchałka	R,	Piwczyński	M.	2017.	Environmental	filtering	triggers	community	assembly	of	forest	understorey	plants	in	Central	European	pine	stands.	Scientific	Reports	7:	274.	Utter		null,	Denny		null.	1996.	Wave-induced	forces	on	the	giant	kelp	Macrocystis	pyrifera	(Agardh):	field	test	of	a	computational	model.	The	Journal	of	Experimental	Biology	199:	2645–2654.	van	der	Valk	AG.	1981.	Succession	in	wetlands:	a	Gleasonian	approach.	Ecology	62:	688–696.	Van	Alstyne	KL,	McCarthy	III	JJ,	Hustead	CL,	Duggins	DO.	1999.	Geographic	variation	in	polyphenolic	levels	of	Northeastern	Pacific	kelps	and	rockweeds.	Marine	Biology	133:	371–379.	Vermeij	GJ.	2012.	The	evolution	of	gigantism	on	temperate	seashores.	Biological	Journal	of	the	Linnean	Society	106:	776–793.	Vermeij	GJ,	Banker	R,	Capece	LR,	Hernandez	ES,	Salley	SO,	Vriesman	VP,	Wortham	BE.	2018.	The	coastal	North	Pacific:	Origins	and	history	of	a	dominant	marine	biota.	Journal	of	Biogeography.	Vogel	S.	1989.	Drag	and	reconfiguration	of	broad	leaves	in	high	winds.	Journal	of	Experimental	Botany	40:	941–948.	Vogel	S.	1994.	Life	in	moving	fluids.	Princeton:	Princeton	University	Press.	Wainwright	SA,	Biggs	WD,	Currey	JD,	Gosline	JM.	1976.	Mechanical	Design	in	Organisms.	Princeton:	Princeton	University	Press.	Warton	DI,	Duursma	RA,	Falster	DS,	Taskinen	S.	2012.	smatr	3–	an	R	package	for	estimation	and	inference	about	allometric	lines.	Methods	in	Ecology	and	Evolution	3:	257–259.	 166 Webb	CO.	2000.	Exploring	the	phylogenetic	structure	of	ecological	communities:	an	example	for	rain	forest	trees.	The	American	Naturalist	156:	145–155.	Webb	CO,	Ackerly	DD,	McPeek	MA,	Donoghue	MJ.	2002.	Phylogenies	and	community	ecology.	Annual	Review	of	Ecology	and	Systematics	33:	475–505.	Weber	MG,	Wagner	CE,	Best	RJ,	Harmon	LJ,	Matthews	B.	2017.	Evolution	in	a	Community	Context:	On	Integrating	Ecological	Interactions	and	Macroevolution.	Trends	in	Ecology	&	Evolution	32:	291–304.	Wernberg	T.	2005.	Holdfast	aggregation	in	relation	to	morphology,	age,	attachment	and	drag	for	the	kelp	Ecklonia	radiata.	Aquatic	Botany	82:	168–180.	Wernberg	T,	Bennett	S,	Babcock	RC,	Bettignies	T	de,	Cure	K,	Depczynski	M,	Dufois	F,	Fromont	J,	Fulton	CJ,	Hovey	RK,	et	al.	2016.	Climate-driven	regime	shift	of	a	temperate	marine	ecosystem.	Science	353:	169–172.	Wernberg	T,	Coleman	M,	Fairhead	A,	Miller	S,	Thomsen	M.	2003.	Morphology	of	Ecklonia	radiata	(Phaeophyta:	Laminarales)	along	its	geographic	distribution	in	south-western	Australia	and	Australasia.	Marine	Biology	143:	47–55.	Wernberg	T,	Krumhansl	K,	Filbee-Dexter	K,	Pedersen	MF.	2019.	Chapter	3	-	Status	and	Trends	for	the	World’s	Kelp	Forests.	In:	Sheppard	C,	ed.	World	Seas:	an	Environmental	Evaluation	(Second	Edition).	Academic	Press,	57–78.	Wernberg	T,	Smale	DA,	Tuya	F,	Thomsen	MS,	Langlois	TJ,	de	Bettignies	T,	Bennett	S,	Rousseaux	CS.	2012.	An	extreme	climatic	event	alters	marine	ecosystem	structure	in	a	global	biodiversity	hotspot.	Nature	Climate	Change	3:	78–82.	Wernberg	T,	Thomsen	MS.	2005.	The	effect	of	wave	exposure	on	the	morphology	of	Ecklonia	radiata.	Aquatic	Botany	83:	61–70.	Wernberg	T,	Vanderklift	MA.	2010.	Contribution	of	Temporal	and	Spatial	Components	to	Morphological	Variation	in	the	Kelp	Ecklonia	(laminariales)1.	Journal	of	Phycology	46:	153–161.	West	GB,	Brown	JH,	Enquist	BJ.	1997.	A	general	model	for	the	origin	of	allometric	scaling	laws	in	biology.	Science	276:	122–126.	West	GB,	Brown	JH,	Enquist	BJ.	1999a.	The	fourth	dimension	of	life:	fractal	geometry	and	allometric	scaling	of	organisms.	Science	284:	1677–1679.	West	GB,	Brown	JH,	Enquist	BJ.	1999b.	A	general	model	for	the	structure	and	allometry	of	plant	vascular	systems.	Nature	400:	664–667.	 167 Westoby	M,	Falster	DS,	Moles	AT,	Vesk	PA,	Wright	IJ.	2002.	Plant	ecological	strategies:	some	leading	dimensions	of	variation	between	species.	Annual	Review	of	Ecology	and	Systematics	33:	125–159.	Westoby	M,	Wright	IJ.	2006.	Land-plant	ecology	on	the	basis	of	functional	traits.	Trends	in	Ecology	&	Evolution	21:	261–268.	Widdowson	TB.	1959.	Some	aspects	of	the	intertidal	ecology	of	marine	organisms	on	Vancouver	Island	between	Victoria	and	Port	Renfrew.	Wood	SN.	2011.	Fast	stable	restricted	maximum	likelihood	and	marginal	likelihood	estimation	of	semiparametric	generalized	linear	models.	Journal	of	the	Royal	Statistical	Society:	Series	B	(Statistical	Methodology)	73:	3–36.	Wright	IJ,	Reich	PB,	Westoby	M,	Ackerly	DD,	Baruch	Z,	Bongers	F,	Cavender-Bares	J,	Chapin	T,	Cornelissen	JHC,	Diemer	M,	et	al.	2004.	The	worldwide	leaf	economics	spectrum.	Nature	428:	821–827.	Zhang	L,	Wang	X,	Liu	T,	Wang	H,	Wang	G,	Chi	S,	Liu	C.	2015.	Complete	Plastid	Genome	of	the	Brown	Alga	Costaria	costata	(Laminariales,	Phaeophyceae).	PLOS	ONE	10:	e0140144.	  168 Appendix A – Supplementary Information for Chapter 2 Table A1. Collection information for samples sequenced in this study Species Collector Herbarium Voucher  Collection Locality GPS Coordinates Akkesiphycus lubricus Yamada & Tanaka H. Kawai KU 199 Japan (from culture) - Pseudochorda nagaii (Tokida) Inagaki H. Kawai KU 482 Japan (from culture) - Chorda asiatica Sasaki & Kawai H. Kawai KU 494 Japan (from culture) - Alaria marginata Postels & Ruprecht K. Demes UBC A93071 Santa Cruz County, CA, USA N 37.182 W 122.389 Aureophycus aleuticus H.Kawai, T.Hanyuda, Lindeberg & S.C.Lindstrom H. Kawai KU d12535 St. George Island, Alaska, USA N 56.608 W 169.682 Lessoniopsis littoralis (Farlow & Setchell ex Tilden) Reinke K. Demes UBC A93072 Soberanes, Monterey County, CA, USA N 36.453 W 121.930 Pterygophora californica Ruprecht S. Starko NA Victoria, BC, Canada N 48.413694, W 123.385152  Pleurophycus gardneri Setchell & Saunders ex J.Tilden S. Starko UBC A93709 Victoria, BC, Canada N 48.413694, W 123.385152  Thalassiophyllum clathratum (S.G.Gmelin) Postels & Ruprecht H. Chenelot UBC A89431 Alaska, USA N 51.83722, W 176.5308 Costaria costata K. Demes UBC A93093 Vancouver, BC, Canada  N 49.37213,   W 123.29231  Neoagarum fimbriatum (Harvey) H.Kawai & T.Hanyuda K. Demes UBC A93091  Catalina Island, CA, USA N 33.444 W 118.485 Dictyoneurum californicum Ruprecht   K. Demes UBC A93094  Hopkins, Monterey Bay, CA, USA N 36.622 121.9022  169 Laminaria ephemera Setchell K. Demes UBC A93077  Still Water Cove, Monterey County, CA, USA N 36.564 W 121.945 Lessonia variegate J.Agardh W. Nelson WELT ASQ124 West Lyall Bay, Wellington, North Island, New Zealand S 41.33956 E 174.7933 Lessonia spicata (Suhr) Santelices E. Macaya NA Cocholgue, Chile S 36.59,  W 72.97 Egregia menziesii (Turner) Areschoug K. Demes NA Botanical Beach, BC, Canada N 48.525961 W 124.447283 Ecklonia arborea (Areschoug) M.D.Rothman, Matio & J.J.Bolton S. Starko NA Tapaltos Beach, BC, Canada N 48.80104, -W 125.18334 Ecklonia radicosa (Kjellman) Okamura H. Kawai NA Japan   Cymathaere triplicata (Postels & Ruprecht) J.Agardh  S. Starko UBC A93708 Victoria, BC, Canada N 48.413694, W 123.385152  Postelsia palmaeformis Ruprecht K. Demes UBC A93076  Soberanes, Monterey County, CA, USA N 36.452778 W 121.930000 Macrocystis pyrifera (Linnaeus) C.Agardh S. Starko NA Dixon Island, Bamfield, BC, Canada N 48.853709, W 125.121658 Pelagophycus porra (Léman) Setchell M. Edwards NA San Diego, California, USA N 32.6982 W 117.2650 Arthrothamnus bifidus N. Yotsukuro NA Hokkaido, Japan N 43.29472 W 145.6002 Saccharina subsessilis (Areschoug) Starko & S.C.Lindstrom S. Lindstrom UBC A93599 Attu Island, Alaska, USA N 52.93019,  W 173.25461 Saccharina nigripes (J.Agardh) Lontin & G.W.Saunders S. Starko NA  Cape Beale, BC, Canada N 48.784116, W 125.215875 Sphacelaria sp. P. Martone UBC A89833  Calvert Island, BC, Canada N 51.66542, -W 128.13530 Analipus japonicas (Harvey) M.J.Wynne   M. Lemay NA Calvert Island, BC, Canada N 51.65077,  W 128.14397  170 Desmarestia aculeata (Linnaeus) J.V.Lamouroux P. Martone UBC A89817  Calvert Island, BC, Canada N 51.66542, -W 128.13530                             171 Table A2. Genbank sequence read archive (SRA) accession numbers for raw sequence reads acquired in this study. Species Biosample ID Akkesiphycus lubricus SAMN09506682 Pseudochorda nagaii SAMN09506683 Chorda asiatica SAMN09506684 Alaria marginata SAMN09506685 Aureophycus aleuticus SAMN09506686 Lessoniopsis littoralis SAMN09506687 Pterygophora californica SAMN09506688 Pleurophycus gardneri SAMN09506689 Thalassiophyllum clathratum SAMN09516489 Costaria costata SAMN09516490 Neoagarum fimbriatum SAMN09516491 Dictyoneurum californicum SAMN09516493 Laminaria ephemera SAMN09516492 Lessonia variegata SAMN09516494 Lessonia splicans SAMN09516495 Egregia menziesii SAMN09516496 Ecklonia arborea SAMN09516497 Ecklonia radicosa SAMN09516498 Cymathaere triplicata SAMN09516531 Postelsia palmaeformis SAMN09516532 Macrocystis pyrifera SAMN09516533 Pelagophycus porra SAMN09516534 Arthrothamnus bifidus SAMN09516535 Hedophyllum subsessile SAMN09516536 Hedophyllum nigripes SAMN09516537 Sphacelaria sp. SAMN10026787 Analipus japonicus SAMN10026788 Desmarestia aculeata SAMN10026789        172 Table A3. Sequencing methods used for each sample and matches to previously published sequences to confirm species identification.  Species Extraction technique Library  Percent complete (Ribosome) Percent complete (Mitome) Percent complete (Plastid) Sequence Match   Alaria marginata CTAB BioO NEXTFlex 100 % 100 % 100 % 100 % match CO1-5P Aureophycus aleuticus CTAB BioO NEXTFlex 100 % 81 % 44 % 100 % match CO1-5P Lessoniopsis littoralis CTAB BioO NEXTFlex 100 % 100 % 100 % 100 % match CO1-5P Pterygophora californica CTAB BioO NEXTFlex 100 % 100 % 100 % 100 % match CO1-5P Pleurophycus gardneri CTAB BioO NEXTFlex 100 % 100 % 100% 100 % match CO1-5P Thalassiophyllum clathratum CTAB BioO NEXTFlex 100 % 100 % 100 % 100 % match CO1-5P Costaria costata CTAB BioO NEXTFlex 100 % 100 % 100 % 100 % match CO1-5P Neoagarum fimbriatum Wizard Clean-Up Kit BioO NEXTFlex 100 % 100 % 100 % 100 % match CO1-5P Dictyoneurum californicum CTAB BioO NEXTFlex 100 % 100 % 100% 100 % match CO1-5P Laminaria ephemera Wizard Clean-Up Kit BioO NEXTFlex 100 % 100 % 100 % 100 % match CO1-5P Lessonia variegata CTAB BioO NEXTFlex 100 % 100 % 100 % 100% match  atp8 Lessonia splicata CTAB BioO NEXTFlex 100 % 100 % 100 % 97 % match CO1-5P1 Egregia menziesii DNEasy Plant Mini Kit Nugen Ovation Ultra Low 100 % 97 % 100 % 100% match CO1-5P Ecklonia arborea CTAB BioO NEXTFlex 100 % 100 % 100 % 100 % match CO1-5P Ecklonia radicosa CTAB BioO NEXTFlex 100 % 100 % 100 % 99 % match  ITS Cymathaere triplicata CTAB BioO NEXTFlex 100 % 99 % 100 % 99 % match CO1-5P Postelsia palmaeformis CTAB BioO NEXTFlex 100 % 99 % 100 % 99 % match CO1-5P Macrocystis pyrifera CTAB BioO NEXTFlex 100 % 100 % 100 % 100 % match CO1-5P Pelagophycus porra CTAB NEB Ultra II 100 % 100 % 100 % 100 % match CO1-5P Arthrothamnus bifidus CTAB NEB Ultra II 100 % 97 % 100 % 99% match  ITS Hedophyllum subsessile CTAB NEB Ultra II 100 % 100 % 100 % 100 % match CO1-5P  173 Hedophyllum nigripes CTAB NEB Ultra II 100 % 100 % 100 % 100 % match CO1-5P Chorda asiatica DNEasy Plant Mini Kit BioO NEXTFlex 100 % 92 % 93 % 98 % match  CO1-5P Pseudochorda nagaii DNEasy Plant Mini Kit BioO NEXTFlex 100 % 80 % 93 % 100 % match CO1-5P Akkesiphycus lubricum DNEasy Plant Mini Kit BioO NEXTFlex 100 % 95 % 98 % 100 % match CO1-5P Sphacelaria sp. Wizard DNA Clean-up BioO NEXTFlex 100 % 94 % 93 % 96% match rbcL – Sphacelaria caespitula Analipus japonicus Wizard DNA Clean-up BioO NEXTFlex 100 % 99 % 98 % 93 % match – CO1-5P, 99% rbcL Desmarestia aculeata Wizard DNA Clean-up BioO NEXTFlex 100 % 93 % 95 % 100 % match CO1-5P 1Lessonia splicata has never been sequenced before and could only be confirmed to genus                 174 Table A4. Sources of previously published genomic data  Species Data Study Genbank or Database number Agarum clathratum Dumortier Transcriptome Jackson et al 2017 Biosample: SAMN05000119 Chorda filum (Linnaeus) Stackhouse Transcriptome Jackson et al 2017 Biosample: SAMN05000111  Desmarestia viridis (O.F.Müller) J.V.Lamouroux Mitochondrion Secq et al 2006 NC_007684.1  Ecklonia radiate (C.Agardh) J.Agardh Transcriptome Jackson et al 2017 Biosample: SAMN05000107 Ectocarpus siliculosus (Dillwyn) Lyngbye Mitochondrion Cock et al 2010 NC_030223.1   Ectocarpus siliculosus (Dillwyn) Lyngbye Plastid Le Corguille et al 2009 FP102296.1  Fucus vesiculosus Linnaeus Mitochondrion Secq et al 2006 NC_007683.1  Fucus vesiculosus Linnaeus Plastid Le Corguille et al 2009 NC_016735.1  Saccharina sculpera C.E.Lane, C.Mayes, Druehl & G.W.Saunders Mitochondrion Zhang et al 2016 NC_029206.1  Laminaria digitata (Hudson) J.V.Lamouroux    Transcriptome (Plastid) Jackson et al 2017 Biosample: SAMN05000120 Laminaria digitata (Hudson) J.V.Lamouroux Mitochondrion Oudot-Le Secq et al 2001 NC_004024.1  Laminaria hyperborea (Gunnerus) Fosile Mitochondrion Zhang & Liu  NC_021639.1  Lessonia nigrescens Bory Transcriptome Jackson et al 2017 Biosample: SAMN05000110 Nereocystis luetkeana (K.Mertens) Postels & Ruprecht Transcriptome Jackson et al 2017 Biosample: SAMN05000112 Pylaiella littoralis (Linnaeus) Kjellman Mitochondrion Oudot-Le Secq et al 2001 NC_003055.1   175 Saccharina angustata (Kjellman) C.E.Lane, C.Mayes, Druehl & G.W.Saunders Mitochondrion Yotsukura et al NC_013473.1  Saccharina cichorioides (Miyabe) C.E.Lane, C.Mayes, Druehl, G.W.Saunders Mitochondrion Yotsukura et al NC_013475.1  Saccharina japonica (Areshoug) C.E.Lane, C.Mayes, Druehl, G.W.Saunders Mitochondrion Yotsukura et al AP011493  Saccharina japonica (Areshoug) C.E.Lane, C.Mayes, Druehl, G.W.Saunders Plastid Wang et al 2013 JQ405663.1  Saccharina latissima (Linnaeus) C.E.Lane, C.Mayes, Druehl, G.W.Saunders Transcriptome Jackson et al 2017 Biosample: SAMN05000109 Saccharina longissima (Miyabe) C.E.Lane, C.Mayes, Druehl, G.W.Saunders Mitochondrion Zhang & Liu NC_021640.1  Saccharina sp yeC Mitochondrion Teng et al 2017 KT315643 Sargassum horneri (Turner) C.Agardh Mitochondrion Liu et al 2014 NC_024613  Undaria pinnatifida (Harvey) Suringar Mitochondrion Li et al 2014 NC_023354  Undaria pinnatifida (Harvey) Suringar Plastid Zhang et al 2015 NC_028503  Alaria esculenta (Linnaeus) Greville Transcriptome Jackson et al 2017 Biosample: SAMN05000119         176 Table A5. Completeness of previously published phylogenomic datasets   Species Data Percent complete (Ribosome) Percent complete (Mitome) Percent complete (Plastid) Genbank or Database number Agarum clathratum Transcriptome 85 % 93 % 45 % Biosample: SAMN05000119 Chorda filum Transcriptome 77 % 46 % 49 % Biosample: SAMN05000111  Desmarestia viridis Mitochondrion - 100 % NA NC_007684.1  Ecklonia radiata Transcriptome 85% 48 % 51 % Biosample: SAMN05000107 Ectocarpus siliculosus Mitochondrion - 100 % NA NC_030223.1   Ectocarpus siliculosus Plastid - NA 100 % FP102296.1  Fucus vesiculosus Mitochondrion - 100 % NA NC_007683.1  Fucus vesiculosus Plastid - NA 100 % NC_016735.1  Kjellmaniella crassifolia Mitochondrion - 100 % NA NC_029206.1  Laminaria digitata Transcriptome (Plastid and Ribosome) 15 % NA 12 % Biosample: SAMN05000120 Laminaria digitata Mitochondrion - 100 % NA NC_004024.1  Laminaria hyperborea Mitochondrion - 100 % NA NC_021639.1  Lessonia nigrescens Transcriptome 85 % 65 % 42 % Biosample: SAMN05000110 Nereocystis luetkeana Transcriptome 46 % 42 % 41 % Biosample: SAMN05000112 Pyllaella littoralis Mitochondrion - 100 % NA NC_003055.1  Saccharina angustata Mitochondrion - 100 % NA NC_013473.1  Saccharina coriaceae Mitochondrion - 100 % NA NC_013475.1   177 Saccharina japonica Mitochondrion - 100 % NA AP011493  Saccharina japonica Plastid - NA 100 % JQ405663.1  Saccharina latissima Transcriptome 81 % 30 % 21 % Biosample: SAMN05000109 Saccharina longissima Mitochondrion - 100 % NA NC_021640.1  Saccharina sp yeC Mitochondrion - 100 % NA KT315643 Sargassum horneri Mitochondrion - 100 % NA NC_024613  Undaria pinnatifida Mitochondrion - 100 % NA NC_023354  Undaria pinnatifida Plastid - NA 100 % NC_028503  Alaria esculenta Transcriptome 100 % 76 % 47 % Biosample: SAMN05000119                178 Table A6. New taxonomic scheme for the “Saccharina” sensu lato  Species New Genus Confirmed by Genbank No. Arthrothamnus bifidus Arthrothamnus Main - Saccharina bongardiana Hedophyllum CO1 (Fig S7) GU097729 Saccharina dentigera Hedophyllum CO1 (Fig S7) MH327956 Saccharina druehlii Hedophyllum CO1 (Fig S7) KJ960273 Saccharina nigripes Hedophyllum Main - Saccharina sessilis Hedophyllum CO1 (Fig S7) FJ409208 Saccharina subsessilis Hedophyllum Main - Saccharina angustata Saccharina Main - Saccharina angustissima Saccharina CO1 (Fig S7) MF156536 Saccharina cichorioides Saccharina Main - Saccharina gyrata Saccharina NAD6 (Fig S8) AY857925 Saccharina japonica Saccharina Main - Saccharina kurilensis Saccharina ITS (99 % similar to S. latissima)  AF318983 Saccharina latissima Saccharina Main - Saccharina longissima Saccharina Main - Saccharina sachalinensis Saccharina NAD6 (Fig S8) AB480844 Arthrothamnus kurilensis Unconfirmed No Data ND Saccharina complanata Unconfirmed No Data  ND Saccharina crassifolia Unconfirmed No Data ND Saccharina gurjunovae Unconfirmed No Data ND Saccharina lanciformis Unconfirmed No Data ND Saccharina longicuris Unconfirmed No Data ND           179 Table A7. Divergence times and 95% HPD of major clades from molecular clock analyses  Clade Stem Age 95% HPD Crown Age 95% HPD Alariaceae 25.5 14.51 - 37.66 16.02 8.69 - 24.38 Agaraceae 29.08 20.08 - 40.31 15.64 8.09 - 23.95 Aureophycaceae 31.43 21.42 - 43.70 25.5 14.51 - 37.66 Arthrothamnaceae 26.39 18.41 - 36.22 25 17.54 - 34.27 Laminariaceae 23.08 14.81 - 32.97 16.05 6.99 - 25.00 Lessoniaceae 23.08 14.81 - 32.97 7.44 2.94 - 12.87 Chordales 73.23 41.15 - 108.47 62.69 33.37 - 95.63                   180  Fig A1. Ancestral state reconstruction of (a) paraphyses structure [unicellular, multicellular, not applicable to life history] and (b) gamete types [oogamous, anisogamous, isogamous] using an equal rates model.       181  Fig A2. Ancestral state reconstruction of (a) stipe differentiation [present or absent] and (b) growth [meristematic or diffuse] using an equal rates model.      182  Fig A3. Ancestral state reconstruction of (a) gametophyte sexual dimorphism [present or absent] and (b) eyespots on zoospore [present or absent] using an equal rates model.         183  Fig A4. Ancestral state reconstruction of life history [annual or perennial] using an equal rates model.       Agarum clathratumThalassiophylum clathrumAgarum fimbriatumDictyoneurum californicumCostaria costataArthrothamnus bifidusSaccharina groenlandicaSaccharina subsessilisSaccharina angustataSaccharina japonicaSaccharina longissimaSaccharina coriaceaeSaccharina latissimaSaccharina sculperaMacrocystis pyriferaPelagophycus porraNereocystis luetkeanaPostelsia palmaeformisPseudolessonia laminarioidesCymathaere triplicataEcklonia arboreaEcklonia radiataEcklonia radicosaEgregia menziesiiLaminaria digitataLaminaria hyperboreaLaminaria ephemeraLessonia nigrescensLessonia splicansLessonia variegataAlaria esculentaAlaria marginataUndaria pinnatifidaLessoniopsis littoralisPleurophycus gardneriPterygophora californicaAureophycus aleuticusAkkesiphycus lubricusPseudochorda nagaiiChorda asiaticaChorda filumEctocarpus siliculosusPylaiella littoralisAgarum clathratumThalassiophylum clathrumAgarum fimbriatumDictyoneurum californicumCostaria costataArthrothamnus bifidusSaccharina groenlandicaSaccharina subsessilisSaccharina angustataSaccharina japonicaSaccharina longissimaSaccharina coriaceaeSaccharina latissimaSaccharina sculperaMacrocystis pyriferaPelagophycus porraNereocystis luetkeanaPostelsia palmaeformisPseudolessonia laminarioidesCymathaere triplicataEcklonia arboreaEcklonia radiataEcklonia radicosaEgregia menziesiiLaminaria digitataLaminaria hyperboreaLaminaria ephemeraLessonia nigrescensLessonia splicansLessonia variegataAlaria esculentaAlaria marginataUndaria pinnatifidaLessoniopsis littoralisPleurophycus gardneriPterygophora californicaAureophycus aleuticusAkkesiphycus lubricusPseudochorda nagaiiChorda asiaticaChorda filumEctocarpus siliculosusPylaiella littoralisLife	HistoryAnnualPerennialHoldfastDiscoidHapteralNot	Applicable 184  Fig A5. Maximum likelihood reconstruction of the kelp phylogeny using a partitioned, concatenated 90-gene alignment of plastid genes. Asterisks (*) indicate 100 % bootstrap support; other values are shown next to branches. Scale bar indicates estimated number of substitutions per site.     185  Fig A6. Bayesian reconstruction of the kelp phylogeny performed using a partitioned, concatenated 90 gene alignment of plastid genes. Asterisks (*) indicate 1.0 posterior probability; other values are shown next to branches. Scale bar indicates estimated number of substitutions per site. Nodes that conflict with combined analyses (Fig 1) are colored red.      186  Fig A7. Maximum likelihood reconstruction of the kelp phylogeny performed using a concatenated 28 gene alignment of mitochondrial genes. Asterisks (*) indicate 100 % bootstrap support; other values are shown next to branches. Scale bar indicates estimated number of substitutions per site. Nodes that conflict with combined analyses (Fig 1) are colored red.          187  Fig A8. Bayesian reconstruction of the kelp phylogeny performed in MrBayes using a concatenated 28 gene alignment of mitochondrial genes Asterisks (*) indicate 1.0 posterior probability; other values are shown next to branches. Scale bar indicates estimated number of substitutions per site. Nodes that conflict with combined analyses (Fig 1) are colored red.         188  Fig A9. Maximum likelihood reconstruction of the kelp phylogeny performed in RaxML using a concatenated 120 gene alignment of plastid, mitochondrial and ribosomal genes. Asterisks (*) indicate 100 % bootstrap support; other values are shown next to branches. Scale bar indicates estimated number of substitutions per site. Nodes that conflict with combined analyses (Fig 1) are colored red.         189  Fig A10. Bayesian reconstruction of the kelp phylogeny performed in MrBayes using a concatenated 120 gene alignment of plastid, mitochondrial and ribosomal genes. Asterisk (*) indicates full support from posterior probability (1.0); other values are shown next to branches. Scale bar indicates estimated number of substitutions per site. Nodes that conflict with combined analyses (Fig 1) are colored red.      190  Fig A11. Time calibrated molecular phylogeny reconstructed in BEAST 1.8.4 using a concatenated alignment of 9,561 basepairs. Branch length are in millions of years and node bars are 95% HPD for node ages. Calibrations are indicated by arrows.        191  Fig A12. Historical biogeography of the kelps with branch lengths in millions of years. A time-adjusted phylogeny generated in BEAST 1.8.4 using two fossil and one biogeographical time calibration and a subset of genetic data (psaA, psbA, rbcL, atp6, cox1, cox3, nad6, LSU). Pie charts on the nodes indicate the most likely biogeographical distribution of that ancestor, as predicted by DEC+J model in BioGeoBears.           192  Fig A13. Phlorotannin content of North Pacific kelps plotted on the phylogeny (see text for methods).     193  Fig A14. Maximum likelihood cox1 gene-tree of Saccharina spp. and allies, reconstructed using RaxML. Node labels indicate bootstrap support. While old species names are given as tip label, the new scheme is indicated to the right if each clade. This tree was used to place S. bongardiana, S. dentigera, S. druehlii, S. sessilis into Hedophyllum, and S. angustissima into Saccharina.       194  Fig A15. Maximum likelihood nad6 gene-tree of Saccharina spp. and allies, reconstructed using RaxML. Node labels represent bootstrap support. While old species names are given as tip label, the new scheme is indicated to the right if each clade. This was used to place Saccharina gyrata (formerly Kjellmaniella gyrata) and Saccharina sachalinensis into Saccharina.  195 Appendix B - Supplementary Information for Chapter 3  Fig B1. Raw drag coefficient data for each species.           196 Table B1. Summary of models used to determine Streamlining indices Species Streamlining Index  Intercept R2 Statistic A. marginata 0.2005 -1.5701 0.727 F = 50.482; P < 0.001 C. costata 0.208 -1.5738 0.598 F = 14.900; P < 0.01 E. menziesii 0.1543 -1.5479 0.747 F = 38.421; P < 0.001 L. littoralis 0.0083 -1.2026 0.003 F = 0.0364; P = 0.852 L. setchellii -0.0774 -1.1718 0.052 F = 0.5442; P = 0.477 M. pyrifera 0.3118 -1.7914 0.787 F = 40.664; P < 0.001 S. nigripes 0.1752 -1.5636 0.508 F = 11.339; P < 0.01 S. sessilis -0.009 -1.2847 0.001 F = 0.0109; P = 0.919             197  Fig B2. Raw drag (at 0.97 ms-2) data for each species            198 Table B2. Summary of models used to predict drag at two size classes Species Slope Intercept R2 Statistic A. marginata 0.7995 1.1147 0.977 F = 803.07; P < 0.001 C. costata 0.7920 1.1111 0.956 F = 216.15; P < 0.001 E. menziesii 0.8457 1.1370 0.989 F = 1154.6; P < 0.001 L. littoralis 1.0083 1.4823 0.98 F = 542.60; P < 0.001 L. setchellii 1.0774 1.5131 0.913 F = 105.31; P < 0.001 M. pyrifera 0.6882 0.8935 0.947 F = 198.09; P < 0.001 S.nigripes 0.8248 1.1214 0.958 F = 251.48; P < 0.001 S. sessilis 1.0090 1.4002 0.913 F = 136.95; P < 0.001                  199  Fig B3. Raw tenacity data for each species.            200 Table B3. Summary of models used to determine Tolerance indices and tenacity predictions  Species Tolerance Index  Intercept R2 Statistic A. marginata 0.2934 2.1052 0.514 F = 25.294; P < 0.001 C. costata 0.2176 1.9783 0.257 F = 5.8549; P < 0.05 E. menziesii 0.5323 2.5288 0.842 F = 70.071; P < 0.001 L. littoralis 0.5621 3.1467 0.853 F = 105.86; P < 0.001 L. setchellii 0.764 3.2025 0.847 F = 67.262; P < 0.001 M. pyrifera 0.3677 2.3542 0.695 F = 51.090; P < 0.001 S. nigripes  0.4354 2.2183 0.558 F = 28.815; P < 0.001 S. sessilis 0.5954 2.7732 0.672 F = 27.669; P < 0.001                   201 Table B4.  Collection information for second moment of area measurements Family Species Sample Size1 Min area (cm2) Max area (cm2) Alariaceae Costariaceae Lessoniaceae Laminariaceae Alariaceae Laminariaceae Laminaraiceae Laminariaceae Alaria marginata Costaria costata Egregia menziesii Laminaria setchellii Lessoniopsis littoralis Macrocystis pyrifera Saccharina nigripes Saccharina sessilis N = 19 NS = 10 NS = 14 NS = 18 NS = 9  NS = 9 NS = 11  NS = 15  20 10 60 20 10 50 10 30 2,510 2,220 1,320 1,160 5,810 3,740 1,250 1,200                      202 Table B5. Summary of models used to predict second moment of area at different sizes  Species Slope Intercept R2 Statistic A. marginata 0.7094 -10.235 0.757 F = 54.02; P < 0.001 C. costata 0.7550 -9.6274 0.840 F = 48.17; P < 0.001 E. menziesii 1.9649 -7.7998 0.890 F = 105.90; P < 0.001 L. littoralis 2.1988 -5.9236 0.930 F = 106.70; P < 0.001 L. setchellii 1.7747 -7.1658 0.913 F = 178.80; P < 0.001 M. pyrifera 1.1429 -9.8862 0.943 F = 150.30; P < 0.001 S. nigripes  0.8198 -9.6201 0.831 F = 50.160; P < 0.001 S. sessilis 1.3889 -8.7952 0.861 F = 87.470; P < 0.001       203 Appendix C - Supplementary Information for Chapter 4. 	 Fig C0.1. Total thallus or leaf surface area of kelps (data from this study) and Arabidopsis thaliana (data from Weradewage et al 2015) plotted against total dry weight. 							−6 −5 −4 −3 −2 −1 0−4−3−2−101Log Total Dry Weight (kg)Log Photosynthetic Area (m2 )KelpsArabidopsis   204 	Table C1. Collection site locations Location Site GPS coordinates Barkley Sound Bamfield Inlet 48.835496N, -125.137379W Barkley Sound Brady’s Blowhole 48.823515N, -125.162177W Barkley Sound Cape Beale 48.787036N, -125.217496W Barkley Sound Edward King Island 48.825605N, -125.212080W Barkley Sound Scott’s Bay 48.834453N, -125.147289W Calvert Island West Beach Boulder Field 51.651020N, -128.143544W Vancouver Kitsilano Beach 49.272502N, -123.169715W Victoria Ogden Point 48.414095N, -123.386003W 																	  205 Table C2. Populations and sample sizes used in A vs MT scaling analysis Species Sample Size Site Alaria marginata N = 5 Eagle Bay, Bamfield Costaria costata N = 5 Brady’s Blowhole, Bamfield Laminaria setchellii N = 5 Eagle Bay, Bamfield Saccharina nigripes N = 5 Eagle Bay, Bamfield Saccharina sessilis N = 5 Brady’s Blowhole, Bamfield Laminaria ephemera N = 5 Edward King, Bamfield Lessoniopsis littoralis N = 5 Brady’s Blowhole, Bamfield Agarum fimbriatum N = 3 Bamfield Inlet, Bamfield Macrocystis pyrifera N = 5 Eagle Bay, Bamfield 																	  206 Table C3. Sources for data from which area and dry weight were reported together in the same study  Species Study Acacia implexa Atkin et al (1998) Alnus rubra Hook et al (1990) Arabidopsis thaliana Weradewage et al (2015) Bellus perennis Gunn and Farrar (1999) Betula papyrifera Aphalo & Lehto (1997) Bromus valdevianus Lopez et al  (2013) Cypressus pygmaea Westman & Whittaker (1975) Dactylis glomerata Harmens et al (2000) Eucalyptus regnans Sillett et al (2010), Sillett et al (2015)1 Hordium vulgare Rebetzke et al (2004) Hordium vulgare Rebetzke et al (2004) Lolium perenne Lopez et al  (2013) Pachycereus pringlei Price and Enquist (2006) Pinus contorta Westman & Whittaker (1975) Poa annua Gunn and Farrar (1999) Sequoia sempervirens Sillett et al (2010)2 Triticum sp.  Rebetzke et al (2004) 												  207 	Table C4. Sources for LMA or SLA data that were used to estimate leaf area from the Niklas & Enquist (2004) dataset.  		Species Source Abies alba Wright et al (2005) Abies veitchii Mitamura et al (2008) Betula pendula Wright et al (2005) Cornus sanguinae Wright et al (2005) Elaeis guineensis  Roupsard et al (2006) Fagus crenata Wright et al (2005) Fagus sylvatica Wright et al (2005) Hevea brasiliensis Ray et al (2004) Ilex aquifolium Wright et al (2005) Ligustrum vulgare Wright et al (2005) Oenothera biennis Wright et al (2005) Picea abies Wright et al (2005) Picea sitchensis Wright et al (2005) Pinus nigra Wright et al (2005) Pinus sylvestria Wright et al (2005) Pinus taeda Drake et al (2012) Populus tremuloides Wright et al (2005) Quercus ilex Wright et al (2005) Quercus robur  Wright et al (2005) Rhamnus caltharticus Wright et al (2005) Rosa canina Wright et al (2005) Rubus chamaemorus Wright et al (2005) Taraxicum officinale Wright et al (2005) Tectona grandis Wright et al (2005) 									  208 Table C5. RMA regression coefficients and confidence intervals for area – dry weight scaling in kelps, seed plants and Pachycereus pringlei   Taxa α  (95% CIs) log β (95% CIs) R2 Kelps (Laminariales)  0.78 (0.69 – 0.87) 0.71 (0.54 – 0.90) 0.88 Seed Plants (Embryophyta)  0.73 (0.69 – 0.77) 0.29 (0.28 – 0.30) 0.97 P. pringlei (succulent species) 0.75 (N/A) 0.017 (N/A) 0.99 																	  209 Table C6. Results of ANCOVAs comparing intertidal and subtidal scaling relationships Model Effect of Habitat Test statistic1 P - value Dfmodel N ML vs MR Intercept 12.23 <0.001 1 109  Slope 1.474 >0.05 1 109 MS vs MR Intercept 0.280 >0.05 1 104  Slope 16.64 <0.001 1 104 ML vs MS Intercept 5.40 <0.05 1 104  Slope 0.27 >0.05 1 104 MF vs MR Intercept 13.52 <0.001 1 109  Slope 1.535 >0.05 1 109 1Wald Statistic for intercepts; Likelihood ratio statistic for slopes  210 References for Appendix C  Aphalo, P. J., & Lehto, T. 1997. Effects of light quality on growth and N accumulation in birch seedlings. Tree Physiology, 17 (2): 125-132.	 Atkin, O. K., Schortemeyer, M., McFarlane, N., & Evans, J. R. 1998. Variation in the components of relative growth rate in 10 Acacia species from contrasting environments. Plant, Cell & Environment, 21(10): 1007-1017.  Burger, D. W., Forister, G. W., & Gross, R. 1997. Short and long-term effects of treeshelters on the root and stem growth of ornamental trees. Journal of Arboriculture, 23: 49-56.  Drake, J. E., Oishi, A. C., Giasson, M. A., Oren, R., Johnsen, K. H., & Finzi, A. C. 2012. Trenching reduces soil heterotrophic activity in a loblolly pine (Pinus taeda) forest exposed to elevated atmospheric [CO 2] and N fertilization. Agricultural and Forest Meteorology, 165: 43-52.  Gunn, S., & Farrar, J. F. 1999. Effects of a 4 C increase in temperature on partitioning of leaf area and dry mass, root respiration and carbohydrates. Functional Ecology, 13: 12-20.  Harmens, H., Stirling, C. M., Marshall, C., & Farrar, J. F. 2000. Is partitioning of dry weight and leaf area within Dactylis glomerata affected by N and CO2 enrichment? Annals of Botany, 86(4): 833-839.  Hook, D. D., DeBell, D. S., Ager, A., & Johnson, D. 1990. Dry weight partitioning among 36 open-pollinated red alder families. Biomass, 21(1): 11-25.  López, I. F., Kemp, P. D., Dörner, J., Descalzi, C. A., Balocchi, O. A., & García, S. 2013. Competitive strategies and growth of neighbouring Bromus valdivianus Phil. and Lolium perenne L. plants under water restriction. Journal of Agronomy and Crop Science, 199(6): 449-459.  Mitamura, M., Yamamura, Y., & Nakano, T. 2009. Large-scale canopy opening causes decreased photosynthesis in the saplings of shade-tolerant conifer, Abies veitchii. Tree physiology, 29(1): 137-145.  Niklas, K. J., and B. J. Enquist. 2004. Biomass Allocation and Growth Data of Seeded Plants. Data set. Available on-line [http://www.daac.ornl.gov] from Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A.   Price, C.A. & Enquist, B.J. 2006. Scaling of mass and morphology in plants with minimal branching: an extension of the WBE model. Functional Ecology 20: 11-20.    211 Ray, D., Dey, S. K., & Das, G. 2004. Significance of the leaf area ratio in Hevea brasiliensis under high irradiance and low temperature stress. Photosynthetica, 42(1): 93-97.  Rebetzke, G. J., Botwright, T. L., Moore, C. S., Richards, R. A., & Condon, A. G. 2004. Genotypic variation in specific leaf area for genetic improvement of early vigour in wheat. Field Crops Research, 88(2): 179-189.  Roupsard, O., Bonnefond, J. M., Irvine, M., Berbigier, P., Nouvellon, Y., Dauzat, J., ... & Mialet-Serra, I. 2006. Partitioning energy and evapo-transpiration above and below a tropical palm canopy. Agricultural and Forest Meteorology, 139(3): 252-268.  Sillett, S. C., Van Pelt, R., Koch, G. W., Ambrose, A. R., Carroll, A. L., Antoine, M. E., & Mifsud, B. M. 2010. Increasing wood production through old age in tall trees. Forest Ecology and Management, 259(5): 976-994.  Sillett, S. C., Van Pelt, R., Kramer, R. D., Carroll, A. L., & Koch, G. W. 2015. Biomass and growth potential of Eucalyptus regnans up to 100m tall. Forest Ecology and Management, 348: 78-91.  Weraduwage, S. M., Chen, J., Anozie, F. C., Morales, A., Weise, S. E., & Sharkey, T. D. 2015. The relationship between leaf area growth and biomass accumulation in Arabidopsis thaliana. Frontiers in plant science, 6: 167.  Westman, W. E., & Whittaker, R. H. 1975. The pygmy forest region of northern California: studies on biomass and primary productivity. The Journal of Ecology, 63(2): 493-520.  Wright, I. J., Reich, P. B., Westoby, M., Ackerly, D. D., Baruch, Z., Bongers, F., Cavender-Bares, J., Chapin, T., Cornelissen, J.H.C., Diemer, M., Flexas, J., Garnier, E., Groom, P.K., Gulia, J., Hikosaka, K., Lamont, B.B., Lee, T., Lee, W., Lusk, C., Midgley, J.J., Navas, M., Niinemets, U., Oleksyn, J., Osada, N., Poorter, H., Poot, P., Prior, L., Pyankov, V.I., Roumet, C., Thomas, S.C., Tjoelker, M.G., Veneklaas, E.J., & Villar, R. 2004. The worldwide leaf economics spectrum. Nature, 428(6985): 821-827.  		      

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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0378025/manifest

Comment

Related Items