UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Molecular and field studies of the life history of Acrosiphonia (codiolales) chlorophyta Sussman, Andrea Vera 2000

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

Item Metadata

Download

Media
831-ubc_2000-487261.pdf [ 8.24MB ]
Metadata
JSON: 831-1.0089864.json
JSON-LD: 831-1.0089864-ld.json
RDF/XML (Pretty): 831-1.0089864-rdf.xml
RDF/JSON: 831-1.0089864-rdf.json
Turtle: 831-1.0089864-turtle.txt
N-Triples: 831-1.0089864-rdf-ntriples.txt
Original Record: 831-1.0089864-source.json
Full Text
831-1.0089864-fulltext.txt
Citation
831-1.0089864.ris

Full Text

MOLECULAR AND FIELD STUDIES OF THE LIFE HISTORY OF ACROSIPHONIA (CODIOLALES) CHLOROPHYTA by Andrea Vera Sussmann B.Sc , The University of British Columbia, 1985 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF Doctor of Philosophy in THE F A C U L T Y OF G R A D U A T E STUDIES (Department of Botany) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A April 2000 © Andrea Vera Sussmann, 2000 U B C Special Collections - Thesis Authorisation Form Page 1 of 1 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r reference and study. I f u r t h e r agree that permission f o r extensive copying of t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the head of my department or by h i s or her r e p r e s e n t a t i v e s . I t i s understood that copying or p u b l i c a t i o n of t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be allowed without my w r i t t e n permission. The U n i v e r s i t y of B r i t i s h Columbia Vane ouve r, Canada htq)://www.library.ubc.ca/spcoll/thesauth.html 19/04/00 11 A B S T R A C T This study employs molecular and field sampling techniques to understand the complex life history of the filamentous green alga, Acrosiphonia, in southern British Columbia, Canada. The D N A sequences of the nuclear ribosomal internal transcribed spacer (ITS) regions conclusively identify the unicellular green algal endophytes, 'Chlorochytrium inclusum' and 'Codiolum petrocelidis', as the alternate life history phases of one or more Acrosiphonia species. 'Chlorochytrium inclusum', a spherical unicell, was found abundantly in the foliose red alga Mazzaella splendens, whereas 'Codiolum petrocelidis', a stalked unicell, densely colonises 'Petrocelis franciscana' (= crustose tetrasporophytic phase of Mastocarpus papillauts). The D N A sequence data supported previous culture studies and Kornmann's hypothesis that the two morphologically different endophytes are alternate phenotypes of the sporophyte of a single Acrosiphonia species. The relationship of Acrosiphonia's endophytic sporophyte and filamentous, free-living gametophyte in nature revealed similar dynamics for three environmentally variable field sites. Filamentous Acrosiphonia plants were seasonally abundant (spring and summer) with fertile cells developing almost immediately after Acrosiphonia's appearance in the rocky intertidal zone. The unicellular sporophytes colonised M. splendens and 'Petrocelis' one to three months later, and showed higher tolerance to abiotic factors than Acrosiphonia's gametophyte: high summer temperatures (which correlated with death of the filamentous free-living plants) were survived and they overwintered in their hosts. Endophytes matured primarily in winter, zoospore release occurred throughout winter and spring and Acrosiphonia's life cycle is completed with subsequent zoospore germination and establishment of filamentous gametophytic plants. The two red algal hosts, M. splendens and 'Petrocelis1, were abundantly available for endophyte colonisation in spring and summer. A number of factors, e.g. herbivory, winter storms and senescence, however, were identified to produce fluctuating seasonal abundance patterns of the hosts, thus potentially affecting endophyte survival. I suggest the endophytes have evolved a strategy whereby duration in the host is synchronised with seasonality of the host. A n investigation of possible hosts for Acrosiphonia's sporophyte established a wide range of hosts. However, 'Codiolum' showed a greater affinity for 'Petrocelis' than for other crusts, and 'Chlorochytrium' colonised foliose red algae characterised by loosely compacted cells in the cortex and medulla and carrageenansand carragars as cell wall Ill constituents e.g. M. splendens, M. heterocarpa and Schizymenia pacifica, more readily than others. A bet-hedging strategy is proposed for Acrosiphonia's life history. Not only have two morphologically different phases adapted to a seasonally variable environment, but the sporophytic phase of at least one Acrosiphonia species can colonise two alternate hosts (crustose and foliose red algae), and low host specificity is evident for both endophytes. iv TABLE OF CONTENTS Page Abstract i i Table of Contents iv List of Tables vii List of Figures ix Acknowledgements xiii THESIS INTRODUCTION 1 CHAPTER 1. Identification of 'Chlorochytrium' and 'Codiolum' as the alternate phase of Acrosiphonia using ITS1 and ITS2 ribosomal D N A sequence data 11 INTRODUCTION 11 M A T E R I A L S A N D METHODS 14 Sample collection and DNA extraction 14 PCR amplification and sequencing 17 Alignment and analysis of sequences 18 RESULTS 19 DNA extraction and PCR amplification 19 DNA sequences 20 Phylogenetic reconstructions 23 DISCUSSION 23 REFERENCES 29 CHAPTER 2. Life history of Acrosiphonia in southwestern British Columbia: 33 a baseline study INTRODUCTION 32 Description of Acrosiphonia's phases in nature 33 Environmental requirements for gametophyte and sporophyte 35 M A T E R I A L S A N D METHODS 36 Field sampling 36 Data analysis 39 RESULTS 39 Acrosiphonia identification 39 Acrosiphonia distribution and seasonal abundance 43 'Chlorochytrium' morphology and location in host 49 'Chlorochytrium' seasonal abundance 50 'Chlorochytrium' size 52 'Codiolum' morphology and location in host 53 'Codiolum' seasonal abundance 56 'Codiolum' size 60 DISCUSSION 61 Acrosiphonia identification in the field 62 Acrosiphonia seasonality 63 Factors affecting Acrosiphonia seasonal abundance and distribution 64 Acrosiphonia arcta and A. coalita differences in the field 66 Endophyte colonisation 67 Endophyte growth, fertility and duration in host 68 Endophyte abundance variability 72 Selection of heteromorphy in Acrosiphonia 73 REFERENCES 76 CHAPTER 3 Host Availability for Colonisation and Survival of'Chlorochytrium' and 'Codiolum' 80 INTRODUCTION 80 MATERIALS AND METHODS 81 'Petrocelis'surface area 81 Mazzaella splendens density 82 'Chlorochytrium' survival in drift Mazzaella splendens blades 82 Data analysis 83 RESULTS 84 'Petrocelis' seasonal abundance 84 Crust dynamics and invertebrate observations 87 Health of crusts 91 Crust recruitment 92 Mazzaella splendens seasonal abundance and condition of blades 92 'Chlorochytrium' survival in drift Mazzaella splendens blades 97 DISCUSSION 97 vi 'Petrocelis'dynamics 97 Mechanism for seasonal pattern o/Mazzaella splendens abundance 99 Implications for 'Codiolum' survival and duration in host 101 Mazzaella splendens dynamics 102 Implications for 'Chlorochytrium' survival 103 REFERENCES 106 C H A P T E R 4 Host specificity of Acrosiphonia's sporophyte, 'Chlorochytrium' and 108 'Codiolum' INTRODUCTION 108 MATERIALS A N D METHODS 110 Putative host species for 'Chlorochytrium' and 'Codiolum' 110 'Chlorochytrium' affinity for life history phase 113 Data analysis 113 RESULTS 114 'Chlorochytrium' hosts 114 Identity of endophytes within foliose red algae 114 Mean densities of 'Chlorochytrium' 115 'Codiolum' hosts 117 'Chlorochytrium' colonisation of gametophytic vs. tetrasporophytic phase 117 DISCUSSION 119 REFERENCES 128 G E N E R A L CONCLUSIONS 132 Suggested future research 134 APPENDIX A : Acrosiphonia percent cover Means 137 APPENDIX B: 'Chlorochytrium'Density Means 138 APPENDIX C: 'Codiolum'Density Means 140 APPENDIX D: Surface area of individual 'Petrocelis' crusts at Burrard Inlet for one year 142 APPENDIX E: Surface area of individual 'Petrocelis' crusts at Sooke from summer 1997 to fall 1998 143 LIST OF TABLES Pag Table 1.1 Species names, collection sites and abbreviations and GenBank accession 15 numbers for isolate sequences in phylogenetic analysis. Note: host names for endophytes are given. Isolates with ambiguities have not been included. Table 2.1 Distinguishing characters for Acrosiphonia arcta and Acrosiphonia coalita in southwestern British Columbia. 42 Table 2.2 Acrosiphonia percent cover. Results of one-way A N O V A on data shown in Figures 2.4A-2.6A and Figures 2.11 A-2.13A. The significant F value indicates a statistically significant difference among two or more sampling dates. 49 Table 2.3 'Chlorochytrium' density. Results of one-way A N O V A on data shown in Figures 2.4B-2.6B. The significant F value indicates a statistically significant difference among two or more sampling dates. 52 Table 2.4 'Codiolum'density. Results of one-way A N O V A on data shown in figures 2.11B-2.13B. The significant F value indicates a statistically significant difference among two or more sampling dates. 60 Table 3.1 Seasonal differences of'Petrocelis' crust surface area. Results of independent samples f-test performed on pairs of maximum and minimum data points shown in Figures 3.1 A and 3.2A. The significant t value indicates a statistically significant difference among maximum and minimum surface area values. Note that pairs of points correspond to (1) (summer) fall and winter and (2) winter and summer. 87 Table 3.2 Seasonal condition of Mazzaella splendens thalli. 96 Table 4.1 Red algal hosts reported for 'Chlorochytrium'. Note all herbarium specimens were examined by me for the presence of 'Chlorochytrium'. 109 Table 4.2 Collection dates for putative host species for 'Chlorochytrium' at Burrard Inlet, Sooke and Bamfield and their relative abundance. A (-) indicates the species may have been present but was not collected. 112 Table 4.3. 'Chlorochytrium' presence in putative host species. Note the much lower sample sizes of all species compared to Mazzaella splendens. 1 Table 4.4 Mean densities of'Chlorochytrium'. Note the lower sample sizes of species compared to Mazzaella splendens. 1 viii Table 4.5 'Chlorochytrium' densities in tetrasporophytic and gametophytic Mazzaella splendens blades. Results of independent samples r-tests performed for Burrard Inlet, Sooke and Bamfield and all three sites combined. The significant t value (*) indicates a statistically significant difference among endophyte densities in tetrasporophytic and gametophytic blades. 118 Table 4.6 Cell wall composition of algal species examined for the presence of endophytes. Note all are members ofthe Rhodophyta, except Ralfsia pacifica, a brown alga. 125 ix LIST OF FIGURES Page Figure 0.1 Acrosiphonia coalita in the low intertidal zone at Sooke, Vancouver Island. 2 Figure 0.2 Life cycle of Acrosiphonia. 3 Figure 0.3 Location of the three study sites. 6 Figure 1.1 A summary of results from culture studies involving 'Chlorochytrium inclusum' and 'Codiolum petrocelidis'. Note that the northeast Pacific workers found 'Codiolum' to give rise to more than one Acrosiphonia species, whereas 'Chlorochytrium' only produced A. arcta. Both endophytes were shown to give rise to Spongomorpha lanosa in Europe. 13 Figure 1.2 Map showing primer positions within nuclear ribosomal D N A cistron. Primer pairs ITS 1/5; ITS4 and ITS2;ITS3 were used to amplify and sequence the ITS1, 5.8s, ITS2 and a short segment ofthe 26s gene. The pair M O l ; ITS4 was used to isolate endophyte from host D N A . 18 Figure 1.3a. Ethidium bromide stained agarose gel of PCR amplified ITS1, 5.8S, ITS2 and partial 26S regions from 'Petrocelis' crust colonised by 'Codiolum petrocelidis'. Note that the 0.5 kb fragment is 'Codiolum' and the 0.75 kb fragment is'Petrocelis'. b. Agarose gel of'Chlorochytrium'and'Codiolum' PCR product re-amplified with M O l and ITS4 primers. 20 Figure 1.4 Sequence variation among 'Chlorochytrium', 'Codiolum' and Acrosiphonia isolates. Six variable sites occur within the ITS regions of four of the five distinct genotype groups. Group 1 differs from group 2 by a single bp change at position 48; group 3 varies from groups 1 and 2 by 4 or 5 bp changes; group 4 differs from group 2 by 2 bp changes and group 5 shows 5 new variable sites in the ITS. The * 'Chlorochytrium' is the only endophyte isolate which does not match an Acrosiphonia isolate. 22 Figure 1.5 Maximum likelihood tree from ITS1 and ITS2 sequences. Monostroma arcticum was used as the outgroup. Numbers above the branches are neighbour-joining bootstrap values (1000 replicates) using a Jukes-Cantor model of nucleotide substitution; numbers below the branches are bootstrap proportions using parsimony (200 replicates). The branch to Monostroma is not proportional to its length of 0.765. Otherwise, the scale bar on the tree indicates the maximum likelihood distance. See Table 1.1 for detailed information on species, collection sites and abbreviations. 24 Figure 2.1. Establishing the timing of events of Acrosiphonia''s life history in nature. 36 X Figure 2.2 A. Acrosiphonia arcta at Burrard Inlet. B. Acrosiphonia arcta and A. coalita at Bamfield. C. E. A. coalita at Sooke. F. A. arcta at Sooke. D. A. arcta and A. coalita at Sooke. Scale bar is 5 cm. arc = A. arcta; co = A. coalita. 40 Figure 2.3 A. Acrosiphonia coalita habit. B. Acrosiphonia arcta habit. C. Fertile cells. D. Empty gametangium with pore for gamete release. E. Compound hook of A coalita. F. Simple hook. G . Many cells fertile in a series, A. arcta. H. Single cells fertile in A. coalita branches. Scale bar is 200 um. Py = pyrenoids; gm = gametangium 41 Figure 2.4 A. Acrosiphonia % cover at Burrard Inlet over two years. Data are means ± S.E. from 30 quadrats placed along transect lines. The S.E. are missing for two data points, due to missing data (theft of field notes). B . 'Chlorochytrium' densities at Burrard Inlet. Data are means ± S.E. from cells counted in 30 Mazzaella splendens blades. 44 Figure 2.5 A. Acrosiphonia % cover at Sooke over two years. Data are means ± S.E. from 30 quadrats placed along transect lines. The S.E. are missing for three data points, due to missing data (theft of field notes). B . 'Chlorochytrium' densities at Sooke. Data are means ± S.E. from cells counted in 30 Mazzaella splendens blades. 46 Figure 2.6 A. Acrosiphonia % cover at Bamfield over two years. Data are means ± S.E. from 30 quadrats placed along one or two transect lines. B. 'Chlorochytrium' densities at Bamfield. Data are means ± S.E. from cells counted in 30 Mazzaella splendens blades. 48 Figure 2.7 A. Surface view of 'Chlorochytrium inclusum' cells (80-160 um) in Mazzaella splendens. B . Transverse section showing 'Chlorochytrium' in cortex region of M. splendens. C. D. E. Vegetative 'Chlorochytrium' cells. F. 'Chlorochytrium'cell near maturity. G . 'Chlorochytrium' cell releasing > 800 zoospores. Scale bar is 50 um. py = pyrenoids. 51 Figure. 2.8 'Chlorochytrium' life history schematic based on Sooke cell size observations over two years. The solid line represents average cell size and the dotted line maximum cell size. 53 Figure 2.9 Polymorphism of 'Codiolum petrocelidis'. A. B . C. E. show cells among 'Petrocelis' filaments. Note both fertile 'Codiolum' and fertile 'Petrocelis' in A. Cells with long stalks are shown in D and E. Juvenile 'Codiolum' are represented by G and I. Large, fertile cells are shown in A. C . H. Scale bar is 50 um. Tet = tetrasporangium. 54 Figure 2.10 Polymorphism of 'Codiolum petrocelidis'. Note the large, fat fertile cells filled with zoospores in the third row. The last cell in the first row is believed to represent the youngest cells. 55 XI Figure 2.11 A. Acrosiphonia % cover at Burrard Inlet over two years. Data are means ± S.E. from 30 quadrats placed along transect lines. The S.E. are missing for three data points, due to missing data (theft of field notes). B. 'Codiolum' densities at Burrard Inlet. Data are means ± S.E. from cells counted in 30 'Petrocelis' patches. 57 Figure 2.12 A. Acrosiphonia % cover at Sooke over two years. Data are means ± S.E. from 30 quadrats placed along transect lines. The S.E. are missing for three data points, due to missing data (theft of field notes). B. 'Codiolum' densities at Sooke. Data are means ± S.E. from cells counted in 30 'Petrocelis' patches. 58 Figure 2.13 A. Acrosiphonia % cover at Bamfield over two years. Data are means ± S.E. from 30 quadrats placed along one or two transect lines. B. 'Codiolum' densities at Bamfield. Data are means ± S.E. from cells counted in 30 'Petrocelis' patches. 60 Figure 2.14 'Codiolum' life history schematic based on observations of cell size and reproductive state of cells over two years at Burrard Inlet, Sooke and Bamfield. 61 Figure 2.15 Timing of events of Acrosiphonia''s life history in nature. 62 Figure 2.16 Generalised life history schematic of Acrosiphonia''s sporophyte. * 'Chlorochytrium' colonisation occurs earlier than 'Codiolum' colonisation. 71 Figure 3.1 A. Surface area of 'Petrocelis' crust at Burrard Inlet over 1 year. Data are means ± S.E. from 26 crusts. Surface area values of zero were omitted from the calculation of the mean. B. ' Codiolum' densities at Burrard Inlet. Data are means ± S.E. from cells counted in 30 'Petrocelis' patches. 85 Figure 3.2 A. Surface area of 'Petrocelis' crust at Sooke over 18 months. Data are means ± S.E. from 30 crusts. Surface area values of zero were omitted from the calculation of the mean. B. ' Codiolum' densities at Sooke. Data are means ± S.E. from cells counted in 30 'Petrocelis' patches. 86 Figure 3.3 Surface area of 'Petrocelis' crust patches at Burrard Inlet (A) and Sooke (B). Series 1 illustrates change in surface area of crusts where values of zero were omitted. Series 2 includes zero surface area values, i.e. crusts which disappeared during the observation period, to illustrate crust survivorship. Data are means ± S.E. from 26 (A) and 30 (B) crusts. 88 Figure 3.4 A. Barnacle settlement on 'Petrocelis' at Burrard Inlet. B. Intricate crust pattern at Sooke. C. D. Limpet abundance and notable tissue loss from October to November. E. Fragmentation of single crust? F. G. H. Green, mucilaginous tissue. I. J. "Vein" pattern on Mazzaella splendens blades. 89 Figure 3.5 Time series of two 'Petrocelis' crusts at Sooke. 90 Xll Figure 3.6 A. Mazzaella splendens density ± S.E. at Burrard Inlet over two years. The point-quarter method of density estimation was based on locating 8 random points along a transect. B. 'Chlorochytrium' densities at Sooke. Data are means ± S.E. from cells counted in 30 M. splendens blades. 93 Figure 3.7 A. Mazzaella splendens density ± S.E. at Sooke over two years. The point -quarter method of density estimation was based on locating 8 random points along a transect. B. 'Chlorochytrium' densities at Sooke. Data are means + S.E. from cells counted in 30 Mazzaella splendens blades. 94 Figure 3.8 A. Mazzaella splendens density ± S.E. at Bamfield over two years. The point -quarter method of density estimation was based on locating 8 random points along a transect. B. 'Chlorochytrium' densities at Bamfield. Data are means ± S.E. from cells counted in 30 Mazzaella splendens blades. 95 Figure 4.1 'Chlorochytrium' density in Mazzaella splendens and in Schizymenia pacijica at Sooke for 1998. Data are means ± S.E. from cells counted in 30 M. splendens blades and in 2-12 S. pacifica blades. 118 Figure 4.2 A comparison of blade cross-sections: Sparlingia pertusa (A) is characterised by a medulla comprised of large, compacted cuboidal cells, whereas Mazzaella splendens (B) exhibits a medulla consisting of loosely interwoven filaments. 123 ACKNOWLEDGEMENTS I wish to acknowledge my supervisor, Robert DeWreede, for his unfailing support and guidance throughout the duration of my graduate work. Also, many thanks to my committee members, Mary Berbee and Jeannette Whitton, for their constructive suggestions and comments on this thesis. In addition, I thank Mary Berbee for the use of her laboratory for the molecular study, and for her expertise with molecular techniques. Mona Pirseyedi and Barbara Mable deserve acknowledgement for their help with the molecular work. The following people were invaluable for their help in the field, often in rain or snow, sometimes in the early hours of the morning: Tania Thenu, Helga Sussmann, Diane Orihel, Ricardo Scrosati, Tamara Allen, Reza Shahidi, Tracy Stevens, Frank and Christel Shaughnessy, Sherri Rodgers and CarolAnn Borden. Diane Orihel assisted me with the tedious task of counting endophytes in red algal blades. This thesis was primarily funded by operating grants from the Natural Sciences and Engineering Research Council of Canada to Robert DeWreede and to Mary Berbee. I am also grateful for the NSERC scholarships and University Graduate Fellowship I was awarded as a Ph.D. candidate. Finally, special thanks to my dear friends and peers, Tania Thenu, Tamara Allen and Tracy Stevens, for their steady support, encouragement, discussions both academic and not and many laughs along the way. THESIS INTRODUCTION 1 The genus Acrosiphonia was established by J. G. Agardh in 1846 for a group of marine green algae composed of branched uniseriate filaments, each cell having a reticulate chloroplast. Filaments of plants tend to be bound together by hooked branchlets or by rhizoids to form large tangled masses, and are abundant on rocks in the low to mid intertidal zone (Fig. 0.1). Difficulty in separating the genus from Spongomorpha has long existed (Wille 1900, Collins 1909, Kornmann 1965, 1970a; Scagel 1966, Jonsson 1991). Agardh (1846) originally delimited the two genera on the basis of cell length to width ratios. Later Wille (1900) suggested that multinucleate plants be called Acrosiphonia and uninucleate ones Spongomorpha. Since culture studies of Acrosiphonia revealed extreme polymorphism in response to temperature and light intensity, Kornmann (1965, 1970a) proposed that reproductive behaviour rather than vegetative characteristics such as length of cells be used for separation ofthe two genera. Based on his studies, Acrosiphonia was characterised by vegetative reproduction and Spongomorpha by a heteromorphic life history. Jonsson (1959a, 1962), however, claimed that Acrosiphonia in France is heteromorphic, and pointed out that Kommann's failure to obtain 'Codiolum' from zygotes of A. arcta (Dillwyn) J.G. Agardh [= Acrosiphonia spinescens (Kutzing) Kjellman according to Hudson (1974) and Jonsson (1986)] may have been due to a complete failure of karyogamy (Jonsson reported that karyogamy in A. arcta is facultative). In the northeast Pacific culture studies (Hollenberg 1958, Fan 1959, Chihara 1969, Hudson 1974) also suggested a heteromorphic life history for Acrosiphonia. Both Kornmann (1962a) and Jonsson (1959a) agreed that the presence of opercula on the gametangia is a distinctive feature of Acrosiphonia. Scagel (1966) questioned the value of Wille's separation into uninucleate and multinucleate genera and Collins (1909) believed that such a delimitation de-emphasised the similarities in thallus structure, other cytological characteristics and reproductive behaviour. Nonetheless, despite these criticisms, Wille's proposal is widely accepted by taxonomists today. Spongomorpha is absent from the northeast Pacific, having only been reported in Scandinavia, Iceland, the Arctic Ocean, Germany and France. There is also a lack of agreement on species delimitation of Acrosiphonia. Scagel (1966) reported five species of Acrosiphonia present in northern Washington and British Columbia. However, Hudson's (1974) intensive culture and taxonomic studies from the Puget Sound region 2 Figure 0.1 Acrosiphonia coalita in the low intertidal zone at Sooke, Vancouver Island. suggest that only two well-defined entities, A. coalita (Ruprecht) Scagel, Garbary, Golden & Hawkes and A. arcta are represented on our coast. Acrosiphonia coalita appears to be present only in the northeast Pacific, whereas A. arcta has been recorded in both the northern and southern hemispheres, including the Northwest Atlantic, the northeast Pacific, the North Sea, Greenland, southern Chile and Antarctica. Other Acrosiphonia species, apparently not present in North America, have been identified occurring in Japan, Germany, France, Scandinavia and Greenland. Criteria commonly used to distinguish species of Acrosiphonia are primarily vegetative characteristics (Setchell & Gardner 1920, Scagel 1966), many of which are variable and responsive to environmental conditions (Kornmann 1965, 1970a; Hudson 1974). In addition to taxonomic uncertainties, the life history of Acrosiphonia is poorly understood. Extensive culture studies conducted primarily in the 1960s and 1970s (Hollenberg 1958, Fan 1959, Jonsson 1959a, 1959b, 1962,1963,1966,1970; Kornmann 1961a, 1964,1970a, 1972; Chihara 1969, Hudson 1974, Miyaji & Kurogi 1976, Miyaji 1984,1996) implicated 3 unicellular green endophytes of the Chlorococcales in the life histories oi Acrosiphonia and Spongomorpha. Endophytic algae are those which are found within the tissues or cells of other algae / plants, but do not elicit symptoms of disease (Wilson 1995). Chlorochytrium inclusum Kjellman, endophytic in foliose red algae and Codiolum petrocelidis Kuckuck, found within red algal crusts, are suspected to be the sporophytic alternate life history phases of the gametophytic Acrosiphonia. (Fig. 0.2). For this reason, all reference to the Codiolum and Chlorochytrium phases appears in single quotes in this thesis. • Fusion Zygote Gametes Gametophyte Sporophyte Red Algal Hosts Acrosiphonia Phase Off 'Chlorochytrium' Phase 'Codiolum Phase Zoospore Meiosis Figure 0.2 Life cycle of Acrosiphonia. 4 Several alternative life histories (other than a heteromorphic alternation of generations) have been reported for Acrosiphonia in culture. In A. grandis Kjellman (type locality Norway) the zygote develops into a 'Codiolum' or 'Chlorochytrium' phase which grows directly into a filamentous plant without formation of zoospores (Kornmann 1970b). Meiosis has not been observed, but Kornmann (1970a) suggested it might occur within the 'Codiolum' vesicle. In A. sonderi (Kiitzing) Kornmann (type locality Helgoland, Germany) and A. arcta filamentous plants of unknown ploidy were found to recycle by means of biflagellate zoospores, sexual reproduction not having been observed (Kornmann 1962a). Filamentous plants of unknown ploidy of A. arcta were also found to recycle by means of isogametes (Kornmann 1962a). Meiosis was not observed, but was suggested to occur during germination of the zygote or possibly in gametogenesis. No evidence exists that any of these alternate life histories occurs in nature. 'Chlorochytrium inclusum' is a spherical to subspherical unicell, generally found among the cells of the cortex of foliose red algal hosts such as Schizymenia, Constantinea, Mazzaella, Neodilsea and Dilsea (Scagel 1966, Chihara 1969, Hudson 1974). It is reported to measure 80 -100 um in diameter at maturity and contains one parietal chloroplast with many pyrenoids. 'Codiolum petrocelidis', on the other hand, is differentiated into an ovoid vesicle with a colourless stalk, and is commonly found embedded in the filamentous system of the red algal crust 'Petrocelis' [= tetrasporophytic phase of Mastocarpus (Hollenberg 1958, Fan 1959, Jonsson 1959a)]. The cell contains a single dense chloroplast and many pyrenoids, and is 125-175 um in length and 25-50 u.m in diameter in the pigmented region. Jonsson (1959a, 1966) and Kornmann (1961a, 1964) observed that when these morphologically different endophytes develop in culture, free of their host, 'Chlorochytrium' cells retain a stalk characteristic of'Codiolum'. Kornmann (1964) thus proposed that phenotypic responses, dictated by the different nature of the two hosts, accounted for the differences in morphology of'Codiolum' and 'Chlorochytrium', i.e. the morphologically distinct endophytes represent the sporophytic phase ofthe same organism. 'Chlorochytrium' / 'Codiolum' phases are not, however, restricted to Acrosiphonia and Spongomorpha. Representatives of the genera Urospora, Ulothrix and Monostroma (present in southwestern British Columbia) also possess a unicellular sporophytic phase (Kornmann 1961b, 1962b, 1963) that develops similarly to Acrosiphonia- and Spongomorpha-reiated 'Codiolum' phase (hence all are members ofthe Codiolales sensu van den Hoek et al. 1995). Although the 5 past culture studies suggest that 'Chlorochytrium inclusum' and 'Codiolum petrocelidis1 are likely to be a part of the heteromorphic life cycle of Acrosiphonia, verification is needed for which genera are most closely related to the endophytes present within an area. The focus of my thesis is on understanding the complex life history Acrosiphonia in southern British Columbia. It is comprised of two major studies: one utilising molecular work to determine whether endophytes present in an area are an alternate life history phase of Acrosiphonia, and the other employing field sampling techniques to investigate the natural dynamics oi Acrosiphonia along the coast of British Columbia . The initial objective of my research was to establish the identity of'Chlorochytrium' and 'Codiolum' cells found in Mazzaella splendens (Setchell & Gardner) Fredericq blades and in the crust 'Petrocelis franciscana' Setchell & Gardner [= tetrasporophytic phase of Mastocarpus papillatus (Agardh) Kutzing], respectively from Vancouver Island and Burrard Inlet. Rather than culturing the endophytes (in the past fraught with poor survival rates and the inability to reproduce natural conditions resulting in morphological or reproductive ambiguities in plants), molecular studies were chosen to investigate whether the endophytes and Acrosiphonia constitute alternating phases in the life history of a single alga. The study in Chapter 1 uses D N A sequence data from the ITS regions of the nuclear ribosomal D N A cistron to provide a phylogenetic comparison of the endophytes, 'Chlorochytrium' and 'Codiolum', and free-living genera of the Codiolales found in the region. Pairwise sequence comparisons and a proposed phylogeny of relationships among endophytes and free-living Codiolales genera establish that the endophytes are more closely associated with Acrosiphonia than with any of the other genera, i.e. the endophytes are identified as the sporophytic phase of Acrosiphonia. In addition, the study supports Kornmann's hypothesis that 'Chlorochytrium' and 'Codiolum' are alternate phenotypes of the sporophyte of a single Acrosiphonia species. Chapters 2 to 4 focus on understanding the natural dynamics of the life history of Acrosiphonia elucidated in Chapter 1. More specifically, gametophyte / sporophyte and host / endophyte associations, as well as the timing of life history events in nature are investigated at three different sites in southern British Columbia: Burrard Inlet (Brockton Point) 49° 34' N , 123° 10' W in Vancouver on mainland British Columbia, and Sooke (Whiffin Spit) 47° 42' N , 123° 48' W and Bamfield (Prasiola Point) 48° 49' N , 125° 10' W on Vancouver Island (Fig. 0.3). The sites selected vary in environmental factors such as wave exposure (Burrard Inlet is relatively sheltered, Bamfield wave-exposed and Sooke of intermediate wave-exposure), 6 Figure 0.2 Location ofthe three study sites. occurrence of summer low tides (afternoons at Burrard Inlet; early morning on Vancouver Island sites), salinity, temperature and precipitation. Burrard Inlet also differs in that human disturbance is common and can be quite destructive, e.g. people overturning rocks and digging in the sand for marine organisms. The effect of environmental conditions on the natural dynamics of Acrosiphonia is addressed. Acrosiphonia was found growing abundantly on boulders or epiphytic on Fucus spp. or seagrass in the low to mid intertidal zone at all three sites, where it is the dominant filamentous green alga in spring and summer. At Burrard Inlet only one morphological species, Acrosiphonia arcta was present, whereas at Sooke and Bamfield the two morphological species, A. arcta and A coalita comprised the majority oi Acrosiphonia plants. The algal communities of the three sites were observed to be dominated by Fucus spp., Ulva spp., Enteromorpha spp., Mastocarpus papillatus and a number of filamentous red algae such as Polysiphonia spp. and Microcladia spp. in the mid intertidal zone and by Mazzaella splendens and kelps in the low intertidal zone. Generally a greater diversity of algal species were found at Sooke and Bamfield than at Burrard Inlet; species absent at Burrard Inlet included Halosaccion glandiforme (Gmelin) Ruprecht, Mazzaella heterocarpa (Postels & Ruprecht) Hommersand , Schizymenia pacifica (Kylin) Kylin, Hedophyllum sessile (Agardh) Setchell, Cladophora spp. and Palmaria mollis (Setchell & Gardner) van der Meer & Bird (= Rhodymenia palmata var. mollis Setchell & Gardner). Constantinea subulifera Setchell was common only at Burrard Inlet. Urospora sp., Ulothrix sp. and Monostroma sp. of the Codiolales were identified from Sooke and Bamfield. The encrusting algae, 'Petrocelis franciscana', Hildenbrandia occidentalis Setchell and Ralfsia pacifica Hollenberg coexisted throughout the intertidal zone at all three sites. 'Chlorochytrium inclusum' and 'Codiolum petrocelidis' had already been observed within M. splendens blades from Sooke and 'Petrocelis' crusts from Sooke and Burrard Inlet, respectively, prior to the field study. It was not known i f the endophytes were present in other algae. At the Sooke and Bamfield study sites the gastropods, Thais spp. and Tectura spp., and the chiton, Katharina tunicata Wood, were observed to be the most abundant invertebrates. Littorinid snails of the genus Littorina were found on Mazzaella splendens blades at Burrard Inlet and Bamfield. At Burrard Inlet the predominant invertebrates were the mussels, Mytilus trossulus Gould, and barnacles (the dominant species being Balanus glandula Darwin); sometimes, especially in the spring, so abundant that dense 'mats' of juvenile mussels and barnacles covered boulders. Limpets present at Burrard Inlet and Sooke were primarily Tectura scutum Rathke. The studies detailed in Chapters 2 to 4 were conducted at the three sites described, Burrard Inlet, Sooke and Bamfield, and utilised field sampling techniques. The main objectives of Chapter 2 are to identify the filamentous gametophytes and endophytic sporophytes of 8 Acrosiphonia in nature (DNA sequences from Chapter 1 aid in the identification), and to establish the relationship of the life history phases, primarily the timing of life history events, in southwestern British Columbia. This is the first study to examine the relationship of Acrosiphonia's sporophyte and gametophyte in nature. I also briefly speculate on the factors responsible for selection of this complex heteromorphic life history. Host availability for colonisation and survival of Acrosiphonia's sporophytic phase (the endophytes, 'Chlorochytrium' and 'Codiolum') is addressed in Chapter 3. The two red algal hosts, Mazzaella splendens and 'Petrocelis franciscana1, were abundant at my three study sites, and thus seemed to represent ideal hosts for 'Chlorochytrium' and 'Codiolum', respectively. Seasonal abundance of M. splendens and 'Petrocelis' was determined (no study has examined seasonality of'Petrocelis' on boulder strewn shores), and 'Chlorochytrium' and 'Codiolum' survival, as well as endophyte duration in host, assessed in relation to availability ofthe hosts. Chapter 4 investigates the degree of specificity of the host / endophyte relationship for Acrosiphonia's sporophytic phase. A l l foliose Rhodophyte species and crustose species found in the intertidal zone at Burrard Inlet, Sooke and Bamfield were examined to compare endophyte densities. A number of mechanisms and strategies are discussed for the relatively low host specificity, but differential colonisation of hosts, observed in this study. These include the role of structural and chemical characters of the host, and a bet-hedging strategy for Acrosiphonia's sporophytic life history phase. 9 REFERENCES Agardh, J. G. 1846. Anadema, ett nytt slagtet bland Algerne. K. Sv. Vet.-Akad. Handl. pp. 1-16. Stockholm. Chihara, M . 1969. Culture study of Chlorochytrium inclusum from the northeast Pacific. Phycologia 8:127-33. Collins, F .S . 1909. The green algae of North America. Tufts College Studies (Sci. Ser.) 2:79-480. Fan, K . C . 1959. Studies on the life histories of marine algae I. Codiolum petrocelidis and Spongomorpha coalita. Bull. Torrey Bot. Club 86:1-12. Hollenberg, G. J. 1958. Observations concerning the life cycle of Spongomorpha coalita (Ruprecht) Collins. Madrono 14:249-52. Hudson, M . L. 1974. Field, culture and ultrastructural studies on the marine green alga Acrosiphonia in the Puget Sound region. Ph.D. Thesis, University of Washington, 206 pp. Jonsson, S. 1959a. L'existence de l'alternance heteromorphe de generations entre VAcrosiphonia spinescens et le Codiolum petrocelidis. C. R. Acad. Sci., Paris. 248:835-7. 1959b. Le cycle de developpement du Spongomorpha lanosa (Roth) Kiitz. et la nouvelle famille des Acrosiphoniacees. Ibid. 248:1565-7. 1962. Recherches sur des Cladophoracees marines: structure, reproduction, cycles compares, consequences systematiques. Ann. Sci. Nat. Bot. 12:25-230. 1963. Sur quelques variations du cycle de developpement dans la famille des Acrosiphoniacees. C. R. Acad. Sci., Paris 256:5187-9. 1966. Sur l'identification du sporophyte du Spongomorpha lanosa. Ibid. 262:626-9. 1970. Localisation de la meiose dans le cycle de VAcrosiphonia spinescens (Kiitz) Kjell. (Acrosiphoniacees). Ibid. 271:1859-61. 1986. Isolement, en culture, de lignees haploides a partir du cycle heteromorphe chez le Spongomorpha aeruginosa. Cryptogamie, Algologie 7:149-59. 1991. Typification of Spongomorpha (Kiitz) Wille and Acrosiphonia (Agardh) Wille (Acrosiphoniales, Chlorophyta). Ibid. 12:171-81. Kornmann, P. 1961a. Uber Spongomorpha lanosa und ihre Sporophytenformen. Helg. Wiss. Meeresunters. 7:195-205. 10 1961b. Uber Codiolum und Urospora. Ibid. 8:42-57. 1962a. Die Entwicklung von Monostroma greveillei. Ibid. 8:195-202. 1962b. Eine Revision der Gattung Acrosiphonia. Ibid. 8:219-42. 1963. Die Lebenszyklus einer marinen Ulothrix-Art. Ibid. 8:357-60. 1964. Zur Biologie von Spongomorpha aeruginosa (Linnaeus) van den Hoek. Ibid. 11:200-8. 1965. Zur Wachstum und Aufbaus von Acrosiphonia. Ibid. 12:219-38. 1970a. Phylogenetische Beziehungen in der Griinalgengattung Acrosiphonia. Ibid. 21:292-304. 1970b. Der Lebenszyklus von Acrosiphonia grandis (Acrosiphoniales, Chlorophyta). Mar. Biol. 7:324-31. 1972. Les sporophytes vivant en endophyte de quelques Acrosiphoniacees et leurs rapports biologiques et taxonomiques. Soc. Bot. Fr., Memoires 75-86. Miyaji, K . 1984. The life history study on a sporophytic form of Spongomorpha-Acrosiphonia complex (Acrosiphoniales, Chlorophyta), Codiolum petrocelidis Kuckuck from Northern Japan. Jpn. J. Phycol. 32:29-36. Miyaji, K . 1996. Taxonomic study of the genus Spongomorpha (Acrosiphoniales, Chlorophyta) in Japan. Phycol. Research 44:27-36. Miyaji, K. & Kurogi, M . 1976. On the development of zoospores of the green alga Chlorochytrium inclusum from eastern Hokkaido. Bull. Jap. Soc. Phycol. 24:121-9. Scagel, R. F. 1966. Marine algae of British Columbia & northern Washington, Part I: Chlorophyceae. Nat. Mus. Can. Bull. No. 207, Biol. Ser. No. 74, 257 pp. Setchell, W. A. & Gardner, N . L. 1920. The marine algae of the Pacific Coast of North America. Part II. Chlorophyceae. Univ. Calif. Publ. Bot. pp. 149-50. van den Hoek, C , Mann, D. G. & Jahns, H . M . 1995. Algae. An Introduction to Phycology. Cambridge University Press, Cambridge, 623 pp. Wille, N . 1900. Die Zellkerne bei Acrosiphonia (J. Ag.) Kjellm. Botanisches Centralblatt 8:1238-9. Wilson, D. 1995. Endophyte - the evolution of a term, and clarification of its use and definition. Oikos 73:274-6. 11 C H A P T E R 1 I D E N T I F I C A T I O N OF C H L O R O C H Y T R I U M ' A N D ' C O D I O L U M ' AS T H E A L T E R N A T E P H A S E OF ACROSIPHONIA USING ITS1 A N D ITS2 R I B O S O M A L D N A S E Q U E N C E D A T A I N T R O D U C T I O N The morphologically different endophytes, 'Chlorochytrium inclusum' and 'Codiolum petrocelidis' from southwestern British Columbia were suspected to be the sporophytic alternate life history phases of free-living gametophytic algae in the Codiolales (sensu van den Hoek et al. 1995). In this study D N A sequence data were used to investigate whether these endophytes are the sporophytes of species belonging to Acrosiphonia. Evidence for the implication of'Chlorochytrium' and 'Codiolum' cells in the life histories of Acrosiphonia and Spongomorpha was established through culture studies conducted largely in the 1960s and 1970s. For example, in Europe, Kornmann (1961a, 1964) demonstrated that both 'C. inclusum' and 'C. petrocelidis' give rise to Spongomorpha lanosa (Roth.) Kutzing [= Spongomorpha aeruginosa (L.) Hoek] in culture. This, and the similar appearance ofthe two endophyte types free-living in culture, led Kornmann (1964) to propose that the two morphologically distinct endophytes are alternate phenotypes of the sporophyte of a single Acrosiphonia or Spongomorpha species, i.e. the morphological variation is solely attributed to the different nature of the two hosts. Jonsson (1959b, 1962,1966) showed zoospores of 'C. inclusum' produce S. lanosa, and zoospores of 'C. petrocelidis' give rise to Acrosiphonia arcta. Spongomorpha is not found in the northeast Pacific, but workers along the coasts of Washington and California have shown that 'C. inclusum' produces A. arcta in culture (Chihara 1969, Hudson 1974), while 'C. petrocelidis' gives rise to both A arcta and Acrosiphonia coalita (Hollenberg 1958, Fan 1959, Hudson 1974). Similarly, in Japan, culture work by Miyaji and Kurogi (1976) and Miyaji (1984, 1996) revealed that 'C. inclusum' and 'C. petrocelidis' maybe the sporophytic phase of several Acrosiphonia species. Figure 1.1 provides a summary of the still rather confused picture of the associations between Acrosiphonia I Spongomorpha and 'Codiolum' / 'Chlorochytrium' that have been established through culture studies. The 'Codiolum' phases are not, however, restricted to Acrosiphonia and Spongomorpha. Representatives ofthe genera Urospora, Ulothrix and Monostroma also possess a unicellular sporophytic 'Codiolum' phase (Kornmann 1961b, 1962,1963) that develops similarly to the 12 Acrosiphonia and Spongomorpha related 'Codiolum' phase. Codiolum gregarium Braun, a free-living macroscopic (l-2mm at maturity) unicell, has been associated with the life history of Urospora (Hanic, 1965). Kornmann (1973) even suggested a new class, Codiolophyceae, to unite all genera with a 'Codiolum' phase. Although it is possible that one or more of the 'Codiolum' phases of Urospora, Ulothrix and Monostroma invades red algal crusts and/or blades, none is known to exist endophytically. While the culture work described suggests that 'Codiolum petrocelidis' and 'Chlorochytrium inclusum' are likely to be a part of the life cycle of one or more species of Acrosiphonia, it would be helpful to establish genotypic relatedness via phylogenetic analysis to verify which genera are most closely related to the endophytes present within an area. Being multinucleate, the genus Acrosiphonia is clearly separated from uninucleate Spongomorpha (Wille, 1900). However, there is a lack of agreement on species boundaries within Acrosiphonia. Criteria commonly used to distinguish species of Acrosiphonia are primarily vegetative characteristics such as diameter of the filaments, length to width ratios of the cells, presence or absence of simple or compound hooks, branching patterns of the plants, shape of the tip cells and number of fertile gametangia cells in a series (Setchell and Gardner 1920, Scagel 1966). Many of these characteristics are very variable when environmental factors such as light intensity, day length and temperature differ (Kornmann 1965, 1970, Hudson 1974), making species delimitation difficult. Due to the taxonomic uncertainty within the genus Acrosiphonia, the focus of this study is to establish whether the endophytes found in the northwestern Pacific region are more closely related to Acrosiphonia than to one of the other genera of Codiolales found in the region, rather than to predict which species oi Acrosiphonia most likely represent the alternate phase of the endophyte life cycle. A phylogenetic comparison of the sporophytic endophytes 'Chlorochytrium' and 'Codiolum' and the gametophytes in the Codiolales, using sequences from the ITS regions of the nuclear ribosomal D N A cistron is provided in this study. Unlike the 18S, 5.8S and 26S ribosomal genes, which have been strongly conserved through time, the ITS regions (ITS1 and ITS2) provide a level of sequence variation suitable for comparisons at the genus and species level in algae (Peters et al. 1997, Burkhardt and Peters 1998, Blomster et al. 1999, Serrao et al. 1999) and more specifically in green algae (Bakker et al. 1995, van Oppen et al. 1993, Pillmann et al. 1997). Based on pairwise sequence comparisons and a proposed phylogeny of relationships among endophytes and free-living genera within the 13 1999, Serrao et al. 1999) and more specifically in green algae (Bakker et al. 1995, van Oppen et al. 1993, Pillmann et al. 1997). Based on pairwise sequence comparisons and a proposed phylogeny of relationships among endophytes and free-living genera within the Spongomorpha lanosa Acrosiphonia spp. • a * o '5" s T- ™ — -5 'Chlorochytrium inclusum' (endophytic in foliose reds) Kornmann (1961a,1964) Germany t Morphological variations ? t 'Codiolum petrocelidis' endophytic in red crusts CO cn O) t » a -Acrosiphonia spp. Hollenberg(1958) Fan (1959), Hudt California, Washington Acrosiphonia arcta Acrosiphonia coalita Figure 1.1 A summary of results from culture studies involving 'Chlorochytrium inclusum' and 'Codiolum petrocelidis'. Note that the northeast Pacific workers found 'Codiolum' to give rise to more than one Acrosiphonia species, whereas 'Chlorochytrium' only produced A. arcta. Both endophytes were shown to give rise to Spongomorpha lanosa in Europe. Codiolales, this study 1) identifies the two morphologically different endophytes as the sporophytic phase of the free-living green algal genus Acrosiphonia, (2) supports Kornmann's hypothesis that 'Chlorochytrium' and 'Codiolum' are alternate phenotypes ofthe sporophyte of a 14 M A T E R I A L S A N D M E T H O D S Sample collection and DNA extraction. Mazzaella splendens blades, 'Petrocelis' crust and Acrosiphonia plants were collected at Sooke (Whiffin Spit) and Bamfield (Prasiola Point) on Vancouver Island, and Burrard Inlet (Brockton Point) in Vancouver on mainland British Columbia. Acrosiphonia species were collected in the spring of 1997 and 1998 when the plants had just appeared and were relatively free of epiphytes. Initial attempts to extract D N A from Acrosiphonia collected in the summer were unsuccessful due to interference from heavy epiphytisation. Van Oppen et al. (1993) and van Oppen (1995) used only cultured material, considering Acrosiphonia from the field too difficult for use in extractions. Mazzaella splendens blades and 'Petrocelis' crusts were collected in the summer of 1997 and 1998, when endophyte density tended to be highest (A. V . Sussmann, personal observation). In total, six 'Codiolum' isolates (two from each site), 21 'Chlorochytrium' isolates (representatives from all sites), and 34 Acrosiphonia isolates (also from all sites) were sequenced in this study. A l l 'Codiolum' isolates sequenced were from 1997 collections, whereas 'Chlorochytrium' and Acrosiphonia isolates from 1997 (3 and 19, respectively) and 1998 (18 and 14, respectively) were sequenced. Two additional sequences were obtained from Acrosiphonia isolates which were collected at Amalga Harbor 58° 29' N , 134° 47' W, Alaska and St. Lawrence Island 63° 48' N , 171° 43' W, Alaska by S. C. Lindstrom. Isolates from a particular site were assigned a number, and collection location is indicated by an abbreviated site name (Table 1.1). Algal host specimens were refrigerated for less than 24 h, and examined for the presence of endophytic green algal unicells. Acrosiphonia specimens (approximately 0.1 g from each plant) collected in 1997 were lyophilised in 1.5 mL microcentrifuge tubes. The D N A from Acrosiphonia plants which were collected in 1998 was obtained from herbarium specimens, and the Alaskan isolates had been dried and stored in silica gel. Different protocols were initially used to extract D N A from 'Chlorochytrium' and 'Codiolum' cells due to the different nature of the host-endophyte associations. Acrosiphonia D N A extraction followed the protocol used for 'Codiolum'. 'Chlorochytrium' cells. In 1997 between 30 and 100 'Chlorochytrium' cells were dissected from the Mazzaella splendens blades and transferred to 100 u,L of 5% Chelex® 100 [Biotechnology grade, Bio-Rad™, Hercules, C A (Walsh et al. 1991)] w/v in d H 2 0 with a drawn-out pipette. Centrifugation at 16,500 x g for ca. 30 s pelleted the cells. The pellet was ground 15 TABLE 1.1 Species names, collection sites and abbreviations and GenBank accession numbers for isolate sequences in phylogenetic analysis. Note: host names for endophytes are given. Isolates with ambiguities have not been included. Species Host Collection site3 GenBank accession no.b 'Chlorochytrium inclusum' Mazzaella splendens Sooke, B .C. (Ski) AFO19256 M. splendens Sooke, B.C. (Sk3/Sk4) AF047681 M. splendens Sooke, B .C. (Sk5) AF161595 M. splendens Sooke, B .C. Sk6-Sk9)) AFO19255 M. splendens Bamfield, B .C. (Bml) AF047681 M. splendens Burrard Inlet, B.C. (BI1-BI4) AF019255 M. splendens Bamfield, B .C. (Bm2-Bm4) AF019255 'Codiolum petrocelidis' 'Petrocelis franciscana' Sooke, B.C. (Skl/Sk2) AFO19256 'P. franciscana' Bamfield, B .C. (Bml) AF047682 'P. franciscana' Bamfield, B .C . (Bm2) AF047681 'P. franciscana' Burrard Inlet, B .C. (BI1/BI2) AFO19256 Acrosiphonia coalita Monterey Bay, California (MB) Friday Harbor, Washington (FH) Bamfield, B .C . (Bml-Bm5) Sooke, B.C. (Skl-Sk3) AF047682 AF047682 Acrosiphonia arcta Helgoland, Germany (Hg) Roscoff, Brittany, France (Ro) Faeroe Islands (Fa) Halifax, Nova Scotia, Canada (Ha) Grotta, Iceland (Ic) Disko Island, Greenland (Di) Palmer Station, Antarctica (PS) King George Isl., Antarctica (KG) Friday Harbor, Washington (FH) Puerto Williams, Southern Chile (SC) Burrard Inlet, B.C. (BI1-BI10) AF019256 Sooke, B .C. (Skl-Sk6) AFO19256 Sooke, B .C. (Sk7-Sk9) AF047681 Bamfield, B .C. (Bml-Bm4) AF047681 Acrosiphonia sonderi Helgoland, Germany (Hg) Disko Island, Greenland (Di) Acrosiphonia sp. Amalga Harbor, Alaska (AH) AF019255 St. Lawrence Island, Bering Sea (BS) AFO19255 16 TABLE 1.1 continued Acrosiphonia sp. Spongomorpha lanosa Urospora penicilliformis (Roth) Areschoug Sooke, B.C. (Sk) Grotta, Iceland (Ic) Helgoland, Germany (Hg) Disko Island, Greenland (Di) Otago Peninsula, New Zealand (Ot) Dunedin, New Zealand (Du) King George Isl., Antarctica (KG) King George Isl., Antarctica (KG) Spitsbergen (Sp) AF161595 Ulothrix implexa (Kiitzing) Kiitzing Monostroma arcticum sensu Bliding a Abbreviated site names and isolate numbers assigned for a particular site appear in brackets. b Sequences without accession numbers are all from van Oppen (1995); all those with accession numbers are new sequences. with a sterile pestle for approximately 10 s or until green colouration indicated lysis. Samples were heated to 95-100° C for 15 min and frozen at -20° C. The samples were thawed, centrifuged again at 16,500 x g for 5 min, and then a 1:2 dilution of Chelex extract in dH 2 0 was used as template for PCR reactions. In 1998 the D N A extraction protocol used for 'Codiolum' was successfully applied to 'Chlorochytrium'. Several 10-25 mm 2 pieces of M. splendens colonised by 'Chlorochytrium' were lyophilised and ground up. Grinding endophyte and host tissue was considerably faster than dissecting individual 'Chlorochytrium' cells from M. splendens. The PCR product was, however, always inferior, resulting in weaker signals from sequencing results. 'Codiolum' cells and Acrosiphonia. Due to the obvious difficulty in dissecting 'Codiolum' cells from filaments of the 'Petrocelis' crust, host tissue colonised by endophytic cells was used for D N A extraction. Small patches (approximately 25 mm ) of crust were lyophilised and then ground to a fine consistency in sterile sand with a Teflon pestle-drill assembly, and 800 uL C T A B extraction buffer [0.7 M NaCl, 1% CTAB(w/v), 50 m M Tris (8.0 pH), 10 m M EDTA, 1% p-mercaptoethanol (v/v)] added. Samples were then incubated at 65° C for 1 h and extracted twice with one volume of chloroform. The D N A was precipitated with isopropanol and centrifuged at 16,500 x g for 10 min. The D N A pellet was washed with ice-cold 70% ethanol, 17 re-suspended in 100-150 uL TE buffer and frozen at -20° C. A 1:2 dilution was performed prior to PCR amplification. For a more detailed account of this protocol refer to McDermottetal. (1989). PCR amplification and sequencing. PCR amplifications were performed in a Perkin-Elmer™ D N A thermal cycler. The initial denaturation step (95° C for 2 min) was followed by 30 cycles with a reaction profile of 1 min at 95° C, 1 min at 48° C, and 45 s at 72° C. With each successive cycle the polymerisation step was increased by 4 s. The terminal extension step was 7 min at 72° C. For each amplification 25 uL reactions were comprised of 12.5 uX of the diluted genomic DNA, 0.2 m M dNTP mixture, 2.5 uL lOx reaction buffer (Ultratherm™, Bio/Can Scientific Inc., Canada), 1.5 m M MgCl 2 ,2.5 uL 50% glycerol, 0.9 uL dH 2 0 , 5 units Taq Polymerase (Ultratherm) and 0.5 uM of each primer. The primer pairs, ITS1 and ITS4, or ITS5 and ITS4 (White et al. 1990) were used to amplify the ITS1 and ITS2 regions and the intervening 5.8S rDNA, as well as a short segment of the 26S gene. PCR products were analysed by electrophoresis of 5 p i of reaction mixture through 1-2% agarose and visualised by ethidium bromide staining. To target only Acrosiphonia sequences, an Acrosiphonia-specific primer, M O l , designed by van Oppen et al. (1993) was used in a second PCR reaction for 'Chlorochytrium' and 'Codiolum' DNA. This primer is located within the ITS1 region, 45 bp from the 5' end (Fig. 1.2), and has the following sequence (5* to 3'): A G G T C T G A C T T G T T G G G C G G C . In the case of the 'Chlorochytrium' D N A , the primary PCR product was diluted 100 fold and then re-amplified with the primer pair M O l and ITS4 using the above reaction conditions but with an annealing temperature of 53° C. For the 'Codiolum' DNA, bands of appropriate length (as compared with Acrosiphonia arcta) were purified from the primary PCR products by electrophoresis through 1.5% agarose and ethidium bromide staining. D N A was recovered from agarose gel blocks by incubation at 65° C for 15 min in 100 uL TE. After a 100 fold dilution, re-amplification with primers M O l and ITS4 was performed as above with an annealing temperature of 56° C. PCR products used as templates for sequencing in 1997 were gel purified according to Qian and Wilkinson (1991). In 1998 purification of PCR products was accomplished with an ethanol spin precipitation following Applied Biosystems Inc. instructions. The D N A nucleotide sequence of the ITS1 (partial sequence due to M O l ' s position), ITS2 and 5.8S gene regions was determined using the AmpliTaq® Dye Terminator Cycle Sequencing kit and the 373 D N A 18 ITS 1/5 Nuclear Small rDNA (18s) ITS 3. — • 5.8s rDNA ITS 1 ITS 2 ITS 2 Nuclear Large rDNA (26s) ITS 4 Figure 1.2 Map showing primer positions within nuclear ribosomal D N A cistron. Primer pairs ITS1(5) / ITS4 and ITS2 / ITS3 were used to amplify and sequence the ITS1, 5.8s, ITS2 and a short segment of the 26s gene. The pair M O l / ITS4 was used to isolate endophyte from host DNA. Sequencer (Applied Biosystems Inc.). The primer pairs M O l / ITS4 (for the endophyte sequences), ITS1 / ITS4 (for Acrosiphonia sequences) and ITS2 / ITS3 (Fig. 1.2), were used so that this entire region could be sequenced from both ends. Alignment and analysis of sequences. Clustal V (Higgins et al. 1992) was used to align the 'Chlorochytrium1, 'Codiolum' and Acrosiphonia sequences to a previously published alignment of Acrosiphonia species from other geographic locations, and species of Urospora, Ulothrix and Monostroma (van Oppen, 1995; see Table 1.1 for species names and collection sites). Three kinds of phylogenetic analyses were conducted using P A U P * (Swofford, 1998). Monostroma was shown as the outgroup because it differed from the other species at about 44% ofthe sites in the alignment (without correcting for multiple hits). Other pairs of sequences in the data set all differed at fewer than 10% of the sites in the alignment. Ten heuristic searches, employing the tree bisection reconnection option and random addition of all 21 taxa, were used to find the most likely reconstructions from the data set; a branch and bound search was used to find all ofthe most parsimonious trees and a neighbour joining tree was found using the Jukes-Cantor model of nucleotide substitution (Jukes and Cantor, 1969). Two bootstrapping approaches were used to estimate sampling variation and support for branches: 1000 neighbour 19 joining bootstrap replicates with a Jukes-Cantor model of nucleotide substitution and 200 bootstrap replicates using parsimony. The number of parsimony bootstrap replicates was restricted, because the program ran out of memory when a particular bootstrap sample missed the few informative sites and generated a very high number of possible trees. To evaluate the relative signal to noise ratio in the data set, the length of the most parsimonious tree found using parsimony was compared to the tree-length distribution generated from searches based on random permutations of the data by using the permutation-tailed-probability (PTP) test option in PAUP*. R E S U L T S DNA extraction and PCR amplification. In Coleman and Goff s (1991) review of available molecular techniques performed on algal material, it is evident that no one method for the isolation of D N A consistently yields good results. In this study the extraction and amplification of D N A from unicellular endophytes obtained from the field was especially difficult and required the use of two different methods for the two morphologically distinct endophytes. The use of 30-100 'Chlorochytrium' cells (from Mazzaella splendens blades) as starting material for the Chelex extraction protocol was successful for Sooke isolates. Difficulty in obtaining genomic D N A from 'Chlorochytrium' cells from Bamfield M. splendens blades, however, resulted in sequencing only one isolate from this site. This and the fact that dissection of individual cells from host tissue was very time consuming, prompted the use of the C T A B / chloroform extraction method for 'Chlorochytrium' / M. splendens tissue (25-75 'Chlorochytrium' cells present). Nonetheless, obtaining adequate amounts of D N A from PCR amplification continued to be somewhat problematic. The CTAB/chloroform extraction method was successful for 'Codiolum'/ 'Petrocelis' samples (200-1000 'Codiolum' cells present) from all sites. Goff and Moon (1993) have shown that, with red algae, it is possible to amplify D N A from as little as 2 p.L of extraction supernatant of spores, which translates to ca. two spores. However, attempts to amplify D N A from a single unicell of 'Chlorochytrium' or 'Codiolum' were unsuccessful. 'Chlorochytrium' and 'Codiolum' D N A was successfully PCR amplified along with host D N A , resulting in two fragments of estimated size 0.5 kb and 0.75 kb (Fig. 1.3a). The host D N A was suspected to be the 0.75 kb fragment based on ITS sizes in other red algae (Steane et al. 20 1991). Upon agarose purification ofthe 'Codiolum' fragments and 100-fold dilution of 'Chlorochytrium' primary PCR product, re-amplification with the Acrosiphonia specific primer M O l , resulted in a single bright band of fragments for each endophyte (Fig. 1.3b). 'Petrocelis' -' C o d i o l u m ' / 1 2 3 4 5 6 7 -1018bp - 506 bp -1018 bp - 506 bp a. b. Figure 1.3a. Ethidium bromide stained agarose gel of PCR amplified ITS1, 5.8S, ITS2 and partial 26S regions from 'Petrocelis' crust colonised by 'Codiolum petrocelidis'. Note that the 0.5 kb fragment is 'Codiolum' and the 0.75 kb fragment is 'Petrocelis'. b. Agarose gel of 'Chlorochytrium' and 'Codiolum' PCR product re-amplified with M O l and ITS4 primers. DNA Sequences. The total length ofthe ITS1, ITS2 and 5.8S regions consists of 471 nucleotides, and it begins 66 bp from the 5' end ofthe ITS1 region and ends 23 bp into the 26S gene. A l l sequence variation was found in the ITS1 and 2 regions; 5.8S and the partial 26S sequences were identical. A l l unambiguous endophyte sequences (22 ofthe 27 isolates) were 100% identical to either Acrosiphonia arcta, Acrosiphonia coalita or Acrosiphonia sp. from Alaska and from Sooke. Among these 'Codiolum', 'Chlorochytrium' and Acrosiphonia isolates five distinct genotypes were found (Fig. 1.4). Six variable sites are present, with the exception 21 of the two isolates, A. sp. from Sooke and its 'Chlorochytrium' match, which show additional variable sites. The 'Codiolum' isolates separated into three genotypes, identical to A. arcta (genotype groups 1 and 2) and A coalita (genotype group 3) isolates. Among the 'Chlorochytrium' isolates, four distinct genotypes were present: (1) identical to A. arcta isolates (genotype groups 1 and 2) (2) identical to A. sp. isolates from Alaska (genotype group 4) and (3) one 'Chlorochytrium' isolate from Sooke ('Chlorochytrium' Sk5) corresponded to a unique Acrosiphonia isolate from Sooke (genotype group 5). This sequence is interesting in that four new nucleotide changes were found within the ITS1 and ITS2 regions and one new change within the 5.8S region. A n ambiguous 'Chlorochytrium' sequence obtained from a Sooke isolate (Sk2) did not match any other sequences. It should be noted here that on occasion, among 'Chlorochytrium', heterogeneity was evident at some of the sites, i.e. chromatograms sometimes revealed strong signal from a mixture of two nucleotides at the same position. This was the case for position 6 of the 'Chlorochytrium' Sk2 sequence and position 303 of the Bm5, Bm6, SklO, Ski 1 sequences, where both T and C were indicated (Fig. 1.4). Among the Acrosiphonia isolates sequenced there were two genotypes for A. arcta (varying by 1 bp at position 48), a single A. coalita genotype (different from A. arcta by nucleotide changes occurring at four or five of the six variable sites) and two additional genotypes for unknown Acrosiphonia species from Alaska and Sooke. The isolates from Alaska differed from A. coalita by 4 bp and from A. arcta by 2 or 3 bp. The Acrosiphonia sp. isolate from Sooke was quite distinct for this geographical region, characterised by five new nucleotide changes. It did, however, show close similarity to an Acrosiphonia isolate from Helgoland (10 of the 11 informative sites were identical; data not shown). The largest amount of variation among endophyte and Acrosiphonia sequences was seen at Sooke, where all five genotypes were present. A. coalita isolates sequenced from Bamfield and Sooke were 100% identical to isolates from Friday Harbor (Washington) and Monterey Bay (California), and A. arcta isolates from Burrard Inlet and six of the nine A. arcta isolates from Sooke were identical to isolates from Friday Harbor. Phylogenetic reconstructions. Maximum likelihood analysis of the rDNA sequences from the endophytes and from 40 isolates belonging to the Codiolales (Acrosiphonia, Spongomorpha, Urospora, Ulothrix and Monostroma species) found a single reconstruction Q. 3 O O O e 4} O CN c o •c co O D. u 13 •o o 0) U U H < < O u u u 3 3 pa i . a s .,OS 3 tS - CQ • | •a J2 O 43 D U U H U U U U U H H U U U U U U U U U H H o a < < < < < < u u < < < < < u < < < u u < < o o o o o o o U U U H H U U < < < O O O D O O O O H H H H O O U U U H ^ U U U U U U y -2 £ CQ CQ - | "6 £ 2, £ o S •a s O 43 u u 1 6 moa 1 1 CQ pa "a "e 3 3 'It 43 43 o o O O O O 2 2 u u SI vo oo 4* oo 4«S oo u 4* o o 4* 00 "B S -3 £ -2 l a 43 o o o 3 U 2 oo S ON • 3 t°? g 00 s a U <3 3 oo M oo "6 3 9> i o V9 ^ 44 44 44 00 00 00 "e "e 'e 3 3 3 I I I 43 43 43 u o o o u o ^ 2 ci cd « GO H CO -S > 00 H h—t 43 g 'J3 VI a - -* Sb43 3 <*> 5 C3 X • 1—4 00 o £ 'co c O in C^ CO u 8 •s a o co 43 CN O u s co O <a -s: 00 Cv S 'ST 1 s CJ CJ o -° S s * ^ CN 2 n, c § o 2 a  o 4=1 O 43 t8 43 O 2 o 43 y 00 c o S O C3 o ts -a | 82 CU U c C3 _ O WD'oO "O tx o 3 co 2 <U co" <U O-CD S C H . u >, .a * § to ^ o o *-43 E fl >^  c3 43 - 2 a o | y E u O 43 co «< ^ fa 2 cb H 22 23 (shown in Fig. 1.5). Six equally parsimonious trees were found with the branch and bound search, and a neighbour joining tree was found using the Jukes-Cantor model of nucleotide substitution (data not shown). The maximum likelihood, parsimony and neighbour joining reconstructions differed in the arrangement of the nearly identical species in the Acrosiphonia clade, for branches where bootstrap support was never more than 50%. A l l analyses, including the neighbour joining and parsimony bootstrap consensus, showed that the endophytes cluster in the Acrosiphonia clade with high bootstrap support (80% - 91%). A l l of the reconstructions also show Spongomorpha as the sister genus to the Acrosiphonia clade. In the maximum likelihood trees Ulothrix appears as the sister taxon to Urospora (Fig. 1.5). In the parsimony and distance trees, however, Ulothrix diverges after Monostroma, at the base of the tree. A significant difference (p < 0.01, PTP test) was found between trees generated using parsimony and those generated from random permutations of the data, suggesting that phylogenetic signal in the data could be distinguished from random noise. A PTP test conducted using only sequences within the Acrosiphonia clade also showed a significant difference (p < 0.01), suggesting that although resolution within the clade was limited, there was phylogenetic signal in the data. The strong support for the Acrosiphonia clade, and its low internal resolution [except for the high bootstrap support relating Acrosiphonia sp. (Sk) to A. sonderi (Kiitzing) Kornmann (Hg)], was due to the low sequence variation between the taxa included in this clade (Figs. 1.4 and 1.5). The relatively long branch lengths between the Acrosiphonia I Spongomorpha clade and the clade including Urospora and Ulothrix species (Fig. 1.5) clearly show that the endophytes sequenced are unambiguously more closely associated with the former group. DISCUSSION The 100% identity of nucleotide sequences of endophyte / Acrosiphonia pairs provides conclusive evidence that the unicells present within Mazzaella splendens and 'Petrocelis franciscana' in southern British Columbia are sporophytes of Acrosiphonia. Each unambiguous 471-bp sequence of the ITS regions of the 'Chlorochytrium' and 'Codiolum' isolates corresponded to the sequence of at least one of the Acrosiphonia isolates (Fig. 1.4). Phylogenetic clustering of the Codiolales (genera which are reported to exhibit 'Codiolum' phases in their life histories) establishes the fact that 'Chlorochytrium' and 'Codiolum', 24 A. coalita MB/FH/ Bm1-5/Sk1-3 'Codiolum'Bml A. sonderi Hg 9 7 LA. sp.Sk 'Chlorochytrium' Sk5 A arcta FH/BI1-10/Sk1-6 'Codiolum' Bl/Sk A. arcta Bm/Sk7-9 'Codiolum' Bm2 'Chlorochytrium' Bm1/Sk3-4 A. sp. AH/BS 'Chlorochytrium'Sk6-9/Bm2-4/BI S. lanosa Ic U. penicilliformis KG/Di I/, penicilliformis Hg U. penicilliformis Du U. implexa KG Acrosiphonia Spongomorpha Urospora M. arcticumSp Ulothrlx | Monostroma 0.01 Figure 1.5 Maximum likelihood tree from ITS1 and ITS2 sequences. Monostroma arcticum was used as the outgroup. Numbers above the branches are neighbour-joining bootstrap values (1000 replicates) using a Jukes-Cantor model of nucleotide substitution; numbers below the branches are bootstrap proportions using parsimony (200 replicates). The branch to Monostroma is not proportional to its length of 0.765. Otherwise, the scale bar on the tree indicates the maximum likelihood distance. See Table 1.1 for detailed information on species, collection sites and abbreviations. 25 sequenced from southern British Columbia, were not the sporophytes of Urospora, Ulothrix or Monostroma (Spongomorpha is not present in the northeast Pacific). A l l 'Chlorochytrium' and 'Codiolum' isolates were placed within the clade including all of the Acrosiphonia isolates sequenced (Fig. 1.5). A l l other genera, including Spongomorpha, whose 'Codiolum' and 'Chlorochytrium' phases are reported to be endophytic in Haemescharia hennedyi (Harvey) Vinogradova and Yakoleva and Polyides rotundus (Huds.) Greville, respectively (Kornmann 1964) were distinguished as separate well-resolved clades, separated by relatively long branch lengths (Fig. 1.5) compared to those within the Acrosiphonia clade. Based on characteristics described by Hudson (1974), only two morphological species of Acrosiphonia, A. coalita and A. arcta, are believed to be present along the British Columbia coast. The ITS sequence data revealed five genotypes present among Acrosiphonia and endophyte isolates from British Columbia. Low bootstrap support and limited resolution within the Acrosiphonia clade of the Codiolales phylogenetic tree do not, however, allow interpretation of (1) whether it is likely that the Acrosiphonia "species" sequenced represent separate biological entities or a single biological entity; or (2) whether the rate of nucleotide substitution in the ITS region is insufficient to resolve relationships at this scale of relatedness. Nonetheless, A. arcta and A. coalita do not appear to be conspecific based on the fact that consistent ITS sequence variation is correlated with morphological differences. Both A. arcta and A. coalita are present at Friday Harbor, Sooke and Bamfield; at each site their ITS sequences consistently differ at four or five nucleotide positions (Fig. 1.5). In the ITS trees the A. arcta isolates do not form a monophyletic clade, but this probably results from lack of information in the ITS sequence data, and possibly reflects problems with species delimitation. Acrosiphonia specimens from Alaska were difficult to identify based on morphology. Yet, shared similarities (absence of hooked branchlets, filament cell diameter and habit) with U B C herbarium specimens identified as Acrosiphonia hystrix (Stromfelt) Jonaason, and ITS sequences that differed from all other Acrosiphonia, suggest consideration as a separate species. The 'Chlorochytrium' isolates from Sooke, Bamfield and Burrard Inlet with identical sequences to A. sp. (Alaska) would then represent its sporophyte (by implication, Acrosiphonia with the Alaskan ITS genotype should also be present at all three study sites). Likewise, a single Acrosiphonia sample obtained higher up in the intertidal zone at Sooke (on a massive boulder) is characterised by a different genotype. It also cannot be identified morphologically (and attempts to collect and sequence more isolates of this genotype were unsuccessful), but warrants 26 consideration as a separate species, since its sequence is almost identical to that of an isolate of Acrosiphonia sonderi (Fig. 1.5). A. sonderi, type locality Helgoland, has not been reported for the Pacific, and sexual reproduction has never been observed for this "species" (Kornmann 1962). This points to the need for resolution within the Acrosiphonia species complex. Further studies, perhaps taking a multiallelic population level approach, might clarify the breeding patterns of these algae, and also better link the endophytes with individual Acrosiphonia species. Regardless of how many Acrosiphonia species are recognised, the ITS sequence data suggest that the morphologically distinct endophytes, 'Chlorochytrium' and 'Codiolum', can be produced by the same Acrosiphonia species, i.e. A. arcta's sporophyte can colonise both blade and crust. 'Chlorochytrium' and 'Codiolum' isolates from Bamfield and from Sooke revealed 100% sequence identity to an A. arcta genotype (Fig. 1.4). This supports Kornmann's (1964) long-standing hypothesis that 'Chlorochytrium' and 'Codiolum' represent alternate phenotypes of the same sporophyte. However, no 'Chlorochytrium' / 'Codiolum' matches with A. arcta were found at Burrard Inlet, and only 'Codiolum' sequences were found to share a genotype with .4. coalita. Among 13 unambiguous sequences of'Chlorochytrium' isolates, 9 were 100% identical to A. sp. from Alaska, three matched A. arcta sequences and one was 100% identical to the Sooke A. sp. None were found to share a genotype with A. coalita. This is consistent with previous culture studies where only 'Chlorochytrium' gave rise to A. arcta (Fig. 1.1), indicating that Acrosiphonia coalita may produce sporophytes which only colonise red algal crusts. More generally, the sporophytes of Acrosiphonia species may exhibit variable host specificity. And yet, consideration of sequence ambiguity provides an alternative interpretation. The sequence ambiguity found in a number of 'Chlorochytrium' isolates, suggests that a heterogeneous group of endophytes may have been present in single D N A extractions (recall that D N A extracts consisted of patches of 'Petrocelis' and Mazzaella splendens colonised by > 30 cells). Theoretically the unicells in one D N A sample could have been derived from different sources, i.e. sporophytes from different Acrosiphonia species or from different genera could cohabit the same host [Miyaji and Kurogi (1976) reported obtaining three species of Acrosiphonia in culture from 'Chlorochytrium' endophytic in the same host plant from Japan] or some endophytes may represent independent species not associated with the life history of another alga. However, since all sequence ambiguity was found at the six variable sites, and fell 27 within the range of variation observed among Acrosiphonia sequences, it seems evident that all unicells were most closely related to Acrosiphonia. Verification of this assumption would require screening of multiple clones from each isolate. Regardless, in instances of unambiguous sequences where possibly only the dominant sequence was amplified, it can definitely be concluded that this predominant endophyte type was related to Acrosiphonia. With regard to the implication that a mixture of endophytic sporophytes from different Acrosiphonia species may have comprised some D N A extracts, i.e. ambiguous sites suggested neither sequence was dominant, 'Chlorochytrium' (Sk2) may represent such a sample. This is the 'Chlorochytrium' isolate from Sooke which did not show 100% sequence identity with any Acrosiphonia sequences. Based on the sites which show heterogeneity, this isolate would consist of sporophytic endophytes from A coalita, A. arcta and/or A. sp. (Alaska). Screening of multiple clones from this isolate could then substantiate whether 'Chlorochytrium' is in fact associated with A. coalita, contrary to culture study results. In California, where A. arcta is not known to occur, 'Chlorochytrium' cells were reported in Mazzaella splendens blades (Smith, 1944), suggesting A. coalita's sporophyte may not be confined to crusts, at least not in that region. Ultimately, though, it seems that utilisation of a sequence region more variable than the ITS1 and ITS2, and D N A extracts from individuals (rather than > 30cells), would enable better understanding of host / endophyte relationships. The fact that D N A sequences of'Chlorochytrium' isolates share the genotype of an Acrosiphonia "species" from Alaska is particularly interesting, because no Acrosiphonia specimens collected from southern British Columbia in both 1997 and 1998 matched the Alaskan isolate sequences. Absence of the genotype in Acrosiphonia isolates sequenced may simply be due to small sample size or low numbers of the genotype present in Acrosiphonia populations. The high proportion of'Chlorochytrium' isolates with 100% identity to A. sp. (Alaska), as well as the failure to detect 1) 'Codiolum' with an A. sp. match and 2) 'Chlorochytrium' at Burrard Inlet with the Acrosiphonia genotype found at Burrard Inlet, is also surprising. Perhaps the sporophyte of Acrosiphonia characterised by the Alaskan ITS genotype is host specific (colonising only red algal blades and so not associated with 'Codiolum'), and is more successful than the sporophyte of A. arcta in colonising Mazzaella splendens. On the other hand, the failure to detect 'Codiolum' sharing a genotype with A. sp. (Alaska) may simply be a function of low numbers of isolates sequenced. Again, a study taking a population genetics 28 approach or examining a more variable sequence region and more isolates, might shed light on these intriguing questions. As already mentioned, Acrosiphonia ITS sequence variation was found to be highest at Sooke, where all five genotypes are present (two genotypes present at Bamfield and only one at Burrard Inlet). This is interesting because the three geographic sites represent environmentally different areas with regard to factors such as wave exposure and salinity. It is perhaps not surprising to find low sequence variation at Burrard Inlet because this is the most wave-sheltered and homogenous site. Acrosiphonia isolates from Sooke were collected from a number of different habitats, i.e. from different heights in the intertidal zone. Possibly, the degree of habitat heterogeneity present at a site may play a role (among other factors) in determining the degree of infrageneric variation present. D N A amplification and sequencing have proven to be powerful tools in the identification of unicellular green algae endophytic in red algal hosts along southwestern British Columbia, Canada. The stalked unicell present in the crust 'Petrocelis franciscana' and the spherical unicell within the blade Mazzaella splendens, constitute the alternate phase of the free-living filamentous alga Acrosiphonia. Regardless of how many Acrosiphonia species are recognised, the ITS sequence data has shown that the morphologically different endophytes, 'Chlorochytrium' and 'Codiolum', can be produced by the same Acrosiphonia species, A. arcta. The sporophyte of A. coalita, on the other hand, has not been shown to be associated with 'Chlorochytrium', suggesting it is host specific, only colonising crusts. These results support previous culture studies (Hollenberg 1958, Fan 1959, Chihara 1969, Hudson 1974) and Kommann's hypothesis that, at least in some cases (e.g. A. arcta), 'Chlorochytrium' and 'Codiolum' represent alternate phenotypes of the same sporophyte. Chapter 2 wil l examine Acrosiphonia's alternation of a filamentous gametophyte with a unicellular, endophytic sporophyte in nature. 29 R E F E R E N C E S Bakker, F.T., Olsen, J.L. & Stam, W.T. 1995. Evolution of nuclear rDNA ITS sequences in the Cladophora albida I sericea clade (Chlorophyta). J. Mol. Evol. 40:640-51. Blomster, J., Maggs, C. A . & Stanhope, M . J. 1999. Extensive intraspecific morphological variation in Enteromorpha muscoides (Chlorophyta) revealed by molecular analysis. J. Phycol. 35:575-586. Burkhardt, E. & Peters, A. F. 1998. Molecular evidence from nRDNA ITS sequences that Laminariocalax (Phaeophyceae, Ectocarpales sensu lato) is a worldwide clade of closely related kelp endophytes. J. Phycol. 34:682-91. Chihara, M . 1969. Culture study of Chlorochytrium inclusum from the northeast Pacific. Phycol. 8:127-33. Coleman, A . W. & Goff, L. J. 1991. D N A analysis of eukaryotic algal species. J. Phycol. 27:463-73. Fan, K . C. 1959. Studies on the life histories of marine algae I. Codiolum petrocelidis and Spongomorpha coalita. Bull. Torrey Bot. Club 86:1-12. Goff, L . J. & Moon, D. A . 1993. PCR amplification of nuclear and plastid genes from algal herbarium specimens and algal spores. J. Phycol. 29:381-4. Hanic, L. A . 1965. Life history studies on Urospora and Codiolum from southern B.C. Ph.D. Thesis, University of British Columbia, Vancouver, B.C. , 152 pp. Higgins, D. G., Bleasby, A . J. & Fuchs, R. 1992. C L U S T A L V : improved software for multiple sequence alignment. Comput. Appl. Biosci. 8:189-91. Hollenberg, G. J. 1958. Observations concerning the life cycle of Spongomorpha coalita (Ruprecht) Collins. Madrono 14:249-52. Hudson, M . L . 1974. Field, culture and ultrastructural studies on the marine green alga Acrosiphonia in the Puget Sound region. Ph.D. Thesis, University of Washington, 206 pp. Jonsson, S. 1959a. L'existence de ralternance heteromorphe de generations entre 1'Acrosiphonia spinescens et le Codiolum petrocelidis. C. R. Acad. Sci., Paris 248:835-7. 1959b. Le cycle de developpement du Spongomorpha lanosa (Roth) Ktitz. et la nouvelle famille des Acrosiphoniacees. Ibid. 248:1565-7. 1962. Recherches sur des Cladophoracees marines (structure, reproduction, cycles compares, consequences systematiques). Ann. Sci. Nat. Bot. 12 :25-230. 30 1966. Sur l'identification du sporophyte du Spongomorpha lanosa. C.R. Acad. Sci., Paris 262:626-9. Jukes, T. H . & Cantor, C. R. 1969. Evolution of protein molecules. In Munro, H . N . (ed.) Mammalian Protein Metabolism. Academic Press, New York, pp. 21-132. Kornmann, P. 1961a. Uber Spongomorpha lanosa und ihre Sporophytenformen. Helg. Wiss. Meeresunters. 7:195-205. 1961b. Tiber Codiolum und Urospora. Ibid. 8:42-57. 1962. Die Entwicklung von Monostroma greveillei. Ibid. 8:195-202. 1963. Die Lebenszyklus einer marinen Ulothrix-Ait. Ibid. 8:357-60. 1964. Zur Biologie von Spongomorpha aeruginosa. Ibid. 11:200-8. 1965. Zur Wachstum und Aufbaus von Acrosiphonia. Ibid. 12:219-38. 1970. Phylogenetische Bezeihungen in der Griinalgengattung Acrosiphonia. Ibid. 21:292-304. 1973. Codiolophyceae, anew class of Chlorophyta. Ibid. 25:1-13. McDermott, J. M . , McDonald, B. A. , Allard, R. W. & Webster, R. K . 1989. Genetic variability for pathogenecity, isozyme, ribosomal D N A and colony color variants in populations of Rhychosporium secalis. Genetics 122:561-5. Miyaji, K . 1984. The life history study on a sporophytic form of Spongomorpha-Acrosiphonia complex (Acrosiphoniales, Chlorophyta), Codiolum petrocelidis Kuckuck from Northern Japan. Jpn. J. Phycol. 32:29-36. Miyaji, K . 1996. Taxonomic study of the genus Spongomorpha (Acrosiphoniales, Chlorophyta) in Japan. Phycol. Research 44:27-36. Miyaji, K . and Kurogi, M . 1976. On the development of zoospores ofthe green alga Chlorochytrium inclusum from eastern Hokkaido. Bull. Jap. Soc. Phycol. 24:121-9. Peters, A . F., van Oppen, M . J. H. , Wiencke, C , Stam, W. T. & Olsen, J.L. 1997. Phylogeny and historical ecology ofthe Desmarestiaceae (Phaeophyceae) support a southern hemisphere origin. J. Phycol. 33:294-309. Pillmann, A. , Woolcott, G. W., Olsen, J. L. , Stam, W. T. & King, R. J. 1997. Inter- and intraspecific genetic variation in Caulerpa (Chlorophyta) based on nuclear rDNA ITS sequences. Eur. J. Phycol. 32:379-86. Qian, L . & Wilkinson, M . 1991. D N A fragment purification: Removal of agarose 10 minutes after electrophoresis. BioTechniques 10:736-37. 31 Scagel, R. F. 1966. Marine algae of British Columbia & northern Washington, Part I: Chlorophyceae. Nat. Mus. Can. Bull. No. 207, Biol . Ser. No. 74, 257 pp. Serrao, E. A . , Alice, L. A . & Brawley, S. H . 1999. Evolution of the Fucaceae (Phaeophyceae) inferred from nrDNA-ITS. J. Phycol. 35:382-394. Setchell, W. A . & Gardner, N . L. 1920. The marine algae of the Pacific Coast of North America. Part II. Chlorophyceae. Univ. Calif. Publ. Bot. pp. 149-50. Smith, G. M . 1944. Marine algae of the Monterey Peninsula. Stanford University Press, 622 pp. Steane, D. A . , McClure, B. A. , Clarke, A . E. & Kraft, G. T. 1991. Amplification of the polymorphic 5.8S rRNA gene from ribosomal R N A from selected Australian gigartinalean species (Rhodophyta) by polymerase chain reaction. J. Phycol. 27:758-62. Swofford, D. L . 1998. PAUP*. Phylogenetic analysis using parsimony * and other methods. Sinauer Associates, Sunderland, Massachusetts. van den Hoek, C , Mann, D. G. & Jahns, H . M . 1995. Algae. An Introduction to Phycology. Cambridge University Press, Cambridge, 623 pp. van Oppen, M . J. H . 1995. Phylogenetic relationships among some Ulotrichalean (Ulvophyceae, Chlorophyta) algae, as inferred from rDNA ITS and partial 18S sequences. In TrackingTrails by Cracking Codes. Ph. D. Thesis, University of Groningen, The Netherlands, pp. 59-76. van Oppen, M . J. H , Olsen, J. L., Stam, W. T., van den Hoek, C. & Wiencke, C. 1993. Arctic-Antarctic disjunctions in the benthic seaweeds Acrosiphonia arcta (Chlorophyta) and Desmarestia viridislwillii (Phaeophyta) are of recent origin. Mar. Biol. 115:381-6. Walsh, P. S., Metzger, D. A. & Higuchi, R. 1991. Chelex 100 as a medium for simple extraction of D N A from PCR-based typing from forensic medicine. BioTechniques 10:506-13. White, T. J., Bruns, T. D., Lee, S. B. & Taylor, J. W. 1990. Analysis of phylogenetic relationships by amplification and direct sequencing of ribosomal genes. In Innis, M . A . , Geleand, D. H , Sninsky, J. J. & White, T. J. (Eds.) PCR Protocols. Academic Press, New York, pp. 315-22. Wille, N . 1900. Die Zellkerne bei Acrosiphonia (J. Ag.) Kjellm. Botanisches Centralblatt 8 1238-9. 32 CHAPTER 2 LIFE HISTORY OF ACROSIPHONIA IN SOUTHWESTERN BRITISH COLUMBIA: A BASELINE STUDY INTRODUCTION Extensive studies, primarily in the 1960s and 1970s (Hollenberg 1958, Fan 1959, Jonsson 1959a 1959b 1962 1963 1966 1970, Kornmann 1961 1964 1970a 1972, Chihara 1969, Hudson 1974, Miyaji and Kurogi 1976, Miyaji 1984 1996) examined the Acrosiphonia - Spongomorpha complex in culture with the aim of elucidating the life cycle. However, no study to date has focused on the natural dynamics of Acrosiphonia, i.e. seasonal abundance and reproductive phenology of the filamentous gametophyte and endophytic sporophyte in nature. Lack of fundamental ecological studies may, in part, be due to the microscopic nature ofthe endophytic sporophytes of Acrosiphonia. This study establishes the relationship of Acrosiphonia and 'Chlorochytrium' and 'Codiolum' in the field, as well as the timing of life history events. Acrosiphonia's alternation of heteromorphic generations (Fig. 0.1, Thesis Introduction) established through the culture studies noted and the molecular study in Chapter 1 is not unique among algae. Heteromorphic life histories occur in all three major groups of macroalgae (Rhodophyta, Chlorophyta, Phaeophyta). Examples include the green algal genera already discussed in Chapter 1 (Urospora, Ulothrix and Monostroma), as well as the red algae, Mastocarpus, Porphyra and Schizymenia (DeCew et al. 1992) and the brown algae, Petalonia, Scytosiphon and all kelps. Each of these algae have two separate ecologically distinct phases which are so dissimilar in morphology that they had been classified as separate species (with the exception of kelp gametophytes) and, in some cases, placed in separate families or orders. This is the case for Acrosiphonia, whose 'Chlorochytrium' and 'Codiolum' sporophytic phases were initially considered independent genera and placed, separate from Acrosiphonia, in the Order Chlorococcales. These algae also exhibit a seasonal or annual upright phase and a non-upright crustose, boring, endophytic or epilithic phase (some kelp sporophytes are perennial; not all genera display obligatory alternations of generations). Kelp gametophytes have recently been found as endophytes in foliose and filamentous red algae (Garbary et al. 1999a, b). 33 Description of Acrosiphonia's phases in nature. Much taxonomic work has been carried out for both separation of Acrosiphonia and Spongomorpha (Agardh 1846, Wille 1900, Collins 1909, Kornmann 1965 1970b, Scagel 1966, Jonsson 1991) and the delineation of Acrosiphonia and Spongomorpha species (Setchell & Gardner 1920, Kornmann 1962, Jonsson 1962 1971, Scagel 1966, Hudson 1974). Spongomorpha is absent from the northeast Pacific, as are the following Acrosiphonia species: A. grandis and A. sonderi reported from Europe and A. spiralis Sakai and A. heterocladia Sakai reported from Japan. Acrosiphonia hystrix and A. duriuscula (Ruprecht) Collins , although present in the northeast Pacific, have not been reported south of Alaska. Scagel (1966) described five species of Acrosiphonia for northern Washington State and British Columbia. However, Hudson (1974) reported variability of morphological characters on which the species were based when plants were subjected to different environmental conditions in the laboratory. Furthermore, a number of plants collected in the Puget Sound region, Washington State, of different "species", were found to intergrade in all diagnostic characters. Nonetheless, allowing for broad range of variation, Hudson selected characters conservatively to establish two morphological species, A. arcta and A. coalita (she actually chose to accept A. spinescens in favour of A. arcta, because the true characteristics of A. arcta type specimens were unclear to her, but A. arcta has taxonomic priority). Her conclusions were based on 1) specimens collected from the Puget Sound Region, Washington State 2) descriptions of the type specimens of A. coalita, A. mertensii (Ruprecht) Setchell & Gardner, A. saxatilis (Ruprecht) Collins, A. arcta and A. spinescens and 3) specimens collected and identified by Collins, Setchell and Gardner, Scagel and Jonsson. For a more detailed account of the taxonomic work on Acrosiphonia refer to Hudson (1974). Acrosiphonia coalita appears to be present only in the northeast Pacific, whereas A. arcta has been recorded in both the northern and southern hemispheres, including the northeast Pacific, the Northwest Atlantic, the North Sea, Greenland, southern Chile and Antarctica. Seasonal abundance and reproductive phenology of Acrosiphonia are poorly understood. Similarly, no studies have described 'Chlorochytrium' seasonal abundance and reproductive phenology in the northeast Pacific. 'Chlorochytrium inclusum' was first described by Kjellman (1883) in Sarcophyllis arctica Kjellman, an Arctic, foliose, red algal species. In its vegetative state 'Chlorochytrium' was observed to be almost spherical, 80-100 um in diameter, attaining diameters of up to 275 um when mature. Other researchers reported 'Chlorochytrium' 34 with maximum diameter of 75-100 um from a number of foliose red algae from the Pacific northeast (Chihara 1969 from Schizymenia, Setchell & Gardner 1920 from Weeksia, Constantinea and Mazzaella), Europe (Jonsson 1959b 1962 1966, Kornmann 1961, 1964 from Polyides and Dilsea) and Japan (Miyaji and Kurogi, 1976 from Farlowia). The alternate phase of some of these has been identified in culture studies. For example, 'Chlorochytrium' from Europe gave rise to Spongomorpha lanosa, while 'Chlorochytrium' from Japan gave rise to Acrosiphonia spiralis, A. heterocladia and A. duriuscula. Two studies on Arctic and European 'C. inclusum' have found it to be fertile in winter (Kjellman 1883, Kornmann 1964). Seasonal abundance and reproductive phenology of 'Codiolum' is also poorly understood, and is based on a handful of studies (none from the northeast Pacific). 'Codiolum petrocelidis' was first described by Kuckuck (1894) from specimens growing in the crust Haemescharia hennedyi in Helgoland, Germany. These cells were reported as having a vesicle portion 125-175 um in length and 25-50 um in diameter and a colourless stalk. A number of workers (Setchell & Gardner 1920, Jonsson 1958, Kornmann 1972) have noted polymorphism (with regard to vesicle and stalk size and shape) exhibited by 'Codiolum' cells. This phenomenon undoubtedly interferes with the ability to follow cell growth and development from specimens collected in the field. The filamentous plants that developed in culture from the zoospores of 'C . petrocelidis' (collected from H. hennedyi in Helgoland) were identified as Spongomorpha lanosa (Kornmann 1961,1964). Other culture studies suggested that 'Codiolum' represents a phase in the life history of a number of Acrosiphonia species. Jonsson's (1959a, 1962) 'Codiolum' found within 'Petrocelis cruenta' J. Agardh [= tetrasporophytic phase of Mastocarpus stellatus (Stackhouse) Guiry] collected at Roscoff, France, were associated with Acrosiphonia arcta. In Japan Miyaji (1984) cultured zoospores of'Codiolum', endophytic in 'Petrocelis' crusts of Mastocarpus pacificus (Kjellman) Perestenko, which developed into filamentous gametophytes identified as A. saxatilis [which may be synonymous with A. arcta according to Hudson (1974)]. On the Pacific Coast of North America 'Codiolum' reported in 'Petrocelis franciscana' gave rise to both A. arcta and A. coalita in culture (Hollenberg 1958, Fan 1959, Hudson 1974). The only study to examine 'Codiolum' development in the field is that of Kornmann (1961). Based on cell sizes and fertility of specimens collected in the winter of 1963 and spring and summer of 1964, Kornmann suggested the youngest stages found within 'Petrocelis' in July pass the following winter in a vegetative state and become fertile the subsequent winter. In the northeast Pacific 'Codiolum' has been collected from 'Petrocelis franciscana' from February to 35 November (Setchell & Gardner 1920, Hollenberg 1958, Fan 1959, Hudson 1974, Dethier 1987), and Dethier (1987) noted dense colonisation in summer in Washington State. Environmental requirements for gametophyte and sporophyte. Hanic (1965) established that 'Codiolum gregarium', the macroscopic (1-2 mm in length at maturity), free-living 'Codiolum' phase of Urospora is abundant and fertile in the winter; fertility being induced by cold temperatures. He observed that Urospora is present year round, but dies off substantially in the summer months. It is 'Codiolum gregarium' which presumably better survives high summer temperatures and desiccation (L. Hanic, pers. comm.). Experimental studies have not identified environmental requirements for the production and growth of Acrosiphonia's sporophytes, 'Chlorochytrium' and 'Codiolum'. Nonetheless, Hudson (1974) and Miyaji (1996) have shown that Acrosiphonia arcta and A. spiralis, respectively, did not produce zoospores at temperatures >15 °C (but did at 5 °C and 10 °C). With regard to Acrosiphonia's gametophyte, Hudson (1974) found that growth of plants in culture was inhibited at 15-20 °C, and long-day (16 hrs light: 8 hrs dark photoregime) conditions were required for the germination of Acrosiphonia filaments. This implies seasonality of Acrosiphonia, and is supported by the fact plants have never been collected in the winter. Collections from the northeast Pacific (Hollenberg 1958, Fan 1959, Hudson 1974), Europe (Jonsson 1959a 1964 1986, Kornmann 1970a 1970b) and Japan (Miyaji 1996) were all carried out in spring and summer. Urospora's suspected bet-hedging strategy, whereby filamentous and unicellular 'Codiolum' phases are adapted to a seasonally variable environment, may also apply to Acrosiphonia's life history. Acrosiphonia's endophytic sporophytes, 'Chlorochytrium' and 'Codiolum', are expected to demonstrate greater tolerance of extreme values of environmental factors such as temperature and photoperiod (keeping in mind that fertility of the endophytes may be triggered by cold temperatures in winter) than the filamentous gametophyte. The purpose of the study detailed in this chapter is to establish the timing of events, illustrated in Figure 2.1, in Acrosiphonia's life history in southwestern British Columbia, and to gain insights into the factors responsible for selection of this complex heteromorphic life history. More specifically, this study 1) identifies the filamentous gametophytes and endophytic sporophytes of Acrosiphonia in nature, 2) illustrates the seasonal abundance of Acrosiphonia (gametophyte) and its sporophytes, 'Codiolum' and 'Chlorochytrium', 3) determines 36 reproductive phenology of both phases and 4) examines endophyte cell size as an indication of age and growth in the field. Gamete release Acrosiphonia establishment When? Endophyte colonisation Zoospore release Figure 2.1. Establishing the timing of events of Acrosiphonia''s life history in nature. M A T E R I A L S A N D M E T H O D S Field sampling. Field studies conducted over two years, 1996-1998, at Sooke, Bamfield and Burrard Inlet monitored seasonal abundance and reproductive phenology of Acrosiphonia, 'Chlorochytrium' and 'Codiolum'. Notable changes in morphology, colour and / or size of organisms were recorded. Due to difficulties with sampling all three sites within a 3-5 day low tide series, Bamfield was sampled much less frequently than Sooke and Burrard Inlet. Acrosiphonia. Green filamentous algae were collected from the low to high intertidal zone at all three sites for identification to genus (Acrosiphonia) and to species, based on morphological / microscopic characters. Position in the intertidal zone and substratum type were noted. Sampling for percent cover and reproductive phenology of Acrosiphonia commenced with the appearance of plants in the spring of 1997. Every 2-6. weeks (depending on whether Acrosiphonia was present) a 20 m transect line was placed parallel to the water line in the Acrosiphonia zone. The same area was sampled on 37 consecutive sampling dates and ranged from 0.1 to 2.0 m above zero tidal level (Canadian Chart Datum). A random number table was used to generate 30 sites for placement of a 20 cm x 20 cm quadrat along the transect. Quadrats were always centred on the transect line. Monofilament divided the quadrat into 400 squares. Where both A. arcta and A. coalita were recognisable, percent cover was noted for each species. When Acrosiphonia appeared in new areas, additional transects were established. This was the case at Sooke where transects needed to be placed in three different intertidal zones (as described in Thesis Introduction) to ensure representative sampling of the entire area colonised by Acrosiphonia. At Bamfield one additional transect was sampled in an adjacent area at the same tidal height. To determine reproductive phenology for Acrosiphonia arcta and A. coalita at Sooke, Bamfield and Burrard Inlet, 10 thalli of each species were haphazardly collected from each Acrosiphonia zone on each sampling date, and brought back to the laboratory for microscopic examination. Microscopic characters used for distinguishing between the two species were useful in confirming the identity of individual plants, especially in juveniles. These thalli were then prepared as herbarium specimens for future reference, and some were later used for D N A extraction (Chapter 1). 'Chlorochytrium'. 'Chlorochytrium' cells were identified in Mazzaella splendens, the predominant foliose red alga at the three study sites. Seasonal abundance of 'Chlorochytrium' at Sooke, Bamfield and Burrard Inlet was based on density estimates of cells within M. splendens blades. A 12 m transect line for determining M. splendens density using the point-quarter method (see Chapter 3, Methods), was also used for the random collection of 32 blades per transect. Blades were collected, alternating between the largest and smallest (> 5 cm) blade per genet, from each quadrant of 8 points along the transect. Collection occurred approximately monthly, except on sampling dates where tides were not low enough to expose M. splendens or scarcity precluded collecting 30 blades. Mazzaella splendens blades from Sooke, Bamfield and Burrard Inlet were either examined immediately or kept in the refrigerator for 2-3 days. Drying of blades was not feasible, since shrunken 'Chlorochytrium' cells did not adequately rehydrate, and were difficult to locate and identify; frozen blades often deteriorated considerably. Although larger cells of'Chlorochytrium' (> 120 um) could be seen in the field by holding thin blades of Mazzaella splendens that were not epiphytised up to the light, magnification was 38 usually required to detect the bright green endophytic cells. This often meant scraping off patches of bryozoans and / or diatom films from the M. splendens blades, especially late in the summer and fall. Furthermore, both sides of the blade required inspection under the dissecting microscope, since the endophyte was found to reside in both the upper and lower cortex region. For blades with < 100 cells (or where single cells were scattered), individual cells could be counted, but densely colonised patches of endophytes needed to be estimated. Estimates were carried out by counting groups of approximately 100 cells (groups were based on the area occupied by 100 cells). Estimates were checked by actually counting five groups of 100 estimated cells from a blade where > 1000 cells and > 10,000 cells were estimated. This procedure was replicated five times. Absolute numbers of 'Chlorochytrium' cells were recorded per blade, as were frequency and range of cell size and location of endophyte patches. To calculate 'Chlorochytrium' densities cm"2, M. splendens blades were photocopied and the surface area determined as a percentage ofthe weight of a sheet of paper of known area. 'Chlorochytrium' cells were dissected from several blades after each sampling trip to check for fertile cells, and to record variation in morphology. Cross-sectioned Mazzaella tissue enabled cells to be examined within their host. 'Codiolum'. 'Codiolum' cells were present in 'Petrocelis franciscana', a conspicuous red algal crust at all three study sites. Seasonal abundance of 'Codiolum' was determined by counting numbers of cells present in patches of 'Petrocelis' roughly every month. Thirty small patches of approximately 5 x 5 mm were randomly scraped by razor blade from 'Petrocelis' growing on boulders or consolidated rock within the low to high intertidal zone (from 0.1 to 5.1 m above zero tidal level, Canadian Chart Datum). About 10 patches were collected from each zone. 'Petrocelis' patches with 'Codiolum' could be kept in the refrigerator for a few days prior to examination. 'Codiolum' cells were detected by squashing the 5 x 5mm 'Petrocelis' patch onto a microscope slide. Density, 'Codiolum' cell size, polymorphism among cells and reproductive state were recorded. In the summer and fall when densities at Sooke and Burrard Inlet exceeded 10,000 cells cm" , it was necessary to estimate numbers present in the field of view under the lOx objective lens. This method of estimation was established by actually counting the cells visible in a field of view for 10 different fields of view for five individual crust patches, i.e. the average cell count per field of view was used for the estimation. 39 Data Analysis. Acrosiphonia, 'Chlorochytrium' and 'Codiolum' data sets are comprised of percent cover and density values (sample size ranged from 30 to 90 fox Acrosiphonia; approximately 30 for 'Chlorochytrium' and 'Codiolum') for 15 - 23 sampling dates. Statistical analyses were performed using SPSS® 9.0 for Windows (1999). The Levene test showed that the assumption of homogeneous variances was violated. Although transformations reduced the Levene statistic, heterogeneity was not reduced to nonsignificant levels. Consequently, the data sets were also checked for normality, since the processes that produce non-normal distributions may also result in variance heterogeneity (Day & Quinn, 1989). The data satisfied the Kolmogorov-Smirnov Test for normality. Since standard non-parametric tests are inherently less powerful than parametric tests (Zar, 1996), and should not be used as a simple means to avoid the problem of unequal variances (Day & Quinn, 1989), A N O V A s (which are generally robust to variance heterogeneity, Zar 1996) were used in SPSS® 9.0 for Windows (1999). One-way A N O V A s and the Games-Howellpost hoc test were performed on square-root transformed data to test for significant differences in Acrosiphonia, 'Chlorochytrium' and 'Codiolum' abundance over time. The Games-Howell test is more powerful than other post hoc tests for unequal variances (Games et al., 1983), and is recommended when number of treatments is small and sample size > 7 (Day & Quinn, 1989). It also does not require equal sample sizes, which is important for Bamfield 'Codiolum' data where sample size was variable (7-30) due to the occasional misidentification of 'Petrocelis' in the field. R E S U L T S Acrosiphonia identification. Two morphological species, Acrosiphonia arcta and A coalita (Figs. 2.2, 2.3), comprise the majority of Acrosiphonia plants in southwestern British Columbia. Species identification was based on a number of morphological / microscopic characters studied by Hudson (1974) for plants in the Puget Sound region and commonly used by other workers (Kjellman 1893, Collins 1909, Setchell and Gardner 1920, Scagel 1966). Those criteria I deemed dependable for distinguishing between species are thallus morphology, filament diameter, hooked branchlet type and number of fertile cells in a branch (Table 2.1). The D N A sequences of the ITS regions of specimens identified as A. arcta and A. coalita showed 40 Figure 2.2 A . Acrosiphonia arcta at Burrard Inlet. B. Acrosiphonia arcta and A. coalita at Bamfield. C . E. A. coalita at Sooke. F . A. arcta at Sooke. D. A. arcta and A coalita at Sooke. Scale bar is 5 cm. arc = A. arcta co = A. coalita. 41 Figure 2.3 A . Acrosiphonia coalita habit B. Acrosiphonia arcta habit. C . Fertile cells. D. Empty gametangium with pore for gamete release. E . Compound hook of A. coalita. F . Simple hook. G . Many cells fertile in a series, A. arcta. H . Single cells fertile in A. coalita branches. Scale bar is 200 um. Py = pyrenoids; gm = gametangium. consistent base pair differences, thus supporting their status as distinct molecular / morphological species (see Chapter 1 Fig. 1.4). Juveniles were difficult to distinguish both in the field and laboratory, because the branching pattern of A. arcta was generally not evident until plants were greater than 2 cm in height, and compound hooks (a hooked branchlet upon another hooked branchlet) tended not to be developed on very young A. coalita. Table 2.1 Distinguishing characters for Acrosiphonia arcta and Acrosiphonia coalita in southwestern British Columbia. A. arcta A. coalita Thallus length 3-6-(10) cm 12-20-(35) cm Habit very bushy; much branched; forming hemispherical to globose tufts tends to form long rope-like branches Texture soft coarse Hooked branches simple; rarely compound simple and compound Filament diameter 60-100-(120) um 120-230 um # cells fertile in series 3 - 5 - (entire branch) 1-2 (less common) Reproductive phenology April (less common) -May March - April (less common) The Acrosiphonia specimens from Alaska, whose ITS D N A sequences were 100 percent identical to a number of'Chlorochytrium' isolates (see Chapter 1, Fig. 1.4), shared similar characters (absence of hooked branchlets and filament diameter of 160-200 um) with UBC herbarium specimen A83088 obtained from Prince William Sound, Alaska and identified as A. 43 hystrix. Although the A. sp. (Alaska) / 'Chlorochytrium' D N A matches found in the molecular study suggest the presence of A. sp. (Alaska) in southwestern British Columbia, no Acrosiphonia plants with this combination of characters were found at Burrard Inlet, Sooke or Bamfield. Interestingly, A. hystrix is not reported to occur south of Alaska (Setchell and Gardner, 1920), and was in fact not included in the five Acrosiphonia species originally identified by Scagel (1966) for British Columbia. One of the Acrosiphonia specimens collected at Sooke, which was very closely related to A. sonderi based on ITS sequences (see Chapter 1, Fig. 1.5), was not morphologically distinguishable as a separate species, nor are descriptions by Kornmann (1962) adequate for identification. Acrosiphonia distribution and seasonal abundance. At Burrard Inlet only Acrosiphonia arcta was present. This was supported by ITS D N A sequences (see Chapter 1 Fig. 1.4). Plants were found on boulders and as epiphytes on Fucus in the low and mid-intertidal zone (Fig. 2.2A). Smaller plants grew on the base ofthe Stanley Park seawall (5.1m above zero tidal level (Canadian Chart Datum) where freshwater runoff provided a moist habitat. Distribution was patchy at all tidal heights. Bright green Acrosiphonia juveniles first appeared sparsely distributed in early March 1997 (sampling day 100 from start, Fig. 2.4A). Although growth rate was not measured, the average size and abundance of plants present on consecutive sampling dates suggests rapid growth. Abundance peaked in April at 12.7 percent cover, with a small proportion ofthe thalli fertile at this time. In Acrosiphonia vegetative intercalary cells located either in side branches or in the main filament ofthe plant differentiate into gametangia. These gametangia change from dark green to brown as they reach maturity (Figs. 2.3C, F-H), and evidence of gamete release is seen by empty gametangia (Fig. 2.3D). Here I refer to fertile thalli as those where brown gametangia were visible. In April some plants were already covered with diatoms and darker green; others were bleached. By late May / June most plants were heavily epiphytised and bleached, and fifty percent of plants sampled were fertile. On several occasions littorinid snails were observed on Acrosiphonia plants. In late June (sampling day 209 from start, Fig. 2.4A) Acrosiphonia was absent from the intertidal zone and seawall, having been replaced by Enteromorpha and Ulva as the predominant algae. In 1998 Acrosiphonia peak percent cover was twice as high as in 1997. Burrard Inlet B. l i , p 0.8 £ 0.6 H. t ° - 4 ^ e u Q 0.2 'Chtorochytrium' 100 200 300 400 500 spring '97 fall '97 spring Sampling Day from Start 800 Figure 2.4 A . Acrosiphonia percent cover at Burrard Inlet over two years. Data are means ± S.E. from 30 quadrats placed along transect lines. The S.E. are missing for two data points, due to missing data (theft of field notes). B . 'Chlorochytrium' densities at Burrard Inlet. Data are means ± S.E. from cells counted in 30 Mazzaella splendens blades. 45 At Sooke and Bamfield both Acrosiphonia arcta and A. coalita were present, and showed patchy distribution (Fig. 2.2B-D). Examination of total Acrosiphonia percent cover at Sooke (from all three transects) revealed a pattern of seasonal abundance similar to Burrard Inlet (Figs. 2.4A, 2.5A). Acrosiphonia plants at Sooke, however, persisted much longer than at Burrard Inlet. Juvenile plants were present early March (sampling days 117 and 480 from start), percent cover peaked in April (approximately 9 percent) and by July / September Acrosiphonia was no longer present. In 1997 a number of healthy thalli were collected at Sooke in mid August, whereas Burrard Inlet plants had disappeared two months earlier. This difference was much less pronounced in 1998. By late July only a few badly deteriorated remnants were found at Sooke. Also observed at Sooke was the occurrence of a second pulse of Acrosiphonia establishment. In both sampling years by mid May mature Acrosiphonia percent cover was drastically reduced and replaced by Ulva and Microcladia, but juvenile Acrosiphonia of 2-3 percent cover was present. A second percent cover peak oi Acrosiphonia adults was observed in July 1997 and to a much lesser extent in June 1998. Overall, Acrosiphonia arcta and A. coalita showed the same pattern of seasonal abundance. The two species did, however, differ in seasonal abundance and dominance among the three different intertidal heights. Nonetheless, lack of clear patterns over the two year study, compounded by the difficulty of distinguishing among species in the field, makes it virtually impossible to separate Acrosiphonia natural dynamics for each species. It was noted, though, that A. coalita and A. arcta initially occupied different intertidal zones: A. coalita was found on boulders and sand and epiphytic on eelgrass in the low intertidal zone; A. arcta was found on the same substrata higher in the intertidal zone. Within one month the two species were observed to coexist throughout the intertidal zone (Fig. 2.2D). Yet A. arcta always remained the dominant species in the high intertidal zone, even though it became more abundant in the low intertidal zone later in the summer. Reproductive phenology was also different for A. arcta and A. coalita. Twenty percent of A. coalita plants were fertile in early March, with 100 percent reproductive plants sampled in mid April (and variable numbers through to July / August). Fertile A. arcta plants (approximately 30 percent), on the other hand, were not found until the end of April (variable percent reproductive plants were recorded through to July / August). Acrosiphonia at Sooke was also observed on a large boulder (about 2 m in height), located in the low intertidal zone. This boulder showed distinct zonation: the middle zone Sooke Acrosiphonia 0 100 200 300 400 500 600 700 800 spring '97 fall '97 spring '98 fall '98 B. 'ChlorochytTiurri 80 -i spring'97 fall'97 spring'98 fall '98 Sampling Day from Start Figure 2.5 A . Acrosiphonia percent cover at Sooke over two years. Data are means ± S.E. from 30 quadrats placed along transect lines. The S.E. are missing for three data points, due to missing data (theft of field notes). B. 'Chlorochytrium' densities at Sooke. Data are means ± S.E. from cells counted in 30 Mazzaella splendens blades. 47 containing Acrosiphonia; the upper dominated by mussels and barnacles and the lower by Mazzaella splendens and kelps such as Alaria, Costaria and Hedophyllum. The plants resembled A. arcta morphologically, but did not become established until mid May, and also seemed to remain bright green with few epiphytes longer than other A. arcta. The ITS D N A sequences of several isolates indicated the majority of the plants on the boulder were A. arcta. One specimen, however, yielded a unique D N A sequence (see Chapter 1, Fig. 1.4). Although fewer data are available for Bamfield, it is evident that seasonal abundance of Acrosiphonia (Fig. 2.6A) follows the pattern elucidated for Sooke and Burrard Inlet. Sampling in 1997 failed to detect recruitment of Acrosiphonia, but the following year revealed establishment at the end of March (sampling day 515 from start), some 3-4 weeks later than at Sooke. A. arcta and A. coalita were found growing together on consolidated rock (Fig. 2.2B). Their distribution was much patchier than at Sooke and Burrard Inlet with peak percent cover only reaching 4.1 percent in 1997. In 1998, however, as was seen at Burrard Inlet, Acrosiphonia increased in abundance, with peak abundance at 8.4 percent. Peak abundance may have occurred later than at Burrard Inlet and Sooke. Since no sampling was carried out between March 26 and May 26 in both years (for logistical reasons), it is difficult to ascertain whether higher abundance may have been observed at an earlier date. Seasonal dominance of one species over another was variable; as at Sooke no clear pattern was illustrated. Persistence of healthy plants was observed well into July (no sampling was conducted in August) for both 1997 and 1998, unlike at Sooke where 1998 plants disappeared earlier. A. arcta and A. coalita became reproductive later than at Sooke (A. coalita becoming mature first). Also, similarly to the large boulder at Sooke, a rock wall at Bamfield showed later establishment and longer presence of plants. Rather than A. arcta, though, it was primarily A. coalita which occupied this rock wall. As Acrosiphonia plants disappeared from the intertidal zone, Cladophora visibly dominated. Biological or statistical significance? Figures 2.4A-2.6A portray the seasonality of Acrosiphonia at Burrard Inlet, Sooke and Bamfield: early March to June / July (August). This seasonality is statistically significant (p < .05, Table 2.2). However, statistically significant differences among the data points are not easily detectable by the Games-Howell post hoc test. At all sites, though, for both 1997 and 1998, one or more spring / summer peak percent covers are significantly higher (p < 0.05) than percent covers during the period of Acrosiphonia establishment or disappearance. Smaller differences in percent cover such as increased 48 Bamfield Acrosiphonia 14 i 12 -spring'97 fall '97 spring'98 fall '98 Sampling Day from Start Figure 2.6 A . Acrosiphonia percent cover at Bamfield over two years. Data are means + S.E. from 30 quadrats placed along one or two transect lines. B. 'Chlorochytrium' densities at Bamfield. Data are means ± S.E. from cells counted in 30 Mazzaella splendens blades. 49 abundance of Acrosiphonia from one year to another (Figs. 2.4A, 2.6A) or a second pulse of Acrosiphonia establishment at Sooke (Fig. 2.5A) were, however, not confirmed by the post hoc test. These phenomena are visible in the graphs and were very noticeable in the field, suggesting that the extreme patchiness of Acrosiphonia, and hence high variance in the data, may have resulted in failure to detect such differences. In any case, Acrosiphonia's seasonality has clearly been established. Table 2.2 Acrosiphonia percent cover. Results of one-way A N O V A on data shown in Figures 2.4A-2.6A and Figures 2.11 A-2.13A. The significant F value indicates a statistically significant difference among two or more sampling dates. Percent cover SS df MS F p Burrard Inlet 352.638 9 39.182 19.963 < 0.0001 Sooke 214.751 11 19.523 9.403 < 0.0001 Bamfield 76.030 6 12.672 8.909 < 0.0001 'Chlorochytrium' morphology and location in host. 'Chlorochytrium' cells were found embedded among the cells of Mazzaella splendens from Burrard Inlet, Sooke and Bamfield. In transverse section they were consistently located in the cortical layer ofthe blade (Fig. 2.7B). Cell morphology ranged from spherical to ovoid to unusually elongated (Fig. 2.7 C,D,E), evidently due to the surrounding host tissue differentially impeding growth. These unusually elongate cells were found in very thin tetrasporophytic blades collected from Sooke in January, February and November 1998. Spherical cells comprised the majority of cells and were grouped into the following six size classes: 40-80 um, 80-120 um, 120-160 um, 160-200 um, 200-240 jam and > 240 um diameters. The largest cells observed were 300 um. The unusually elongate cells were 10-20 percent longer than the diameter ofthe largest spherical cells observed. A net-like chloroplast and numerous pyrenoids were visible in vegetative cells (Fig. 2.7). Based on a small number of observations in blades collected from Sooke late February 1998 and November 50 1998 blades (also from Sooke) maintained in a seawater tank for observation, fertile cells tended to (1) form protuberances which extended toward the surface of the host (Fig. 2.7F), (2) darken at the apex, (3) change from bright green to olive coloured, and (4) became homogeneous and bumpy with chloroplasts and pyrenoids no longer distinguishable. A single 'Chlorochytrium' cell, 160 um in diameter, from material collected late February was observed releasing > 800 zoospores (Fig. 2.7G). The zoospores were approximately 5 um in diameter, possessing red eyespots and four equal length flagella. Other than the fact that large 'Chlorochytrium' cells tended to be concentrated at the base of numerous M. splendens blades collected from late August to January, no distinct pattern of distribution on blades was detectable. Cells were either scattered over the entire blade surface, grouped in patches along blade margins or at the apex or base of the blade or some combination of patches and scattering of individual cells. 'Chlorochytrium' seasonal abundance. 'Chlorochytrium' seasonality (Figs. 2.4B-2.6B) is statistically significant (Table 2.3). At Sooke 'Chlorochytrium' abundance peaked in May (sampling days 177 and 545 from start, Fig. 2.5B), approximately one month after establishment of Acrosiphonia. 'Chlorochytrium' densities from September to late March are significantly lower (p < 0.05) than densities in April / May 1997 (sampling days 177 / 204 from start) and June 1998 (sampling day 576), and coincide with Acrosiphonia absence. No other significant differences were detected when the post hoc test was performed on square root transformed data. At Burrard Inlet and Bamfield peak densities in 1998 of 0.70 cells cm"2 or 81.62 cells per blade and 14 cells cm"2 or 2060 cells per blade, respectively, were lower than at Sooke (53 cells cm"2 or 5573 cells per blade). Very few 'Chlorochytrium' cells were found in Mazzaella splendens at Burrard Inlet in 1997 (Fig. 2.4B). The following year, however, a significantly higher number of cells (p < 0.05) were present from June to August (sampling days 563-623), six weeks after Acrosiphonia establishment. Abundance of 'Chlorochytrium' at Bamfield was also extremely low in 1997 (Fig. 2.6B). The three sampling dates October 27/96 (day 1), July 21/98 (day 632) and October 6/98 (day 710) show significantly higher densities (4th root transformations of data, p < .05) than all other sampling days. It is interesting to note that 'Chlorochytrium' peak density at Bamfield (at least for 1998) occurred much later than at Burrard Inlet or Sooke. 51 Figure 2.7 A . Surface view of 'Chlorochytrium inclusum' cells (80-160 um) in Mazzaella splendens. B. Transverse section showing 'Chlorochytrium' in cortex region of M. splendens. C. D. E . Vegetative 'Chlorochytrium' cells. F. 'Chlorochytrium'cell near maturity. G . 'Chlorochytrium' cell releasing > 800 zoospores. Scale bar is 50 um. py = pyrenoids. 52 Table 2.3 'Chlorochytrium' density. Results of one-way A N O V A on data shown in Figures 2.4B-2.6B. The significant F value indicates a statistically significant difference among two or more sampling dates. Density SS df MS F P Burrard Inlet 19.890 10 1.989 9.501 < 0.0001 Sooke 633.455 19 33.340 5.534 < 0.0001 Bamfield * 1.731 15 0.115 7.263 < 0.0001 * 4th root transformations were carried out When absolute numbers, rather than densities of 'Chlorochytrium', were used to generate data points, the same seasonal abundance patterns resulted for all three study sites. 'Chlorochytrium' size. Average and maximum 'Chlorochytrium' cell sizes recorded for each sampling date were used to generate a 'Chlorochytrium' life history schematic (Fig. 2.8). The schematic is based on Sooke data since it is the most complete (frequent sampling and occurrence of high 'Chlorochytrium' densities). Smallest cells (average 40 um) were observed in greatest abundance in early spring. Throughout the spring and summer average cell size gradually increased to 200 (im, evidently due to growth of cells. Average cell size decreased in late summer with the observation of a large number of small cells, and then increased again in the fall and winter as cells grew. The following spring only small cells were present with the largest cells 120 um. During all other seasons maximum cell size was > 120 um, and remained > 280 nm from June (sampling day 239 from start) to February (sampling day 480). Schematics for Burrard Inlet and Bamfield would differ from the Sooke schematic, because high densities of the smallest cells were not observed in late summer. The average cell size curve would therefore not decrease in late summer (Fig. 2.8), but instead remain fairly level until spring, at which point the curve would again coincide with the Sooke schematic. A n interesting phenomenon not evident in the schematic is the presence of small cells still observed in December 1997 and January 1998 at Sooke and Bamfield, long after Acrosiphonia has disappeared. 53 ' » i ' 1 1 1 1 I T"™ " " T I I "I 165 239 324 439 534 590 spring '97 summer '97 fall '97 winter '97 spring '98 Sampling Date (Days from Start) Figure. 2.8. 'Chlorochytrium' life history schematic based on Sooke cell size observations over two years. The solid line represents average cell size and the dotted line maximum cell size. 'Codiolum' morphology and location in host. 'Codiolum' cells were found attached to and completely embedded in the filamentous system of 'Petrocelis franciscana' (Fig. 2.9). They are generally distinguished from 'Chlorochytrium' by their differentiation into an ovoid vesicle and colourless stalk. However, extreme polymorphism exhibited by individual cells included 'Codiolum' cells where the stalk was lacking. In general the ovoid vesicle was slender (120-160 um x 30 um, 50-80 um x 15 um) or fat (100 um x 60 um, 80 um x 40 um, 50 um x 30 um); the stalk relatively long (70-100 um), short (20-50 um) or non-existent (Figs. 2.9, 2.10). Infrequently 'Codiolum' cells were seen where a secondary stalk was developing from the primary stalk (Fig. 2.10). Vegetative cells varied from yellow-green to bright green. The presence of oil droplets obscured any internal structures, e.g. chloroplast and pyrenoids. Although no fertile cells were observed discharging zoospores, mature 'Codiolum' were 54 Figure 2.9 Polymorphism of 'Codiolum petrocelidis'. A . B. C . E. show cells among 'Petrocelis' filaments. Note both fertile 'Codiolum' and fertile 'Petrocelis' in A . Cells with long stalks are shown in D and E. Juvenile 'Codiolum' are represented by G and I. Large fertile cells are shown in A . C . H. Scale bar is 50 um. Tet = tetrasporangium. Figure 2.10 Polymorphism of 'Codiolum petrocelidis'. Note the large, fertile cells filled with zoospores in the third row. The last cell in the first row is believed to represent the youngest cells. 56 identified by darkening of the cell and division of the entire contents of the cell into spores approximately 5 um in length. Fertile cell vesicles were generally > 80 um in length and 40 um in width (Figs. 2.9, 2.10). Reproductive phenology of 'Codiolum' coincided with that of 'Petrocelis', from late fall to spring (Fig. 2.9A). 'Codiolum' seasonal abundance. 'Codiolum' density curves for Burrard Inlet (Fig. 2.1 IB) and Sooke (Fig. 2.12B) show clear patterns of seasonality supported by A N O V A s (Table 2.4) and post hoc tests (Figs. 2.1 IB, 2.12B). Unlike 'Chlorochytrium' abundance, 'Codiolum' densities at Sooke and Burrard Inlet peaked late June in 1997 (sampling days 234 and 209 from start, respectively) and in July of 1998 (days 619 and 593, respectively). This is approximately one to two months later than 'Chlorochytrium' peaked and two to three months after Acrosiphonia establishment. At Sooke peak 'Codiolum' densities were 17,900 cells cm" of 'Petrocelis' in 1997 and 22,400 cells cm" of'Petrocelis' in 1998; there was no significant difference (p > 0.05) between the two years. At Burrard Inlet, however, 'Codiolum' peak density of 16,500 cells cm"2 of'Petrocelis' in 1997 (comparable to Sooke peak densities) was significantly lower than the 1998 peak density of 99,100 cells cm"2. At Bamfield the pattern of'Codiolum' seasonality is not clearly elucidated (Fig. 2.13B), and few significant differences in densities were detected. This may in part be due to less sampling and smaller sample sizes for some data points ('Petrocelis' at Bamfield was often difficult to distinguish from other algal crusts, because they tended to exist together, overlapping with one another.) Yet, it is evident that much lower densities of 'Codiolum' colonised 'Petrocelis' at Bamfield than at Sooke and Burrard Inlet. Even peak density in October 1998 (last sampling day), which is significantly higher than any other sampling day for Bamfield, is almost 10 times lower than peak densities at the other study sites. Furthermore, peak density in 1997 does not exceed 255 cells cm" , and, due to high variance, the mean is not statistically different from zero. The October 1998 peak density occurs at least four months after Acrosiphonia establishment and three months later than 'Codiolum' peak densities at Sooke and Burrard Inlet. Endophyte density curves suggest 'Codiolum' peak densities were maintained much longer than 'Chlorochytrium' peak densities (Figs. 2.5B, 2.1 IB, 2.12B). Burrard Inlet Figure 2.11 A. Acrosiphonia percent cover at Burrard Inlet over two years. Data are means ± S.E. from 30 quadrats placed along transect lines. The S.E. are missing for three data points, due to missing data (theft of field notes). B . 'Codiolum' densities at Burrard Inlet. Data are means ± S.E. from cells counted in 30 'Petrocelis' patches. Sooke Acrosiphonia 0 100 200 300 400 500 600 700 800 spring '97 fall '97 spring '98 fall '98 B. ,Codiollrm, 30000 - i spring'97 fall '97 spring'98 fall '98 Sampling Day from Start Figure 2.12 A. Acrosiphonia percent cover at Sooke over two years. Data are means ± S.E. from 30 quadrats placed along transect lines. The S.E. are missing for three data points, due to missing data (theft of field notes). B. 'Codiolum'densities at Sooke. Data are means ± S.E. from cells counted in 30 'Petrocelis' patches. 59 Bamfield B. 'Codiolum' 4000 -j 3500 -spring '97 fall '97 spring '98 fall '98 Sampling Day from Start Figure 2.13 A . Acrosiphonia percent cover at Bamfield over two years. Data are means ± S.E. from 30 quadrats placed along one or two transect lines. B . 'Codiolum' densities at Bamfield. Data are means ± S.E. from cells counted in 30 'Petrocelis' patches. 60 Table 2.4 'Codiolum' density. Results of one-way A N O V A on data shown in figures 2.11B-2.13B. The significant F value indicates a statistically significant difference among two or more sampling dates. Density SS df MS Burrard Inlet Sooke Bamfield 3706916.3 1145984.4 53047.159 22 22 15 168496.194 44.246 < 0.0001 52090.202 44.846 < 0.0001 3536.477 34.687 < 0.0001 Sample sizes, means, standard deviations and standard errors for untransformed Acrosiphonia, 'Chlorochytrium' and 'Codiolum' data are provided in Appendices A , B and C. 'Codiolum' size. Due to the extreme polymorphism exhibited by 'Codiolum' cells, it was not possible to monitor cell size changes the same way as for 'Chlorochytrium' cells. Instead, cell size for each sampling date was based on vesicle size only and recorded as either large ( > 80 um in length) or small (< 80 um in length). The 'Codiolum' life history schematic in Figure 2.14 was generated from data from all three study sites, and shows that large cells (many of them fertile) were consistently most abundant during fall and winter, whereas only small cells were found in 'Petrocelis' patches in the spring. Summer sampling revealed variable numbers of large and small cells, but no fertile cells. Surprisingly, as was the case for 'Chlorochytrium' from Sooke and Bamfield, many small 'Codiolum' cells were observed in the winter of 1997 at Sooke and Burrard Inlet, long after Acrosiphonia's disappearance. 61 Large Cells Large Cells FERTILE / \ FERTILE Small / Large / Cells / / \ \ Small Cells 7 Colonization F a l l / W i n t e r S p ri n g / S u m m e r F a l l / W i n t e r Figure 2.14 'Codiolum' life history schematic based on observations of cell size and reproductive state of cells over two years at Burrard Inlet, Sooke and Bamfield. DISCUSSION The relationship between Acrosiphonia's gametophyte and sporophyte in nature is illustrated in Figure 2.15. Filamentous Acrosiphonia plants are relatively short-lived, abundant during spring and summer (Figs. 2.4A, 2.6A). Growth is rapid and fertile cells develop almost immediately after Acrosiphonia is established. Observation of empty gametangia indicates release of gametes throughout the spring and summer. The unicellular endophytic sporophytes, 'Chlorochytrium' and 'Codiolum', colonise Mazzaella splendens and 'Petrocelis', respectively one to three months after Acrosiphonia establishment (Figs. 2.4B-2.6B, 2.11B-2.13B). Colonisation inferred from cell size may occur primarily as two major events (in spring and summer, Fig.2.8) corresponding to two Acrosiphonia pulses or continuously over the spring and summer. Growth of 'Chlorochytrium' and 'Codiolum' appears to be rapid during this time. Endophytes mature primarily in winter, and zoospore release must occur in winter and / or spring: large fertile 62 'Codiolum' cells were consistently found in fall / winter, but were no longer present the following spring (Fig. 2.14); several fertile 'Chlorochytrium' cells were observed in February, and maximum 'Chlorochytrium' cell size decreased substantially from winter to spring (Fig. 2.8). Germination of zoospores in spring and summer gives rise to the filamentous gametophytic plants and completes Acrosiphonia's life cycle. Gamete release Spring / Summer SPOROPHYTE Acrosiphonia establishment Early spring / Summer Endophyte colonisation Spring / Summer G A M E T O P H Y T E Zoospore release Winter / Spring Figure 2.15 Timing of events of Acrosiphonia's life history in nature. Acrosiphonia identification in the field. Distinguishing among Acrosiphonia species in the field is problematic. As already mentioned, it is proposed that the two morphological species A. arcta and A. coalita (which consistently show 4 or 5 bp differences in their ITS D N A sequences, Chapter 1) comprise the majority of plants at the three study sites. However, as also previously noted, juvenile plants are difficult to identify in the field. Furthermore, some mature A. arcta plants were found without any hooked branchlets. Others were found with short branchlets beginning to form on simple hooked branchlets and so resembled compound hooks, a diagnostic character of A. coalita. Hudson (1974) concluded that the presence or absence of hooked branchlets is a character which should not be given a great deal of weight in distinguishing among plants with otherwise similar appearance. She suggested that the presence 63 or absence of hooks on plants growing in apparently the same conditions may be due to 1) the plants being genetically the same but growth conditions actually sufficiently different to produce a morphological difference or 2) formation of hooks being sporadic. Furthermore, light intensity and day length were found to affect such characters as cell diameter, length to width ratios and amount of branching (Kornmann 1965, Hudson 1974). The Alaskan Acrosiphonia specimens, found to lack hooked branchlets and have large cell diameters, may then not be distinct from A. arcta (with small cell diameter). Indeed, Setchell & Gardner (1920) claimed A. hystrix, resembling the Alaskan specimens, to be a high northern species of the arcta group, differing primarily from A. arcta in the greater diameter of filaments. The Acrosiphonia specimen similar to A. sonderi based on ITS sequences (see Chapter 1), was indistinguishable morphologically, again suggesting the diagnostic morphological characters on which species are defined are too variable and responsive to environmental conditions to adequately diagnose Acrosiphonia species. The ITS D N A sequences supported the recognition of A. arcta and A. coalita as species, but too few informative sites were present in the ITS region to resolve any further infrageneric relationships within Acrosiphonia. I will reiterate that there is great need for a molecular study utilising a more variable region to resolve the taxonomic confusion plaguing the genus Acrosiphonia. Furthermore, clarification ofthe number of Acrosiphonia species and identification of their sporophytes at a given site, along with documentation of seasonal abundance and reproductive phenology, may make a more detailed account of Acrosiphonids life history in the field possible. Acrosiphonia seasonality. Seasonal abundance of Acrosiphonia differed among A. coalita and A. arcta and among the three study sites, but seasonality was consistent with collections reported in the literature. Hudson (1974) obtained fertile^, arcta and A. coalita plants from the Puget Sound region, Washington State throughout the spring and summer of 1970-73, and Fan (1959) found A. coalita at Moss Beach, California from April to October in 1957. In Europe Kornmann (1970a, 1970b) collected A. arcta and A. grandis (not reported from North America) from Helgoland, Germany in the spring and summer of 1969 and 1970, and Jonsson (1959a, 1964, 1986) obtained^, arcta from Brittany, France in the spring of 1958,1964 and 1986. It is not unusual that A. coalita plants were found in October in California, because specimens were also found during this study in early October 1997 from Seppings Island (near Bamfield), Barkley Sound. 64 Factors affecting Acrosiphonia seasonal abundance and distribution. Temperature and photoperiod are important factors in the germination and growth of Acrosiphonia. Hudson (1974) found that in culture upright filaments of Acrosiphonia sprouted from rhizoid pieces under long-day (16:8 photoregime) conditions, but not under short-day (8:16) conditions at the same temperatures. Also, at 15 °C A. arcta plants grew less well than at 5 °C and 10 °C under the same photoregime, and sporophytes ('Chlorochytrium' or 'Codiolum') died before division into zoospores could occur. Miyaji (1996) found this to also be true for A spiralis plants collected in Japan. Vegetative growth of A coalita, on the other hand, was only inhibited at 20 °C. Furthermore, when both A. arcta and A. coalita plants were grown under low light intensity (below 9.2 pmol m"2 s"1), motile gametes were not produced. These studies suggest longer days in the spring trigger germination of Acrosiphonia, increased light intensity plays a role in the production of viable gametes and the effects of high summer temperatures contribute to the mortality of Acrosiphonia. It also seems reasonable to conclude that afternoon low tides in summer, resulting in additional effects of heat and desiccation, are responsible for the earlier disappearance of Acrosiphonia plants at Burrard Inlet than at Sooke and Bamfield (Figs. 2.4A-2.6A) where low tides occur in the morning. In 1999, after sampling had ceased, Acrosiphonia was seen at Burrard Inlet (although at a new site) several weeks later than its disappearance in 1997 and 1998. This may be attributed, at least in part, to the lack of sunny days in June and July 1999 (pers. observ.), and so would imply local weather conditions also affect Acrosiphonia's seasonal abundance. It may then not be surprising that Acrosiphonia disappeared at Sooke almost a full month earlier in 1998 than the previous year (Fig. 2.5A), since considerably less rainfall than usual was recorded (Environment Canada, Climate Data Services) the summer of 1998 (an E l Nino year). Yet, that same year Acrosiphonia at Bamfield did persist as in the previous year, and, as mentioned, specimens were collected from Seppings Island (near Bamfield) in early October. This, and the fact variable abundance of Acrosiphonia occurred from one year to the next among study sites, e.g. increased abundance at Burrard Inlet and Bamfield in 1998, but no difference recorded for Sooke between 1997 and 1998 (Figs. 2.4A - 2.6A) suggests micro-habitat factors such as salinity, nutrient levels and herbivore species and abundance also play a significant role in Acrosiphonia's seasonal abundance. The greater abundance of Acrosiphonia observed at Burrard Inlet and Bamfield in 1998 may be explained by 1) increased survivorship of endophytes (but not necessarily high densities, 65 since low numbers of both 'Codiolum' and 'Chlorochytrium' were recorded for Bamfield in 1997) 2) increased dispersal of endophytes and drift reproductive Acrosiphonia thalli and 3) increased survivorship of overwintering rhizoids of Acrosiphonia. As suggested by Scagel (1966), it is still unproven whether Acrosiphonia gametophytes can overwinter by means of rhizoids and, i f they do, how important this is to the maintenance of Acrosiphonia's population. Purely vegetative propagation by rhizoids was shown under short-day conditions in culture (Hudson, 1974), yet the following observations lead me to speculate that overwintering via rhizoids may not be very important in Acrosiphonia's life history at my study sites: 1) establishment of Acrosiphonia in new areas from year to year at all three study sites, 2) 0 percent cover of Acrosiphonia found in preliminary 20 cm x 20 cm marked areas formerly noted with 75 - 80 percent cover at Bamfield and 3) the inability to detect rhizoids after disappearance of the upright plant. Dispersal patterns of gametes, reproductive spores or propagules of marine algae have been little studied (Anderson & North 1966, Amsler & Searles 1980, Reed et al. 1988, Kendrick & Walker 1991, 1995). Acrosiphonia arcta plants (though lacking hooked branchlets) were found on Kitsilano Beach, Vancouver, B.C. in May 1997, despite the fact that Mazzaella splendens and 'Petrocelis' were absent. In fact, the only foliose algae present were Laminaria saccharina (L.) Lamouroux and Ulva, neither of which were found to be colonised by 'Chlorochytrium', and no crustose algae were found. This finding supports the idea that dispersal of unicellular sporophytes, zoospores or reproductive Acrosiphonia thalli may be long range, e.g. from Burrard Inlet (approximately 6 km). This also implies the survival and overwintering of free-living sporophytes, perhaps in rock crevices. A number of researchers (Jonsson 1959a 1959b 1962 1966, Kornmann 1961 1964, Chihara 1969, Hudson 1974, Miyaji 1996) demonstrated the ability of'Chlorochytrium' and 'Codiolum' to survive free-living in culture. The epilithic existence of Acrosiphonia's sporophyte in the field could be tested experimentally by placing sterilised plates in the intertidal zone among fertile Acrosiphonia in the summer, and examining these for settlement and growth of the unicellular sporophyte later in the year. Furthermore, unlike at my study sites where hosts are abundant for the survival of Acrosiphonia's sporophyte, vegetative propagation of Acrosiphonia by overwintering rhizoids may be important in maintenance of populations in habitats (such as at Kitsilano Beach, Vancouver) devoid of hosts for 'Chlorochytrium' and 'Codiolum'. 66 Although not statistically significant, both Acrosiphonia arcta and A. coalita showed a second pulse at Sooke in both sampling years, i.e. after Acrosiphonia abundance dropped off, percent cover peaked a second time. This was much less pronounced in 1998 (an El Nino year), perhaps due to lower than normal precipitation (Environment Canada, Climate Data Services 1998) resulting in increased desiccation and hastening mortality of plants. Possible explanations for two Acrosiphonia pulses at Sooke include: 1) delayed germination of zoospores of another 'species' of Acrosiphonia, 2) differential maturity of sporophytes resulting in differential release of zoospores, 3) parthenogenesis of unfused gametes germinating directly into gametophytes, 4) zoospores developing within Acrosiphonia filaments and germinating directly into gametophytes and 5) vegetative propagation from rhizoids of old plants. I was unable to determine whether a second pulse occurred at Bamfield and at other locations where Acrosiphonia seasonality extends beyond July. Parthenogenesis (Jonsson 1963, Miyaji 1996) and recycling of filamentous plants by zoospores (Kornmann, 1962) has been observed in A. arcta, A. sonderi and A. spiralis in culture. However, sexual reproduction was by far the dominant mode of propagation, and whether parthenogenesis and asexual reproduction occur in nature remains to be demonstrated. Acrosiphonia arcta and A . coalita differences in the field. The two Acrosiphonia species, A. arcta and A. coalita, occupy the same habitat, but showed variable patterns of abundance and reproductive phenology at Sooke and Bamfield. Acrosiphonia coalita was absent from Burrard Inlet. Hudson's culture work (1974) showed different effects of environmental factors such as temperature on the two species: A. coalita seems to have a higher tolerance for high temperatures. Hudson (1974)noted that A. coalita persisted longer than A. arcta in the intertidal zone ofthe Puget Sound region, Washington State, and that A. arcta had not been reported south of Washington State. Although A. arcta did not consistently disappear from the intertidal zone earlier than A. coalita at Sooke and Bamfield, A. arcta's abundance decreased at Sooke in the high intertidal zone and increased in the low intertidal zone as the summer progressed. However, A. arcta was always the dominant species higher in the intertidal zone, suggesting that in southwestern British Columbia temperature tolerance alone does not determine species distribution and abundance. The absence of A. coalita from Burrard Inlet may be due to a lack of unidentified micro-habitat requirements. On the other hand, A. coalita may simply have never become established. One way to test for this is to plant A. coalita in the intertidal zone at Burrard Inlet. Identification 67 of micro-habitat requirements may best be undertaken in the laboratory with more extensive culture studies. Contrary to observations in the field, in culture A. arcta became fertile before A. coalita (Hudson, 1974). At both Sooke and Bamfield fertile A. coalita plants were found almost immediately after establishment of plants, whereas fertile A. arcta (also at Burrard Inlet) were not detected until the following sampling date, about one month later. Since sampling only occurred monthly, it is possible that A. arcta may actually have become reproductive less than one month after A coalita. Furthermore, the percentage of reproductive plants was higher for A coalita throughout the sampling period. More detailed studies are necessary to pin-point differential reproductive phenology. Endophyte colonisation. Colonisation of the sporophytic endophytes was deduced by examining both cell density and cell size over time. An important assumption is that the smallest cells represent the youngest cells. Polymorphism of 'Codiolum' cells complicates this assumption, and so cells were categorised on the basis of fertility and vesicle length (greater or less than 80 um) regardless of stalk length. Based on cell density increase (and the monthly sampling regime), the onset of 'Chlorochytrium' colonisation occurred in late April about one month after Acrosiphonia establishment (Figs. 2.4B-2.6B), and one month earlier than for 'Codiolum' (Figs. 2.11B-2.13B) at all three study sites. This is somewhat counter-intuitive, since A coalita became fertile before A arcta, and its sporophyte appears to be 'Codiolum' based on culture and molecular work (see Chapter 1). The onset of endophyte colonisation was also implied by the occurrence of the greatest abundance of smallest cells (Figs. 2.8 and 2.14). Kornmann's (1964) findings (within Haemescharia hennedyi from Germany) agree with the present study in that the greatest abundance of young 'Codiolum' cells occurred in late spring, four to six weeks after reproductive Spongomorpha lanosa plants were detected. The presence of smallest cells from spring to early fall in the present study implies colonisation is continuous and synchronised with Acrosiphonia gamete release. Dethier's (1987) finding of high densities of 'Codiolum' within 'Petrocelis' from Washington State in summer supports spring / summer colonisation. The time required and the mechanism for zygote / endophyte penetration oi Mazzaella splendens and 'Petrocelis' remains poorly understood. More detailed culture and microscopy studies are 68 needed to shed light on these events, and to better understand the time-lag between Acrosiphonia gamete release and endophyte establishment. It was possible to detect two major colonisation events by endophytes at Sooke (Fig. 2.8), which coincide with the two Acrosiphonia pulses observed at Sooke (Fig. 2.5 A). However, due to an inability to obtain average 'Codiolum' cell size (categories were simply large or small) and the difficulty in distinguishing relative densities of newly colonised cells from older cells because of polymorphism, it is unlikely that two events for 'Codiolum' colonisation at Sooke could have been detected. An added complication is the fact that no experimental data are available for endophyte growth rates. Colonisation of'Chlorochytrium', and especially 'Codiolum', at Bamfield occurred considerably later than at Burrard Inlet or Sooke. This may in part be explained by the fact that Acrosiphonia percent cover peaked later and plants persisted longer. Endophyte growth, fertility and duration in host. Growth of 'Chlorochytrium' and 'Codiolum' is believed to be rapid, judging by the range of cell sizes observed from one sampling date to the next. Note the steepness of the "maximum size" curve and "growth" portion of the 'Chlorochytrium' life history schematic curve (Fig. 2.8). Most 'Chlorochytrium' cells first observed in abundance at Sooke in early spring were 40 urn in diameter, but cells up to 120 urn were also already present. Less than two months later cells of 200 - 240 um were commonly found. Setchell & Gardner (1920), Scagel (1966), Chihara (1969) and Hudson (1974) described 'Chlorochytrium inclusum' cells from the northeast Pacific of size range 75 - 100 um. Many of the collections were, however, made early in the spring. 'Chlorochytrium inclusum', endophytic in Farlowia (Rhodophyta) from Japan, and associated with three Acrosiphonia species unknown to southwestern British Columbia, was also reported to be < 100 urn in diameter (Miyaji and Kurogi, 1976). Kjellman's (1883) description o f ' C . inclusum' within the arctic alga Sarcophyllis arctica Kjellman (Rhodophyta) is the only case I have found where 'Chlorochytrium' cells are noted to reach 275 um, in agreement with the size attained by 'Chlorochytrium' cells in Mazzaella splendens in this study. Chihara (1969) detected larger 'Chlorochytrium' cells closer to the base of the foliose host, whereas Kjellman (1883) noted individuals in the central part ofthe blade to be largest. Hudson (1974) observed the greatest concentration of cells toward the base. None of these observations were substantiated in my study, but might be explained by the much smaller sample 69 size of blades examined by the other researchers and the fact they collected blades primarily in spring and summer. In late August to January a large number of M. splendens blades from Sooke and Bamfield were found with large 'Chlorochytrium' cells (160 - 300 um) concentrated at the base of the blade. A possible explanation is that 1) the oldest (largest) cells are positioned closer to the base (the oldest part) of the blade and 2) newly colonised cells are sparsely scattered over the entire blade (the disappearance of Acrosiphonia implies gradual cessation of zygote settlement), and are growing slower than in the summer due to decreased light in the fall and winter months. 'Chlorochytrium' fertility and zoospore release occur in winter / early spring (Fig. 2.8) coinciding with Mazzaella splendens reproductive phenology (90 - 100 percent M. splendens blades collected from all three sites from October to February were reproductive). Kjellman (1883) found 'Chlorochytrium' fertile within Sarcophyllis arctica in winter, but did not observe zoospore release (unlike this study, however, he noted that 'Chlorochytrium' was most abundant in winter). Kornmann (1964) also collected fertile 'Chlorochytrium' (from Polyides, a foliose red seaweed, in Helgoland, Germany) in the winter, just months before Spongomorpha lanosa establishment. 'Chlorochytrium' of 80-100 um in diameter, obtained from Schizymenia (a bladed red seaweed) in Washington State, became fertile in culture and gave rise to 32, 64 or more zoospores 10 um long (Chihara, 1969). Unlike Chihara's fertile 'Chlorochytrium', the few fertile cells detected in material collected from Sooke were much larger, 160-200 um , and more than 800 zoospores, 5 um long, were released from one mature sporophyte. The absence of large cells in spring, after zoospore release, leads me to believe 'Chlorochytrium' cells spend less than one year in their host. However, it is unknown whether cells colonising M. splendens late in the summer remain small during the winter, e.g. 120 um, to become reproductive the following winter, as was proposed by Kornmann (1964) for 'Codiolum' in Helgoland, Germany. Chapter 3 will investigate this possibility. Juvenile cells of 'Codiolum', first detected in 'Petrocelis' in spring were 40 - 50 um x 10 - 20 um, apparently without stalk, and densely entangled among 'Petrocelis' filaments. It was difficult to detect stalks of newly colonised 'Codiolum' cells, but this does not mean the cells lacked them. Culture studies (Fan 1959, Jonsson 1970, Kornmann 1972, Miyaji 1984) demonstrated that germlings of 'Codiolum', only 10-15 um in length, were capable of becoming differentiated into vesicle and stalk within 5 - 1 5 days. A number of researchers (Setchell & Gardner 1920, Jonsson 1958, Kornmann 1972) have reported polymorphism among 70 'Codiolum' cells. This polymorphism was evident in specimens collected from all three study sites throughout the year, except in spring when juvenile cells were more uniform and distinct from older cells. The secondary stalk observed on some specimens (Fig. 2.10) was also described by other investigators (Printz 1926, Kornmann 1961, Hanic 1965) and referred to as a lateral stipe appendage. Fertile 'Codiolum' cells observed in the present study were generally > 80 nm in vesicle length and tended to have a proportionately large diameter (Figs. 2.9, 2.10). They were detected in 'Petrocelis' samples collected from fall through early spring, coinciding with the reproductive phenology of their host (tetrasporangia were present in 'Petrocelis' from October / November to February / March). Hanic (1965) found fertile 'C . petrocelidis' in tetrasporic 'Petrocelis franciscana' collected from Sooke in December 1963, and Kornmann (1961, 1964) reported fertile 'Codiolum' associated with Spongomorpha lanosa in tetrasporophytic Haemescharia hennedyi, a crustose red alga, from Helgoland in December 1963 and February 1964. As with 'Chlorochytrium', zoospore release of'Codiolum' generally occurs in the winter (Fig. 2.14), and the detection of only small juvenile cells the following spring suggests 'Codiolum' cells spend less than one year in their host. This is in contrast with Kornmann's (1961, 1964) finding: a mixture of 'Codiolum' cell sizes were found together in Haemescharia hennedyi at the time of 'Codiolum' colonisation, indicating some cells may spend about 18 months in their host, not releasing zoospores until the second winter. It seems unlikely to me that the very small cells observed in my study in greatest abundance in the spring and summer were comprised of both newly colonised cells and cells which had colonised the crust the previous spring or summer. Duration of'Codiolum' in 'Petrocelis' will be examined further in the next chapter. A simplified sporophyte life history schematic, illustrated in Figure 2.16, integrates the dynamics of'Chlorochytrium' and 'Codiolum' in Acrosiphonia's life history. Large, fertile cells are present in fall and winter; following zoospore release endophyte colonisation and growth take place throughout the spring and summer. 'Chlorochytrium' colonisation occurs before 'Codiolum' colonisation. 71 Lg Cells Sm cells Fall /Winter Spring / Summer Figure 2.16 Generalised life history schematic of Acrosiphonia''s sporophyte. * 'Chlorochytrium' colonisation occurs earlier than 'Codiolum' colonisation. Many 'smallest' cells of both 'Chlorochytrium' and 'Codiolum' were observed in January 1998 at the study sites where relatively high densities of endophytes were present. Since Acrosiphonia plants are long gone by this time, what is the origin of these putative juvenile 'Chlorochytrium' and 'Codiolum'? Perhaps they represent zygotes / sporophytes which colonised their host late in the season, and grow slowly during the fall and winter. They may grow quickly in the spring and still become reproductive for spring zoospore release, or remain vegetative until the winter or following spring, in which case Kornmann's hypothesis would be supported. Yet, this scenario does not explain the fact that a large number of such small cells was not detected in the fall of 1997. Furthermore, in the spring, prior to Acrosiphonia establishment, some juvenile Mazzaella splendens blades and 'Petrocelis' patches of new growth were found to be colonised by endophytes. One possible explanation, although never observed in culture, is that 'Chlorochytrium' and 'Codiolum' cells reproduce themselves, i.e. zoospores are released within the host or zoospores of free-living 'Chlorochytrium' and 'Codiolum' colonise M. splendens or 'Petrocelis'. An alternative explanation is that these small cells, although morphologically resembling 'Chlorochytrium' and 'Codiolum', are not in fact the sporophytes of Acrosiphonia. However, this seems rather unlikely, since PCR and D N A sequencing did not 72 detect any bp ambiguities which would indicate the presence of cells not associated with Acrosiphonia (see Chapter 1). Endophyte abundance variability. Higher abundance of Acrosiphonia in 1998 than in 1997 at Burrard Inlet and Bamfield is correlated with significantly higher endophyte densities in 1998 for both study sites (Figs. 2.4, 2.6, 2.11, 2.13). The significantly lower abundance of 'Chlorochytrium' and 'Codiolum' at Bamfield in 1997 (Figs. 2.6, 2.13), however, can not be explained in terms of lower 1997 Acrosiphonia abundance. Bamfield's much lower endophyte colonisation than Sooke's, over the entire sampling period, may be due to cells having more difficulty in settling on and penetrating their hosts at wave-exposed sites. At Burrard Inlet the discrepancy between low numbers of 'Chlorochytrium' in 1997 and much higher densities in 1998 [(Fig. 2.4B) more than would be expected simply from a doubling of Acrosiphonia percent cover] may result from zygotes from Acrosiphonia sp. (Alaska) exhibiting higher colonisation success than A arcta. The four 1998 'Chlorochytrium' D N A sequences from Burrard Inlet isolates were 100 percent identical to the Alaskan Acrosiphonia isolates (Chapter 1). Endophyte abundance in 1997 at Bamfield showed no correlation with 1998 Acrosiphonia abundance, i.e. low endophyte densities did not result in low Acrosiphonia percent cover. Endophyte survivorship (discussed in Chapter 3), however, may have been high. Furthermore, some Acrosiphonia populations may not be solely dependent on endophyte abundance and survivorship, i.e. vegetative propagation or dispersal as discussed earlier may affect the distribution and abundance of Acrosiphonia. Endophyte densities, particularly 'Chlorochytrium' densities, unexpectedly decreased after onset of colonisation (Figs. 2.5B, 2.6B, 2.1 IB, 2.12B). Two possible explanations are 1) Mazzaella splendens blades increase in size as the summer progresses causing endophyte densities to decrease correspondingly, and 2) excessive crowding of'Chlorochytrium' cells results in the mortality of some cells. The first explanation can be ruled out, because seasonal abundance patterns generated from absolute numbers of 'Chlorochytrium' and from densities of 'Chlorochytrium' were similar, i.e. 'Chlorochytrium' abundance actually decreased in summer. I therefore propose that the decrease in 'Chlorochytrium' cells may be due to mortality of cells caused by crowding of neighbouring cells. During the summer, especially at Sooke, 'Chlorochytrium' cells were often seen touching one another (Fig. 2.7A). High densities of 73 'Codiolum' were also observed (Fig. 2.9B), but more space seems to be available among the filaments of'Petrocelis' than among the cells of the cortex region of M. splendens. Greater knowledge of M. splendens growth rate, and isolation of other factors which may affect endophyte survivorship are fundamental to a better understanding of'Chlorochytrium' and 'Codiolum' survival. Selection of heteromorphy in Acrosiphonia. Speculation on selective mechanisms for maintenance of heteromorphy in haplodiploid life cycles (Stebbins & Hi l l 1980, Valero et al. 1992, Klinger 1993) has resulted in two hypotheses. The first proposes that the two morphologically distinct phases differ with respect to their tolerance of environmental conditions (Conway et al. 1976). The second suggests that the phases represent mutually exclusive adaptations to fluctuations in grazing pressure, i.e. the upright phase is adapted to counter increased competition (growth and reproduction) when grazing pressure is low, and the non-upright phase is adapted to survive high grazing pressure by being grazer resistant or exploiting micro-habitats which offer protection from herbivory (Lubchenko & Cubit 1980, Slocum 1980, Littler & Littler 1980). This study, carried out in the field at three environmentally different sites, has clearly shown that the morphologically different gametophytic and sporophytic phases of Acrosiphonia occur in different environmental conditions. The filamentous gametophyte is seasonally abundant [March - June / July (October)]. Based on culture results, Acrosiphonia's seasonality may be dependent on abiotic factors such as photoperiod, light intensity, desiccation and temperature. The endophytes, 'Chlorochytrium' and 'Codiolum' show high tolerance to environmental extremes. They survive high summer temperatures (which appear to ultimately kill the filamentous Acrosiphonia plants), and spend the winter (possibly too adverse for Acrosiphonia, which seems to require high light intensity for gamete production and long-day conditions for optimal growth) within Mazzaella splendens and 'Petrocelis franciscana'. The host may offer protection from the physical environment, as well as from herbivores and competition from other algal species. With regard to Acrosiphonia's life history representing an adaptation to variable grazing pressure, i.e. during high herbivore pressure the filamentous gametophyte has disappeared, while the endophytes are protected within their hosts, no empirical data are available on differential susceptibility of Acrosiphonia, M. splendens and 'Petrocelis' to grazing. Laboratory feeding 74 experiments and / or herbivore exclusion experiments may help to determine whether 1) herbivory is a factor in the seasonal distribution and abundance of Acrosiphonia and 2) grazing pressure is an evolutionary force in the maintenance of Acrosiphonia's heteromorphic life history. This study also demonstrates Acrosiphonia's success (and adaptability) in different environments. Present at three study sites, characterised by differences such as wave exposure, salinity and timing of summer low tides, Acrosiphonia plants showed variable seasonality (e.g. persisted over a longer period in more wave-exposed environments) and abundance (dependent on factors such as local weather, micro-habitat factors and survivorship of endophytes). In addition, variable abundance of the two morphological species, A. arcta and A. coalita, at different tidal heights, and A. coalita's absence at Burrard Inlet, suggests adaptations at the species level, and Acrosiphonia's appearance in new areas from year to year implies its ability to successfully colonise new habitats / sites. 'Chlorochytrium' D N A sequences which shared a genotype with Acrosiphonia sp. (Alaska), and the unique Acrosiphonia sequence (specimen collected from the large boulder at Sooke, see Chapter 1), hint at colonisation of Acrosiphonia "species" other than A. arcta and A. coalita, e.g. A. sp. (Alaska) and /or A. sp. (Sk5), and / or adaptability oi Acrosiphonia to new habitats. The role the endophytic sporophytes play in Acrosiphonia's unique life history is still poorly understood, and many fundamental questions remain unanswered. For example, it is unknown whether the sporophyte phase is actually required for the maintenance of Acrosiphonia populations in southwestern British Columbia. Perhaps Acrosiphonia can successfully recycle through rhizoid overwintering of gametophytic plants or by gametophytic plants giving rise to haploid zoospores (only shown in culture) which survive the winter. If so, the relative contribution of sexual reproduction (and the role of sporophytes) may be less significant compared to asexual reproduction and / or vegetative propagation in Acrosiphonia's life history. In some algae, e.g. the brown alga, Analipus japonicus (Harvey) Wynne, and the coralline alga, Lithothrix aspergillum Gray, loss of sexual reproduction may actually occur at the northern range of the species (Nelson 1980, DeWreede & Vandermeulen 1988). Also of interest is the question of whether the unicellular sporophytes can survive epilithically or planktonically. Given the evolution and maintenance of this alternation of heteromorphic generations, what is the adaptive significance of sporophytes evolving an endophytic condition? To experimentally test in nature whether the sporophyte phase (endophytic or free-living) is required for the maintenance of 75 Acrosiphonia populations, patches in the intertidal zone could be cleared / sterilised in the fall. The appearance of juvenile Acrosiphonia plants in the spring would suggest they are derived from zoospores of the unicellular sporophyte (provided all putative rhizoids were removed and disregarding the possibility of asexual reproduction in the field). Of course, should no juvenile Acrosiphonia appear, this would not mean the sporophyte phase is insignificant in Acrosiphonia's life history, since patchy distribution of Acrosiphonia in the present study suggests patchy zoospore dispersal. The presence of 'Chlorochytrium' and 'Codiolum' at all three study sites, although variable in abundance from year to year, suggests to me the endophytes are successful at colonisation. At Bamfield much lower endophyte densities imply lower success - it may be more difficult to colonise blades or crusts in a more wave-exposed environment - such that asexual or vegetative means of reproduction may contribute (more) to maintenance of Acrosiphonia. Consequently, and considering the lack of evidence to support alternate life histories for A arcta and A. coalita in the field [e.g. Kornmann (1962) has shown sexual reproduction in cultured A. sonderi to be absent, and the 'Codiolum' phase of A. grandis in culture to develop directly into the filamentous gametophyte (1970a)], success of this heteromorphic life history seems apparent. Heteromorphy can be seen as a bet-hedging strategy, whereby different phases are adapted to a seasonally variable environment or variable grazing pressure. With regard to Acrosiphonia, not only have two phases evolved, but molecular data (Chapter 1) have now shown that the sporophytic phase of at least A. arcta can colonise two alternate hosts (crustose and foliose red algae). The availability of these hosts is another important factor in Acrosiphonia's life history, and will be explored in the following chapter. 76 REFERENCES Agardh, J. G. 1846. Anadema, ett nytt slagtet bland Algerne. K. Sv. Vet.-Akad. Handl. pp. 1-16. Stockholm. Amsler, C. D. & Searles, R. B. 1980. Vertical distribution of seaweed spores in a water column offshore of North Carolina. J. Phycol. 16:617-9. Anderson, E. K. & North, W. J. 1966. In situ studies of spore production and dispersal in the giant kelp, Macrocystis. In Young, E. G. , McLachlan, J. L. (Eds) Proc. 5th Int. Seaweed Symp. Pergamon Press, Oxford pp. 73-86. Chihara, M . 1969. Culture study of Chlorochytrium inclusum from the northeast Pacific. Phycologia 8:127-33. Collins, F . S . 1909. The green algae of North America. Tufts College Studies (Sci. Ser.) 2:79-480. Conway, E., Mumford, T. & Scagel, R. F. 1976. The genus Porphyra in British Columbia and Washington. Syesis 8:185-244. Day, R. W. & Quinn, G. P. 1989. Comparisons of treatments after an analysis of variance in ecology. Ecol. Monogr. 59:433-63. DeCew, T. C , Silva, P. G. & West, J. A . 1992. Culture studies on the relationship between Schizymenia and Haematocelis (Gigartinales, Rhodophyceae) from the Pacific Coast of North America. J. Phycol. 28:558-66. Dethier, M . N . 1987. The distribution and reproductive phenology of intertidal fleshy crustose algae in Washington. Can. J. Bot. 65:1838-50. DeWreede, R. E. and Vandermeulen, H. 1988. Lithothrix aspergillum (Rhodophyta): regrowth and interaction with Sargassum muticum (Phaeophyta) and Neorhodomela larix (Rhodophyta). Phycologia 27:469-76. Environment Canada, Climate Data Services 1998. Fan, K. C. 1959. Studies on the life histories of marine algae I. Codiolum petrocelidis and Spongomorpha coalita. Bull. Torrey Bot. Club 86:1-12. Games, P. A. , Keselman, J. H . & Rogan, J. C. 1983. A review of simultaneous pairwise multiple comparisons. Statistica Neerlandica 37:53-8. Garbary, D. J., Kim, K. Y . , Klinger, T. & Duggins, D. 1999a. Red algae as hosts for endophytic kelp gametophytes. Mar. Biol. 135:35-40. 77 Garbary, D. J., Kim, K . Y . , Klinger, T. & Duggins, D. 1999b. Preliminary observations on the development of kelp gametophytes endophytic in red algae. Hydrobiologia 398/399:247-52. Hanic, L. A . 1965. Life history studies on Urospora and Codiolum from southern B.C. Ph.D. Thesis, University of British Columbia, 152 pp. Hollenberg, G. J. 1958. Observations concerning the life cycle of Spongomorpha coalita (Ruprecht) Collins. Madrono 14:249-52. Hudson, M . L. 1974. Field, culture and ultrastructural studies on the marine green alga Acrosiphonia in the Puget Sound region. Ph.D. Thesis, University of Washington, 206 pp. Jonsson, S. 1958. Sur la structure cellulaire et la reproduction de Codiolum petrocelidis Kuck., Algue verte unicellulaire endophyte. C. R. Acad. Sci., Paris 247:325-8. 1959a. L'existence de l'alternance heteromorphe de generations entre YAcrosiphonia spinescens et le Codiolum petrocelidis. Ibid. 248:835-7. 1959b. Le cycle de developpement du Spongomorpha lanosa (Roth) Kiitz. et la nouvelle famille des Acrosiphoniacees. Ibid. 248:1565-7. 1962. Recherches sur des Cladophoracees marines: structure, reproduction, cycles compares, consequences systematiques. Ann. Sci. Nat. Bot. 3:25-230. 1963. Sur quelques variations du cycle de developpement dans la famille des Acrosiphoniacees. C. R. Hebd. Seance. Acad. Sci., Paris 256:5187-9. 1964. Existence d'une caryogamie facultative chez YAcrosiphonia spinescens (Kiitz) Kjell. Ibid. 258:6207-9. 1966. Sur 1'identification du sporophyte du Spongomorpha lanosa. Ibid. 262:626-9. 1970. Localisation de la meiose dans le cycle de XAcrosiphonia spinescens (Kiitz.) Kjell. (Acrosiphoniacees). Ibid. 271:1859-61. 1971. Des lignees apocaryogamiques et leur segregation dans la population de VAcrosiphonia arcta, Chlorophycee marine d'Heligoland. Helg. Wiss. Meeresunters. 22:281-94. 1986. Isolement, en culture, de lignees haploides a partir du cycle heteromorphe chez le Spongomorpha aeruginosa. Cryptogamie, Algologie 7:149-59. 1991. Typification of Spongomorpha (Kiitz) Wille and Acrosiphonia (Agardh) Wille (Acrosiphoniales, Chlorophyta). Ibid. 12:171-81. 78 Kendrick, G. A. & Walker, D. I. 1991 Dispersal distances for propagules of Sargassum spinuligerum (Sargassaceae, Phaeophyta) measured directly by vital staining and venturi suction sampling. Mar. Ecol. Prog. Ser. 79:133-8. Kendrick, G. A. & Walker, D. I. 1995. Dispersal of propagules of Sargassum spp., Sargassaceae, (Phaeophyta): observations of local patterns of dispersal and possible consequences for recruitment and population structure. J. Exp. Mar. Biol. Ecol. 192:273-88. Kjellman, F. R. 1883. The algae of the Arctic sea. K. Svensk. - Akad. Handl. 20:320-1. 1893. Studien dfver chlorophyces lagtet Acrosiphonia J. G. Agardh och dess skandinaviska arter. Bihang till K. Sv. Vet - Akad. Handl. vol. 18 no. 5 Stockholm. Klinger, T. 1993. The persistence of haplodiploidy in algae. Trends Ecol. Evol. 8:256-8. Kornmann, P. 1961. Uber Spongomorpha lanosa und ihre Sporophytenformen. Helg. Wiss. Meeresunters. 7:195-205. 1962. Eine Revision der Gattung Acrosiphonia. Ibid. 8:219-42. 1964. Zur Biologie von Spongomorpha aeruginosa (Linnaeus) van den Hoek. Ibid. 11:200-8. 1965. Zur Wachstum und Aufbaus won Acrosiphonia. Ibid. 12:219-38. 1970a. Der Lebenszyklus von Acrosiphonia grandis (Acrosiphoniales, Chlorophyta). Mar. Biol. 7:324-31. 1970b. Phylogenetische Bezeihungen in der Griinalgengattung Acrosiphonia. Ibid. 21:292-304. 1972. Les sporophytes vivant en endophyte de quelques Acrosiphoniacees et leurs rapports biologiques et taxonomiques. Soc. Bot. Fr., Memoires 75-86. Kuckuck, P. 1894. Bemerkungen zur marinen Algenvegetationen von Helgoland. Wiss Meeresunters. Abt. Helgoland, N . F. 1:223-63. Littler, M . M . & Littler, D. S. 1980. The evolution of thallus form and survival strategies in benthic marine macroalgae: field and laboratory tests of a functional form model. Am. Nat. 116:25-44. Lubchenco, J. & Cubit, J. 1980. Heteromorphic life histories of certain marine algae as adaptations to variations in herbivory. Ecology 61:676-87. 79 Miyaji, K . 1984. The life history study on a sporophytic form of Spongomorpha-Acrosiphonia_comp\ex (Acrosiphoniales, Chlorophyta), Codiolum petrocelidis Kuckuck from Northern Japan. Jpn. J. Phycol. 32:29-36. Miyaji, K . 1996. Taxonomic study of the genus Spongomorpha (Acrosiphoniales, Chlorophyta) in Japan. Phycol. Res. 44:27-36. Miyaji, K . and Kurogi, M . 1976. On the development of zoospores of the green alga Chlorochytrium inclusum from eastern Hokkaido. Bull. Jap. Soc. Phycol. 24:121-9. Nelson, W. A . 1980. Analipus japonicus (Harv.) Wynne (Phaeophyta): Studies of its biology and taxonomy. Ph. D. Thesis, The University of British Columbia, 244 pp. Printz, H . 1926. Chlorophyceae. In Engler, A . and Prantl, K . (Eds.). Die natiirlichen Pflanzenfamilien, 2nd Ed. Vol.3 Leipzig, 464 pp. Reed, D. C , Laur, D. R. & Ebeling, A . W. 1988. Variation in algal dispersal and recruitment: the importance of episodic events. Ecol. Monogr. 58:321-35. Scagel, R. F. 1966. Marine algae of British Columbia & northern Washington, Part I: Chlorophyceae. Nat. Mus. Can. Bull. No. 207, Biol. Ser. No. 74, 257 pp. Setchell, W. A . & Gardner, N . L. 1920. The marine algae of the Pacific Coast of North America. Part II. Chlorophyceae. Univ. Calif. Publ. Bot. pp. 149-50. Slocum, C. J. 1980. Differential susceptibility to grazers in two phases of an intertidal alga: advantages of heteromorphic generations. J. Exp. Mar. Biol. Ecol. 46:99-110. SPSS® version 9.0 for Windows. 1999. SPSS Inc. Chicago, IL. Stebbins, G. L. & Hi l l , J. C. 1980. Did multicellular plants invade the land? Am. Nat. 115:342-53. Valero, M . , Richerd, S., Perrot, V . & Destombe, C. 1992. Evolution of alternation of haploid and diploid phases in life cycles. Trends Ecol. Evol. 7: 25-29. Wille, N . 1900. Die Zellkerne bei Acrosiphonia (J. Ag.) Kjellm. Botanisches Centralblatt 8 1238-9. Zar, J. H . 1996. Biostatistical Analysis, 3rd Ed. Prentice Hall, New Jersey, 662 pp. 80 C H A P T E R 3 HOST A V A I L A B I L I T Y F O R C O L O N I S A T I O N A N D S U R V I V A L O F C H L O R O C H Y T R I U M ' A N D C O D I O L U M ' I N T R O D U C T I O N Two red algal hosts commonly provide a habitat for the unicellular sporophytes 'Chlorochytrium' and 'Codiolum' in British Columbia. One is the foliose alga, Mazzaella splendens, conspicuous on rocky shores in the low intertidal ofthe northeast Pacific; the other the crustose alga, 'Petrocelis franciscana', predominant throughout the intertidal ofthe southern British Columbia coast. In general, the host organism provides the specific ecological niche required by the endophyte. Benefits to the endophyte may include providing colonisable substrate, ensuring its exposure to light and nutrients and escape from harsh environmental conditions and potential herbivores. Mazzaella splendens and 'P. franciscana' appear to be ideal hosts: they are present in high abundance presumably year round [pers. observation and Paine et al.'s (1979) study on the longevity of 'Petrocelis'], providing relatively high surface area for maximum colonisation, exposure to light and nutrients and survival. But are they in fact ideal hosts? What is known about variation in host abundance from year to year or month to month? And can variable host abundance play a role in endophyte survival? Since many algae exhibit marked seasonal changes in abundance, it is important to establish whether host seasonal patterns exist. Some work has documented seasonal changes in Mazzaella splendens density (Dyck and DeWreede 1995, F. Shaughnessy pers. comm.) along the coast of southwestern British Columbia, but little is known about the ecology of 'Petrocelis'. Paine et al.'s (1979) extensive study examined growth patterns and longevity of 'Petrocelis' found on a sandstone wall in Washington, but no studies have closely examined 'Petrocelis' seasonal abundance on boulder strewn shores. M y study quantifies monthly abundance of M. splendens (over two years) and of 'Petrocelis' (over 12 and 18 month periods at different sites) in order to elucidate any seasonal patterns, and considers possible factors in producing such patterns. Finally, 'Chlorochytrium' and 'Codiolum' survival are assessed in relation to availability of M. splendens and 'Petrocelis', and I speculate on the length of time the endophytes remain in their hosts and why this strategy may have evolved. 81 M A T E R I A L S A N D M E T H O D S Host availability for 'Chlorochytrium' and 'Codiolum' was examined at Burrard Inlet, Sooke and Bamfield. As mentioned in Chapter 2, Bamfield was sampled less intensively. Two investigations using seawater tanks were carried out to examine 'Chlorochytrium' survival in dislodged Mazzaella splendens blades. 'Petrocelis' surface area. 'Petrocelis' crusts were sampled at Sooke from May 1997 to November 1998 and at Burrard Inlet from August 1997 to August 1998. Sampling occurred approximately monthly within the low to high intertidal (between 0.2 to 5.1 m above the Canadian Chart Datum). Thirty granite boulders were haphazardly chosen (10 each from the low, mid and high intertidal) from the total number occupied by 'Petrocelis' crust. Each boulder was marked with West System® epoxy (Gougeon Brothers Inc., Bay City, MI) after an area had been cleared and dried with a propane flame torch. Within each epoxy patch (variable in size but on average 5 cm in diameter) a numbered Dymo® LD.™ ribbon was embedded. In this way it was possible to identify each of the 30 'Petrocelis' crusts. Weathering by water motion, however, resulted in ribbon numbers not being legible over time, so that crusts were then only identifiable by the unique shape of the epoxy marker. Repeated sampling of known crusts was chosen rather than random sampling as with M. splendens blades, because I wished to determine the persistence and performance of specific crusts, i.e. are changes in patch size consistent for all crusts? To facilitate locating boulders, especially when low tides occurred after dark, VA inch rebar marked with flagging tape was pounded into the ground at each corner of the sampling area at Sooke. The total area measured approximately 20m x 20m. At Burrard Inlet the area sampled was bounded by the seawall and measured about 10m x 20m. Sampling at Burrard Inlet began later than at Sooke, because dense overgrowth of boulders by juvenile barnacles, mussels and colonial diatoms made it difficult to locate 30 boulders with 'Petrocelis' patches from May to August 1997. On each sampling day all 'Petrocelis' crusts were photographed with 100 A S A Kodak film (Eastman Kodak Co., Rochester, NY) ; flash photography was used at night. A five cm ruler was placed next to the crust for scale. Care was taken to photograph perpendicular to the crust and to select crusts with as little parallax as possible. It was not always possible to locate and photograph all 30 crusts, since boulders became overgrown with other algae and sessile invertebrates (especially Ulva and Fucus and juvenile barnacles and mussels in spring and 82 summer) making it difficult to locate the epoxy markers. In addition, with time epoxy markers deteriorated to the point where by November 1998 (after 20 months of monitoring) only 23 markers could still be found at Sooke. This was particularly true at Burrard Inlet, where overgrowth by juvenile barnacles contributed to the loss of epoxy markers: only one year after boulders had been marked, ten epoxy markers could no longer be located. The photographic series also permitted examination of limpet abundance, barnacle and mussel settlement and the establishment of new crusts. Twelve Tectura scutum limpets were collected from Sooke in February 1997 for gut analysis as possible evidence for herbivory. 'Petrocelis' crusts could not be adequately monitored at the Bamfield site, because rocky land masses rather than boulders comprise the shore, and the 'Petrocelis' usually overlaps with both Ralfsia and Hildenbrandia, making it difficult to mark and identify individual 'Petrocelis' patches. Instead, the condition ofthe crusts was monitored, noting evidence of new growth (thickness and colour) and the presence of green colouration and unhealthy-looking tissue. The presence or absence of limpets on 'Petrocelis' was also recorded. Mazzaella splendens density. Sampling of Mazzaella splendens occurred over a two year period from October / November 1996 to October / November 1998, and was monthly except during those periods where tides were not low enough to expose Mazzaella or blades were very scarce. Both the condition, i.e. health and degree of epiphytisation, and density of M. splendens was monitored. At each site a 12 m transect line was randomly placed, parallel to the sea, in an area where M. splendens was present. The same area, between 0.2 and 1.0 m above zero tidal level (Canadian Chart Datum), was sampled each time. A random number table was used to choose eight points along the length of the transect. These points were permanently marked with flagging tape, and served as points for the point-quarter distance method of density estimation (Krebs, 1999). 'Chlorochytrium' survival in drift Mazzaella splendens blades. Drift Mazzaella splendens blades collected at Bamfield were found to be colonised by 'Chlorochytrium', prompting the question of 'Chlorochytrium' survival in dislodged blades. The following two investigations were conducted (1) Twelve M. splendens blades colonised by 'Chlorochytrium' (average size 120-160 um) were collected from the low intertidal at Sooke in November 1998. The blades were left in a laboratory seawater tank with constant flow at 11°C under 12 umol m'2 83 s"1 of cool white light and examined every 7-10 days for two months to determine 'Chlorochytrium' condition, i.e. average size, health and whether fertile. Observations were not made beyond two months due to disintegration of algal blades. (2) Mazzaella splendens blades were collected from the low intertidal at Bamfield in December 1998. Those blades colonised by 'Chlorochytrium' were placed in an outdoor tank (approximately 1.5 m x 1.5 m; steady flow, water temperature from 6-8°C and natural light). Plexiglas plates (approximately 20 cm x 20 cm) were positioned on the bottom of the tank to establish whether zoospores would be released and settle on the plates. A neighbouring tank lacking M. splendens blades served as a control tank. The tanks were checked one and two months later. Data analysis. A l l 'Petrocelis' crust photographs were scanned and enhanced (painted black or white to contrast with the background) in Adobe Photoshop® 5.02 for Windows (1998). The imaging program SigmaScan Pro® 5.0 for Windows (1999) was then used to calculate the surface areas of crusts for each sampling date. Images were calibrated with the 5 cm ruler present in each image. For each time period an average value was calculated for 'Petrocelis' crust surface area. Seasonal differences in mean crust surface area were determined by locating maximum (highest surface area) and minimum (lowest surface area) points from the graphed data. These points corresponded to summer / fall (maxima) and winter (minima). Significance of differences between pairs of maximum / minimum points was tested using an independent samples / test. Transformations of some of the data did not reduce heterogeneity of variances to nonsignificant levels [Levene Test, SPSS® 9.0 for Windows (1999)], but the Kolmogorov-Smirnov Test showed normality. Furthermore, the t test is generally robust to variance heterogeneity (Zar, 1996), and the independent samples t test in SPSS® 9.0 for Windows (1999) includes an option for unequal variances which I used. Seasonality of 'Petrocelis' abundance and 'Codiolum' density were compared. Mean densities of Mazzaella splendens were plotted against time, and significance of seasonal differences determined by checking for overlap of 95% confidence intervals. M. splendens seasonality was compared to seasonal densities of'Chlorochytrium' to assess availability as a host. 84 RESULTS 'Petrocelis' seasonal abundance. Marked seasonal fluctuations in 'Petrocelis' abundance were evident both at Sooke and Burrard Inlet. 'Petrocelis' abundance peaked in summer / early fall, and was at its lowest in the winter at both sites. This seasonal change was visible both from the photographic time series and from plotting mean surface areas against time (zero values for surface area were omitted from calculations). From October 1997 to January / February 1998 mean crust surface area decreased by 43 % at Burrard Inlet (Fig. 3.1 A) and by 58 % at Sooke (Fig. 3.2A); during the 6 month period from January / February 1998 to July / August 1998 crust surface area increased by 45 % at Burrard Inlet and 61 % at Sooke. These represent significant differences (all p's < .05, Table 3.1). Burrard Inlet crusts. 'Petrocelis' crusts at Burrard Inlet at the start of observation ranged from 1.95 to 20.49 cm 2 with a mean surface area of 6.29 (1 S.E. = 1.15) cm 2. Crust distribution was much patchier than at Sooke, and in general 'Petrocelis' abundance was notably lower. Within one year 39% of the crusts for which epoxy markers could still be found had disappeared. The mean net change in surface area in one year was - 20%, reflecting this high mortality (Fig. 3.3 A). This may in part be attributed to dense settlement of juvenile barnacles on crusts (9/18 crusts in August 1998 were partially or totally covered by barnacles) as shown in Figure 3.4A. It is not clear i f crust tissues can persist beneath barnacle settlement, because barnacles were not removed from crusts. Among the 61% of crusts which survived during the observation period, mean net change in surface area in one year was + 30%. Change in surface area of individual crusts could not be examined, due to the theft of field notes which identified the individuals. Sooke crusts. Crust patches at Sooke were larger, from 2.58 to 46.4 cm with a mean surface area of 14.31 (1 S.E. = 1.75) cm . At Bamfield 'Petrocelis' crust size was difficult to interpret due to overlapping with other crusts. Nonetheless, on average, patches were much larger than at Sooke, reaching sizes greater than 100 cm 2. After 18 months, of those crusts still identifiable at Sooke, 55% showed a net increase, 17% a net decrease and 28% had disappeared. This translates to a mean net change in surface area of - 24%, again reflecting the mortality of individual crust patches (Fig 3.3B) as was the case for Burrard Inlet. However, among the crust individuals which did not go extinct, the mean net change in surface area is +21%. Although a considerable amount of variation exists among the individual crust patches (variances are high), ° - 'Codiolum' 120000 -1 spring'97 fall '97 spring'98 fall '98 Sampling Day from Start Figure 3.1 A . Surface area of 'Petrocelis' crust at Burrard Inlet over 1 year. Data are means ± S.E. from 26 crusts. Surface area values of zero were omitted from the calculation of the mean. B . ' Codiolum' densities at Burrard Inlet. Data are means ± S.E. from cells counted in 30 'Petrocelis' patches. Sooke 'Codiolum' 30000 i spring '97 fall '97 spring '98 fall '98 Sampling Day from Start Figure 3.2 A . Surface area of 'Petrocelis' crust at Sooke over 18 months. Data are means ± S.E. from 30 crusts. Surface area values of zero were omitted from the calculation of the mean. B. ' Codiolum' densities at Sooke. Data are means ± S.E. from cells counted in 30 'Petrocelis' patches. 87 Table 3.1 Seasonal differences of 'Petrocelis' crust surface area. Results of independent samples f-test performed on pairs of maximum and minimum data points shown in Figures 3.1 A and 3.2A. The significant t value indicates a statistically significant difference among maximum and minimum surface area values. Note that pairs of points correspond to (1) (summer) fall and winter and (2) winter and summer. Burrard Inlet n Mean S.D. S.E. t p Oct. 1997 26 8.5240 8.9499 1.7552 2.555 0.016 Feb. 1998 17 3.7070 2.8374 1.8853 Feb. 1998 17 3.7070 2.8374 0.6882 3.873 0.001 Aug. 1998 11 8.1798 3.2049 0.9663 Sooke N Mean S.D. S.E. t p Oct. 1997 28 18.1921 11.7437 2.2193 2.619 0.012 Jan. 1998 26 10.9112 8.5354 1.6739 Jan. 1998 26 10.9112 8.5354 1.6739 2.351 0.024 July 1998 14 16.3614 6.0001 1.6036 on average 'Petrocelis' crust surface area decreases significantly in winter and increases significantly in summer (Fig. 3.2A). No definitive relationship between crust size and change in surface area was noted, except that the smallest individuals (2.5-3.5 cm2) tended to undergo little fluctuation in surface area, instead exhibiting slow and steady growth in size. Crust dynamics and invertebrate observations. On average crusts both at Sooke and at Burrard Inlet grew during the summer and early fall of 1997 as illustrated in the time series in Figure 3.5. New growth of 'Petrocelis' was also apparent at Bamfield in the summer. In the field older parts of the crust were observed to be lighter in colour and thinner. Limpets were rarely present in crust photographs, nor observed in any abundance in the field during the summer at all three sites. 'Petrocelis' A. Burrard Inlet 12 , E 10 u GO 6 4 2 I 0 • • - - • Series 1 - « — Series2 100 200 300 fall '97 400 500 600 700 spring '98 fall '98 B. a 14 3 OO Sooke Seriesl — • — Series2 100 200 300 400 500 600 700 800 fall '97 spring '98 fall '98 Figure 3.3 Surface area of 'Petrocelis' crust patches at Burrard Inlet (A) and Sooke (B). Series 1 illustrates change in surface area of crusts where values of zero were omitted. Series 2 includes zero surface area values, i.e. crusts which disappeared during the observation period, to illustrate crust survivorship. Data are means ± S.E. from 26 (A) and 30 (B) crusts. Figure 3.4 A. Barnacle settlement on 'Petrocelis' at Burrard Inlet. B. Intricate crust pattern at Sooke. C. D. Limpet abundance and notable tissue loss from Oct. to Nov. E . Fragmentation of single crust? F. G. H. Green, mucilaginous tissue. I. J. "Vein" pattern on Mazzaella splendens. 9 0 91 Following this growth period, in the winter, 80% of individuals at Sooke decreased in crust surface area or experienced patch fragmentation. This was also the pattern for the average of 30 crusts at Sooke and Burrard Inlet. In general, it took patches a minimum of 6 - 8 months to begin to disintegrate, but loss of crust tissue could easily be seen from one month to another. At Sooke primarily central tissue seemed to be lost, with perimeter tissue more persistent. This was not as evident at Burrard Inlet where crusts were smaller and barnacle overgrowth made it difficult to assess growth patterns. The limpets Tectura scutum and Tectura persona Rathke were found in abundance at Sooke and Burrard Inlet, and often observed on 'Petrocelis' crust, from October to January and during this time photographs often revealed loss of crust tissue from one month to the next (Fig. 3.4C,D). Densities of 160 limpets m"2 were estimated from Sooke crust photographs taken in October and November 1997. Crust patches with intricate patterns such as illustrated in Figure 3.4B were not uncommon at Sooke. Gut analyses of Tectura scutum limpets revealed a multitude of diatoms, filamentous red and green algal fragments, foraminiferans and possibly crust cellular components. No distinctive 'Petrocelis' filaments could be identified. Limpets, as well as the chiton, Katharina tunicata were commonly seen on the rocky shores at Bamfield during the winter. Some fragmented crust patches began to regenerate lost tissue (most had not fully regenerated by the end of the monitoring period) in the spring and summer. Other fragmented patches continued to lose tissue, decaying until eventually none of the original patch remained. New growth of crust tissues was also observed. On a few occasions at Sooke it was difficult to assess whether the original marked crust had actually been a remnant of an older larger, continuous patch, parts of which still existed on a different area of the boulder (Fig. 3.4E). At Burrard Inlet difficulty in determining crust mortality stemmed from dense overgrowth by barnacles obscuring crust patches. Of those crusts at Sooke which underwent another growth period in the spring and summer of 1998 (64%), a few produced intriguing time series patterns as a result of variable loss and regrowth of tissues. Often internal spaces (loss of central tissue of a crust) became filled in with new tissue and subsequent loss of tissues resulted in a negative-image effect [Fig. 3.5 (4,5)]. Health of crusts. Green-coloured, mucilaginous surface tissue (Fig. 3.4F,G,H) was noted for a number of crusts in July and August of 1998 at all three sites. Most individuals appeared healthy-looking again by fall 1998, but a few seemed to continue to decay. This was 92 less evident at Bamfield, because individuals could not be identified. The phenomenon had also been recorded (to an even greater extent with many crusts disintegrating) in the winter of 1996 prior to the photographic time series. Smallest individuals, however, were never observed as greenish or mucilaginous. At Bamfield olive-green crust existed intermingled with the darker red/brown 'Petrocelis'. Unlike the green mucilaginous crust this crust appeared healthy-looking, and upon examination, revealed the same tissue structure as 'Petrocelis franciscana'. It has been suggested (R. Scrosati, pers. comm.) that this "healthy" crust may be the tetrasporophytic phase of Mastocarpus jardinii West; in culture it strongly resembles 'Petrocelis franciscana' (West et al. 1978). Crust recruitment. During the sampling period only three crust recruits were detected, two of these at Burrard Inlet in July 1998, the other at Sooke in June 1998. One occupied the area on a boulder where a known crust had disappeared, and the other two were observed in the vicinity of a known crust which was still present. From the photograph of the Sooke crust in question, it seems the new crust may be a resurrected crust, i.e. it may have been present at an earlier time and have been part of a larger continuous crust, since many fragments were still present on the boulder (Fig. 3.4E). In the spring and summer of 1997 prior to sampling, several new crust patches were noted on the base of the seawall at Burrard Inlet. Figures 3.1 and 3.2 show that 'Petrocelis' seasonal abundance, i.e. highest and lowest mean surface area, corresponds to highest and lowest 'Codiolum' densities in these crusts at Sooke and at Burrard Inlet. Mazzaella splendens seasonal abundance and condition of blades. Seasonal changes in density of M. splendens were similar for Burrard Inlet, Sooke and Bamfield sites (Figs 3.6, 3.7, 3.8). The highest densities occurred in the summer months, corresponding to the highest 'Chlorochytrium' densities in these blades for Sooke, Burrard Inlet (summer 1998 only) and Bamfield (summer 1998 only). This pattern was less evident for Bamfield. During the winter months M. splendens densities were lowest (and significantly lower than summer densities based on 95% confidence intervals not overlapping). In fact at Sooke and Bamfield, densities of < 0.1/m2 and absence of blades in the intertidal were recorded; at Burrard Inlet blades appeared to persist year round, although markedly decreased in number. 'Chlorochytrium' densities were Burrard Inlet B. 'Chlorochytrium' 1 i ^ 0.8 -"I 0.6 -5 0.4 -I 0.2 -0 i • •»•••,••,• >, > * • * 0 100 200 300 400 500 600 700 800 spring '97 fall '97 spring '98 fall '98 Sampling Day from Start Figure 3.6 A . Mazzaella splendens density ± S.E at Burrard Inlet over two years. The point-quarter method of density estimation was based on locating 8 random points along a transect. B . 'Chlorochytrium' densities at Burrard Inlet. Data are means ± S.E. from cells counted in 30 M. splendens blades. S o o k e Mazzaella splendens 15 i spring'97 fall'97 spring'98 fall'98 B. 'CWorochyrriurri 80 i spring '97 fall '97 spring '98 fall '98 Sampling Day from Start F i g u r e 3.7 A . Mazzaella splendens density ± S.E. at Sooke over two years. The point-quarter method of density estimation was based on locating 8 random points along a transect. B. 'Chlorochytrium' densities at Sooke. Data are means ± S.E. from cells counted in 30 Mazzaella splendens blades. 95 Bamfield Figure 3 . 8 A . Mazzaella splendens density ± S.E. at Bamfield over two years. The point-quarter method of density estimation was based on locating 8 random points along a transect. B . 'Chlorochytrium' densities at Bamfield. Data are means ± S.E. from cells counted in 30 Mazzaella splendens blades. 96 lowest late fall to early spring. A total of 1840 Mazzaella splendens blades were microscopically examined over two years, and a number of observations regarding the seasonal condition of thalli was recorded (see Table 3.2 for a detailed account). Table 3.2 Seasonal condition of Mazzaella splendens thalli. Site Blade condition Epiphytes/-zoans Littorinids Other Burrard Inlet bleached blades Microcladia "veins"1 (May '96/'97/'98, Enteromorpha + 30-70 % July/Aug. '98) bryozoans Apr - Nov '98 tattered / diatoms "chewed" blades juvenile mussels (Aug - Jan) Sooke bleached blades Enteromorpha "veins"1 (Apr./Aug.'98) bryozoans 20-30 % tattered / diatoms Apr - Nov '98 "chewed" blades (Aug-Jan) few blades Dec-Jan Bamfield bleached blades Microcladia "veins"1 (May/Jun '98) (>50 % Jun-Oct) + 10-30 % tattered/ Enteromorpha June-Oct '98 "chewed" bryozoans blades (Aug-Jan) diatoms Endodictyon few blades Dec- Porphyra 1996 1 see Fig. 3.4 I, J 97 'Chlorochytrium' survival in drift Mazzaella splendens blades. After 7-10 days 50% of the 'Chlorochytrium' cells examined (about 50 cells) from the 12 "drift" M. splendens blades from Sooke had darkened at the apex and become olive-coloured with a more uniform texture. Some had developed a protuberance at the apex; pyrenoids were no longer visible and a few cells showed red pigment suggesting the presence of eyespots, as was observed in fertile cells releasing zoospores. None released zoospores. After one month 'Chlorochytrium' cells were still healthy-looking and most had acquired the fertile-looking condition; only two M. splendens blades remained healthy-looking. By mid-January (two months in laboratory seawater tank) host blades were in very poor shape, but average 'Chlorochytrium' size had increased to 180-200 um (from 120-160 um), cells looked fertile and healthy and zoospores could be seen within cells. The Mazzaella splendens blades colonised by 'Chlorochytrium' which had been placed in an outdoor tank at Bamfield in December 1998 were found dead a month later. When the tank was checked in early February green filaments (1-2 cm long) growing in the tank were identified as Acrosiphonia arcta. No green filaments were present in the control tank. DISCUSSION 'Petrocelis' dynamics. Conspicuous seasonal changes in 'Petrocelis' abundance occur in habitats such as those examined at Sooke and Burrard Inlet, contrary to work conducted by Paine et al. (1979) at the entrance to the Strait of Juan de Fuca, near Neah Bay, Washington State, and Dethier (1994) in the San Juan Islands, Washington State. Since 'Petrocelis' has the capacity for rapid growth and regeneration (as shown in the photographic time series, Fig. 3.5), it is necessary to monitor abundance over short time intervals, i.e. i f individual crusts are only sampled once or twice a year, it may appear to the observer as though crust patches are static, when indeed they may not be. Mean crust surface area at Sooke and Burrard Inlet peaked in summer / early fall 1997 and summer 1998 with the lowest surface area occurring in winter 1997. On average 'Petrocelis' crusts experienced as much as -58% surface area change over the winter and +61% in summer / fall; hence surface area is clearly not static. The net change of +21/30% over 12/18 months for patches which survived the observation period reflects 'Petrocelis' ability for rapid regeneration and vegetative growth. Dethier (1994) found 'Petrocelis' to have the fastest growth 98 rate (some individuals doubling in size over two years) among crusts sampled in the San Juan Islands, but she did not detect seasonal changes in 'Petrocelis' abundance, perhaps due to infrequent sampling (twice yearly). Paine et al. (1979) followed the fate of individual crusts at 1 to 6 month intervals over 6-7 years. Unlike my study, crusts examined by Paine and co-workers (1979) in general grew much more slowly (4% yr 1 increase in surface area) and on average experienced a period of variable growth (about 3 years), a period of size stability (3-4 years) and a terminal period of patch fragmentation. In the present study 'Petrocelis' patches at Sooke and Burrard Inlet fragmented within 6-8 months. Major differences between the two studies which may account for differences in crust performance are habitat type and crust patch size: the average size of patches was much greater (> 100 cm2) at the San Juan Islands site where the habitat consisted of a sandstone wall (occupied by few macroalgae) with 'Petrocelis' found 2.3-2.4 m above the zero tidal datum. At Sooke and Burrard Inlet 'Petrocelis' grew on granite boulders covered by a high diversity of macroalgae (absent in the winter), and was present throughout the intertidal. Paine et al. (1979) did, however, note that patches from closely neighbouring sites tended to behave differently, with some patches showing a seasonal change in percent cover. Furthermore, they were surprised by the microtopographic variability in individual crust performance (some increasing in size while others fragmented). This is consistent with results from Sooke: although the average of 30 individuals shows the pattern in Figure 3.2, large variation was evident among individuals. Had it been feasible to follow discrete crust patches at Bamfield, where crust size of individuals was comparable to San Juan Islands crusts, but substratum consisted of consolidated rock, it is possible that patches would have been observed to be more static. Although individual crust patches were much smaller at Sooke and Burrard Inlet than in the San Juan Islands and Bamfield, several boulders observed at Sooke were covered in crust fragments (Fig. 3.4E). This observation suggests patches on the same boulder may have once made up a single large continuous patch. There are of course other explanations such as habitat type and age of crust patches which may account for smaller patch size at Sooke and Burrard Inlet. "Resurrections" were suggested by this and the study of Paine et al. (1979), and are thought to occur as a result of algal tissue remaining among the rock matrix after crust patch disintegration (Paine et al. 1979, Kitting 1980). 99 The relatively low survivorship (61-72%) of crusts followed for up to 18 months has also not been reported previously. Paine et al. (1979) found 70% survivorship over 6-7 years, with many patches requiring a minimum of 4 years before undergoing fragmentation, and mortality being confined to the smaller individuals. In my study individual crusts began to break up within 6-8 months, and the smallest individuals (2.5-3.5 cm2) did not seem any more susceptible to mortality than larger patches. In light of the phenomenon of crust resurrection, mortality of individual crusts can only be ascertained with confidence over a much longer time period. Furthermore, juvenile barnacle overgrowth at Burrard Inlet may account for the apparent mortality of many of the patches. The question of whether 'Petrocelis' really lives for decades (Paine et al. 1979) on boulder strewn shores therefore remains unanswered. Recruitment is then also difficult to assess (recruit or resurrected crust?), especially when crust patches exist nearby. Also, the basal system of gametophytic Mastocarpus is indistinguishable from the crustose tetrasporophyte in the field (Dethier 1987). Misidentification of new crusts could occur either early in the spring when gametophytes are becoming newly established or when herbivore pressure is high and removes the upright portion oi Mastocarpus. Nonetheless, considering only three new crust patches were observed while monitoring 60 crusts at two sites over 12-18 months, it seems recruitment is low as concluded by Dethier (1994) and Paine et al. (1979). Furthermore, in the northern part of the latitudinal distribution range in the northern hemisphere, direct development of the bladed phase oi Mastocarpus from carpospores is believed to predominate over a heteromorphic alternation of generations (Polanshek and West 1977, Guiry and West 1983, Zupan and West 1988). Accordingly, crustose tetrasporophytes would not be expected to become established with any great frequency. With such low recruitment it seems puzzling that high rates of mortality would be the norm. Perhaps many of the crust patches which disappeared at Sooke and Burrard Inlet are still to be "resurrected" with time. Mechanism for seasonal pattern of 'Petrocelis' abundance. It seems intuitive that spring and summer represent favourable periods for growth of crusts at Sooke and Burrard Inlet. Other macroalgae, such as Ulva, Fucus, the bladed phase of Mastocarpus, Porphyra and filamentous reds present on boulders, also experienced growth at this time. Paine et al. (1979), unlike this study, documented an increase in percent cover (for those crusts exhibiting a seasonal pattern) from roughly October to May with lowest percent cover occurring in the summer. This 100 summer decrease is purported to be due to the annual sloughing of central tissues. Dethier (1987) found some reproductive crusts to be greenish and mucilaginous, and believed this to be a result of the sloughing of loose filaments with tetraspores. Since fertile 'Petrocelis' has only been found in late October to January (pers. observation, Dethier 1987), this phenomenon does not coincide with the systematic sloughing of tissue observed by Paine et al. (1979) in the summer. In my study green mucilaginous tissue that appeared to be disintegrating was noted in central portions of some crust patches only in the fall of 1996 (these crusts were reproductive) and in August 1998. Consequently, decreases in crust surface area and fragmentation at Sooke and Burrard Inlet can not be attributed to annual episodes of sloughing of reproductive tissue or senescing internal tissues. Tectura scutum, the predominant limpet at Sooke and Burrard Inlet, was observed in abundance on crusts in fall and winter, the period coinciding with loss of crust tissue. This led to speculation on the role of herbivory in reducing 'Petrocelis' crust abundance. Laboratory feeding experiments (Dethier 1994) and underwater listening techniques and observations (Kitting 1980) revealed that Washington and California Tectura scutum do readily feed on 'Petrocelis', and are capable of damaging and removing tissue. However, both Dethier (1994) and Kitting (1980) failed to suggest herbivory may play a role in decreasing 'Petrocelis' abundance. Kitting (1980) concluded that high densities of limpets on central California vertical rock faces did not decrease algal abundance detectably, but cited at least two cases where rock was exposed within a crust patch while limpets were observed feeding. Photographs showed that such damage persisted for a year or more while tissue was generated to slowly fill in the space. Dethier's (1994) experimental limpet manipulations in the San Juan Islands revealed significantly more loss of crust percent cover in treatments with normal (and higher than normal) densities of limpets than in limpet exclusions. What does the available evidence from this study suggest? Tectura scutum gut analysis was unsuccessful in revealing 'Petrocelis' filaments, but based on the above studies we can assume this limpet readily consumes and is capable of damaging 'Petrocelis'. A n abundance of T. scutum on crusts in Sooke and Burrard Inlet photographs in the fall and winter (the period which coincides with loss of crust tissue) and low densities observed in the summer, suggest herbivore pressure may be seasonal. And yet, can variable herbivore pressure alone explain crust decrease (detected within just one month for some crust individuals), fragmentation within 101 6-8 months and mortality, when contradictory evidence is offered by Kitting (1980) and Dethier (1994)? A combination of factors may better support a mechanism for this phenomenon. Herbivory may visibly decrease 'Petrocelis' abundance in winter in association with senescence of tissue of older crusts and abiotic factors. Environmental conditions are less favourable for growth in winter, and may actually tend to weaken central portions of crust tissues (e.g. the effect of higher than normal rainfall) making them more susceptible to grazing. Furthermore, environmental events such as higher than normal water temperature caused by El Nino may result in variable crust health from year to year. Senescence of tissue of older crusts may result in visible sloughing, or also contribute to weakening central portions of'Petrocelis'. Other factors which may affect crust health and contribute to decreasing crust abundance include bacteria cover and barnacle settlement. Thick cover of bacteria was shown by Dethier's (1994) laboratory study to result in sloughing of 'Petrocelis' tissues. In contrast to Paine et al.'s (1979) study where the presence of'Petrocelis' was concluded to inhibit the settlement of barnacles, at Burrard Inlet, where barnacle densities were high, the photographic time series unequivocally revealed the settlement of juvenile barnacles on 'Petrocelis1. Failure to recover in spring and summer from tissue loss in fall and winter (e.g. extremely high herbivore pressure, adverse environmental conditions and/or extensive senescence of tissues) could result in a decay process whereby the more permanent peripheral tissue becomes discontinuous and eventually the isolated fragments disappear. Further experimentation, e.g. limpet exclusion and Tectum scutum feeding behaviour studies, and observation over a longer time period are necessary to identify the key factors playing a role in this complex phenomenon at Sooke and Burrard Inlet. Implications for 'Codiolum' survival and duration in host. On average, regeneration of lost crust tissue and growth of'Petrocelis' occur from spring to early fall, coinciding with high densities of small endophytic 'Codiolum' cells (Figs 3.1, 3.2). Consequently 'Petrocelis' is abundantly available for colonisation by 'Codiolum'. It is unknown i f dense barnacle settlement on 'Petrocelis' (as shown at Burrard Inlet) may interfere with 'Codiolum' colonisation. The decrease in crust surface area observed in fall and winter coincides with the reproductive phenology of'Codiolum' (fertile 'Codiolum' cells were present from October - April, Chapter 2), implying that many endophytic cells may lose their host before they have had a chance to release zoospores. The question of whether 'Codiolum' can survive an epilithic or pelagic existence is 102 without answer. It is interesting to note that there exists synchrony of phenology between host and endophyte: 'Petrocelis' crust patches with divided tetraspores were observed from October/November - February/March, coinciding with fertile 'Codiolum'. What this also means is that some reproductive crust tissue is lost. This may, however, be of little consequence for 'Petrocelis' persistence if, as already mentioned, the foliose phase of Mastocarpus commonly develops directly from carpospores rather than the tetraspores of'Petrocelis' in the northern reaches of Mastocarpus distribution (growth may play a larger role than sexual reproduction in the maintenance of'Petrocelis'). Depending on herbivore pressure, environmental conditions affecting the health of crust tissues and the extent of any decay of tissue through senescence, 'Codiolum' survivorship may be quite variable from year to year. However, those 'Codiolum' present in crust portions which are consumed by limpets may be able to survive partial digestion in the herbivore gut. Defecation of still viable seaweeds is known from sea urchins (Santelices et al. 1983) and molluscs (Jernakoff 1985). The tendency for 'Petrocelis' to undergo conspicuous fluctuations in abundance at Burrard Inlet and Sooke, implies that it may not be as ideal a host as suggested by Paine et al. (1979). The duration of'Codiolum' of < 1 year was suggested by size class data and the 'Codiolum' life history schematic (Fig. 2.13, Chapter 2). This is consistent with large seasonal decreases in 'Petrocelis' abundance: if host tissue loss occurs every fall / winter, the exit of 'Codiolum' zoospores and consequent termination of Acrosiphonia's endophytic existence until new colonisation of'Codiolum' every spring, is not surprising. Kornmann (1964) concluded 'Codiolum' persists for more than one year within 'Petrocelis' in Helgoland, Germany (Chapter 2). To my knowledge 'Petrocelis' seasonal patterns are unknown for Helgoland. The fact that 'Petrocelis' crust patches on boulders in the intertidal of southwestern British Columbia are not as static as they were believed to be, suggests endophytes have evolved a strategy whereby they minimise their time spent in the host, becoming reproductive when host tissue loss is most likely to occur. As such their life history can be considered to be synchronised with that of their host. Mazzaella splendens dynamics. The seasonal changes in Mazzaella splendens density shown at Burrard Inlet, Sooke and Bamfield are similar to those observed by Dyck and DeWreede (1995) at Burrard Inlet and F. Shaughnessy (pers. comm.) at Bamfield in 1990-1991. Mazzaella splendens density peaks in the summer months when environmental conditions are favourable for growth, and decreases drastically in the winter. Lower light and temperature are 103 correlated with decreased abundance, and heavy winter storms dislodge blades. In relatively exposed environments such as Sooke and Bamfield M. splendens densities of < .1/m are not uncommon in winter. Heavy fouling of blades in late summer by bryozoans, diatoms and other epiphytes may also affect survival of M. splendens blades through fall and winter. Although littorinid snails have not been shown to graze on M. splendens, (they may instead be grazing on diatoms covering the blades), distinct marks that resemble "chewing" have been observed. Furthermore, Kim and DeWreede (1996) showed that littorinids readily consume Mazzaella cornucopiae (Postels & Ruprecht) Hommersand, a related species found higher in the intertidal, and concluded that these snails have a significant effect on the abundance of M. cornucopiae at Bamfield. Herbivory, then, when combined with other factors may also play a role in reducing M. splendens densities. Implications for 'Chlorochytrium' survival. The bladed red algal host, Mazzaella splendens, is present in high densities during 'Chlorochytrium' colonisation (Figs. 3.7, 3.8, 3.9). Factors which may reduce 'Chlorochytrium' survival during this time are episodes of bleaching of blades (desiccation and nutrient depletion) and epiphytisation which reduces available blade surface area. Mazzaella splendens densities decrease in the winter at the time when both M. splendens (90-100 % M. splendens blades collected from all three sites from October-February were reproductive) and 'Chlorochytrium' are reproductive. Again, as with fertile 'Petrocelis1, loss of reproductive M. splendens blades is implied. This may not be very detrimental for maintenance of M. splendens populations, since some blades may survive long enough to release spores, and more importantly, the perennial basal crusts not lost give rise to large numbers of new blades (Hansen 1977). Furthermore, fertile M. splendens from the subtidal (not subjected to dislodgement by winter storms) also help in maintaining intertidal M. splendens populations. When Mazzaella splendens blades are dislodged by winter storms many carry 'Chlorochytrium' which have not yet released zoospores. This phenomenon seems difficult to comprehend: how would such an endophyte / host relationship have evolved? There are several possibilities. Firstly, the dislodgement of M. splendens blades may serve as a trigger for 'Chlorochytrium' fertility. Alternatively, drift blades may be able to survive for quite some time with endophytes intact, or the endophytes may remain viable despite decay of drift blades. In years where winter storms are more severe than usual and dislodge a larger proportion of blades in the intertidal, subtidal M. splendens may still persist and provide host availability for 104 'Chlorochytrium' through the winter. And finally, 'Chlorochytrium' which have colonised the basal portion of M. splendens blades would be more likely to survive the winter. Based on observations of Mazzaella splendens blades colonised by 'Chlorochytrium' that had been collected at Sooke in November and left in a seawater tank, it does not seem that M. splendens blade dislodgement triggers 'Chlorochytrium' to become reproductive. Nor do M. splendens blades remain healthy long enough for 'Chlorochytrium' to develop and release zoospores. However, the fact that 'Chlorochytrium' cells remained healthy-looking after two months within decaying fragments of M. splendens blades, and were capable of increasing in size and resembling pre-zoospore release cells, suggests 'Chlorochytrium' may be able to achieve reproductive status despite deterioration of its host. Additional evidence for this hypothesis is provided by the Bamfield experiment: Acrosiphonia arcta was found growing in the same tank where M. splendens blades colonised by 'Chlorochytrium' had been found dead one month prior in January 1998. Although further experimentation with replication is necessary to confirm this result, the real possibility exists that 'Chlorochytrium' can persist and become fertile, releasing zoospores, despite mortality of its host. If this is the case, it is also significant in terms of means of dispersal for Acrosiphonia. Russell (1967) showed that drifting fertile fragments of the Ectocarpaceae were able to stay alive for several months and produce sporangia. He and others (Hoffmann 1987, van den Hoek 1987) argued that drifting mature thalli are an important means of dispersal. However, the fact that M. splendens blades are not particularly buoyant and thus sink, must be kept in mind when considering endophyte survival and dispersal via drift thalli. Of course, as with 'Codiolum', the question of whether 'Chlorochytrium' can survive free-living, once its host has disappeared, is of extreme interest. The fact culture studies (Jonsson 1959a 1959b 1962 1966, Kornmann 1961 1964, Chihara 1969, Hudson 1974, Miyaji 1996) showed 'Chlorochytrium' and 'Codiolum' capable of settling on glass and growing independent of their host, does not rule out this possibility. The presence of subtidal Mazzaella splendens year round (F. Shaughnessy, pers. comm.) could be important for 'Chlorochytrium' colonisation during years where winter storms are severe and leave little M. splendens attached in the intertidal. Although it was not possible to identify 'Chlorochytrium' cells in the basal crust of M. splendens (tissue being too thick), cells were often observed near the base ofthe blades. If generally more distal parts of the blade are lost, then endophytic cells located closer to the base stand a higher chance of surviving the winter within their host. 105 The life history schematic for 'Chlorochytrium' based on unicell size (Figure 2.7, Chapter 2) suggested an annual cycle for the endophytic cells. Like 'Codiolum', 'Chlorochytrium' in southwestern British Columbia, may have evolved to synchronise time spent in its host with the one year lifespan of blades of intertidal Mazzaella splendens. Even those M. splendens blades which persist through the winter, and thus enable 'Chlorochytrium' to become fertile within its host, are not believed to survive much longer than one year (pers. observation, pers. comm. L. Dyck). No data are available for the longevity of subtidal M. splendens other than the fact they are present in the winter. For those 'Chlorochytrium' which are vegetative at the time their host is dislodged from the substratum, evidence discussed in this chapter suggests they may still be capable of reproducing and releasing zoospores. Furthermore, should 'Chlorochytrium' colonise the basal crust of M. splendens, yet another means for 'Chlorochytrium' survival to reproduction exists. This chapter has established host availability and illustrated the large number of factors and complexity of interactions that play a role in the survival and reproductive maturity of 'Codiolum' and 'Chlorochytrium', Acrosiphonia's sporophytic phase. These host / endophyte relationships may be highly specific, i.e. obligate symbiotic associations involving a limited number of host species, or non-specific with endophytes employing a wide variety of algal hosts. Chapter 4 will investigate host specificity of Acrosiphonia's sporophyte, and whether the natural dynamics of the endophyte can be generalised for the range of hosts. 106 REFERENCES Adobe Photoshop® version 5.02. 1998. Adobe Systems Inc., San Jose, C A . Chihara, M . 1969. Culture study of Chlorochytrium inclusum from the northeast Pacific. Phycologia 8:127-33. Dethier, M . N . 1987. The distribution and reproductive phenology of intertidal fleshy crustose algae in Washington. Can. J. Bot. 65:1838-50. Dethier, 1994. The ecology of intertidal algal crusts: variation within a functional group. J. Exp. Mar. Biol. Ecol. 177:37-71. Dyck, L. J. and R.E. DeWreede. 1995. Patterns of seasonal demographic change in the alternate isomorphic stages of Mazzaella splendens (Gigartinales, Rhodophyta). Phycologia 34: 390-5. Guiry, M.D. and J.A. West. 1983. Life history and hybridization studies on Gigartina stellata and Petrocelis cruenta (Rhodophyta) in the North Atlantic. J. Phycol. 19:474-94. Hansen, J. E. 1977. Ecology and natural history of Iridaea cordata (Gigartinales, Rhodophyta) growth. J. Phycol. 13: 395-402. Hoffmann, A. J. 1987. The arrival of seaweed propagules at the shore: A review. Bot. Mar. 30: 151-65. Hudson, M . L. 1974. Field, culture and ultrastructural studies on the marine green alga Acrosiphonia in the Puget Sound region. Ph.D. Thesis, University of Washington, 206 pp. Jernakoff, P. 1985. Interactions between the limpet Patelloida latistrigata and algae on an intertidal rock platform. Mar. Ecol. Progr. Ser. 88: 287-302. Jonsson, S. 1959a. L'existence de l'alternance heteromorphe de generations entre VAcrosiphonia spinescens et le Codiolum petrocelidis. C. R. Acad. Sci., Paris 248:835-7. 1959b. Le cycle de developpement du Spongomorpha lanosa (Roth) Kiitz. et la nouvelle famine des Acrosiphoniacees. Ibid. 248:1565-7. 1962. Recherches sur des Cladophoracees marines: structure, reproduction, cycles compares, consequences systematiques. Ann. Sci. Nat. Bot. 12:25-230. 1966. Sur l'identification du sporophyte du Spongomorpha lanosa. Ibid. 262:626-9. Kitting, C. L. 1980. Herbivore-plant interactions of individual limpets maintaining a mixed diet of intertidal marine algae. Ecol. Monogr. 50: 527-50. 107 Kim, J. H . and R. E. DeWreede. 1996. Distribution and feeding preference of a high intertidal littorinid. Bot. Mar. 39: 561-9. Kornmann, P. 1961. Uber Spongomorpha lanosa und ihre Sporophytenformen. Helg. Wiss. Meeresunters. 7:195-205. 1964. Zur Biologie von Spongomorpha aeruginosa (Linnaeus) van den Hoek. Void. 11: 200-8. Krebs, C. J. 1999. Estimating abundance: line transects and distance methods. In Ecological Methodology. Addison-Wesley Educational Publishers, Calif, pp. 158-84. Miyaji, K . 1996. Taxonomic study ofthe genus Spongomorpha (Acrosiphoniales, Chlorophyta) in Japan. Phycol. Research 44:27-36. Paine, R. T., Slocum C. J. and D. O. Duggins. 1979. Growth and longevity in the crustose red alga Petrocelis middendorffii. Mar. Biol. 51: 185-92. Polanshek, A.R. and J.A. West. 1977. Culture and hybridization studies on Gigartina papillata. J. Phycol. 11:434-9. Russell, G. 1967. The genus Ectocarpus in Britain II. The free-living forms. J. Mar. Biol. Ass. U . K . 47: 233-50. Santelices, B., Correa, J. and M . Avila. 1983. Benthic algal spores surviving digestion by sea urchins. J. Exp. Mar. Biol. Ecol. 70: 263-9. SigmaScan Pro® version 5.0 for Windows. 1999. SPSS Inc. Chicago, IL. SPSS® version 9.0 for Windows. 1999. SPSS Inc. Chicago, IL. van den Hoek, C. 1987. The possible significance of long-range dispersal for the biogeography of seaweeds. Helgol Wiss. Meeresunters. 41: 261-72. West, J. A. , Polanshek, A. R. and D. E. Shevlin. 1978. Field and culture studies on Gigartina agardhii (Rhodophyta). J. Phycol. 14: 416-26. Zupan, J.R. and J.A. West. 1988. Geographic variation in the life history of Mastocarpus papillatus (Rhodophyta). J. Phycol. 24:223-9. Zar, J. H . 1996. Biostatistical Analysis, 3rd Ed. Prentice Hall, New Jersey, 662 pp. 108 CHAPTER 4 HOST SPECIFICITY OF ACROSIPHONIA'S SPOROPHYTE, 'CHLOROCHYTRIUM' AND 'CODIOLUM' INTRODUCTION A number of red algal hosts have been reported for 'Chlorochytrium inclusum' (Table 4.1), whereas 'Codiolum petrocelidis' has only been found in the red algal crusts 'Petrocelis' spp. and Haemescharia hennedyi. It is unknown i f all of the endophytic green unicells identified as 'Chlorochytrium' represent the sporophytic phase of Acrosiphonia. Nonetheless, the presence of putative 'Chlorochytrium' cells in such a wide range of red algal species, suggests Acrosiphonia's sporophyte may exhibit relatively low host specificity. In this study the range of algal host species for 'Chlorochytrium' and 'Codiolum' in southwestern British Columbia was investigated. Endophytes representing red, brown and green algae have been reported from numerous algal hosts (Garbary 1979, Boney 1980, Tarn et al. 1986, Correa et al. 1987, Peters 1991, Ellertsdottir & Peters 1997). Few studies, other than the extensive work on red algal parasites reviewed by Goff (1982), have, however, investigated host specificity of algae associated with other algae; fewer still have focused on host specificity of endophytes, and most of these were based on field observations. Recent cross-infection experiments, primarily with the green algal endophytes Acrochaete operculata Correa & Nielsen and Acrochaete heteroclada Correa & Nielsen (Correa & McLachlan 1991) and Endophyton ramosum Gardner (Sanchez et al. 1996), have more rigorously examined the range of hosts for algal endophytes. Colonisation of macroalgae by green algal endophytes has demonstrated low (lima & Tatewaki 1987, Correa & McLachlan 1991) to high (Correa & McLachlan 1991) host specificity. Sanchez et al. (1996) showed that endophytes may colonise a greater variety of hosts in laboratory cross-infection experiments than in nature, since Endophyton ramosum was capable of colonising algae that have not been found colonised in the field. The mechanisms involved in algal host specificity remain poorly understood. Host specificity may be ecologically determined, i.e. the endophyte requires a specific set of physiological conditions which are met in the micro-niche established on the host's surface (Dickson & Waaland 1985, Hansen 1986). On the other hand, since the endophyte must breach structurally and chemically diverse host components, by mechanical or enzymatic means or both, 109 Table 4.1 Red algal hosts reported for 'Chlorochytrium'. Note all herbarium specimens were examined by me for the presence of'Chlorochytrium'. Species Family Order Location Collector / Reference Mazzaella splendens Gigartinaceae Gigartinales Southwestern B.C. Puget Sound, Wash. Chapters 2,3 Hudson, 1974 Schizymenia pacifica Schizymeniaceae Gigartinales Puget Sound, Wash. Hudson, 1974 Chihara, 1969 Mastocarpus papillatus Phyllophoraceae Gigartinales Burrard Inlet, B.C. Puget Sound, Wash. C. Borden pers.comm. Hudson, 1974 Polyides rotundus Polyideaceae Gigartinales Helgoland, Germany Kornmann, 1964 Weeksia sp. Weeksiaceae Cryptonemiales Sitka, Alaska Setchell & Gardner, 1920 Constantinea subulifera Dumontiaceae Cryptonemiales Sooke, B.C. UBC Herbarium #A33090, 1968 Freeman, 1899 Dilsea californica Dumontiaceae Cryptonemiales Europe Cohn, 1872 Farlowia sp. Dumontiaceae Cryptonemiales Japan Miyaji & Kurogi, 1976 Sarcophyllis arctica Dumontiaceae Cryptonemiales Arctic Sea Kjellman, 1883 Kallymenia sp. Kallymeniaceae Cryptonemiales Vancouver Isl., B.C. UBC Herbarium #A33546, 1968 Halymenia sp. Cryptonemiaceae Cryptonemiales Ridley Isl., B.C. UBC Herbarium #A64855, 1981 Palmaria mollis Rhodymeniaceae Rhodymeniales Scandinavia Prince William Sound, Alaska Printz, 1926 UBC Herbarium #A60616, 1979 Porphyra sp. Bangiaceae Bangiales Vancouver Isl., B.C. UBC Herbarium #A39186, 1969 interactions between the chemical and structural composition of the host may determine host specificity. For colonisation of red macroalgae by green endophytes, both structural (Bird et al, 1981) and chemical (Correa & McLachlan 1991, Sanchez et al. 1996) characters have been implicated in host specificity. Recent studies of red algal hosts have established a close relationship between susceptibility to colonisation and the type of polysaccharides present in the host's cell wall. Endophyton ramosum, the causative agent of green patch disease in Mazzaella laminarioides (Bory de Saint-Vincent) Fredericq, was found to rarely infect agarophytes in cross-infection trials (Sanchez et al. 1996). Similarly, Acrochaete operculata, a green endophyte, only colonised the carrageenophytes Chondrus crispus Stackhouse and Iridaea cordata (Turn.) Bory when subjected to infection experiments with both agarophytes and carrageenophytes (Correa & McLachlan 1991). Structurally similar hosts, but with different cell wall compositions were 110 differentially infected, whereas structurally different hosts, but with similar cell wall compositions, showed similar susceptibility to infection. These studies suggest endophytes may be capable of discriminating between agar and carrageenan-producing algal hosts. In the Gigartinaceae the isomorphic generations, although structurally similar, have different carrageenan types in their cell walls (McCandless et al. 1975, 1983; DiNinno et al. 1979, Whyte et al. 1984, Correa-Diaz et al. 1990, Chopin et al. 1999). The sporophyte has X-type carrageenans (and any combination of 0, a, n carrageenans) and the gametophyte K-type carrageenans (including i carrageenan). Although Acrochaete operculata penetrated the outer wall of sporophytic and gametophytic fronds of Chondrus crispus in Correa and McLachlan's study (1991), subsequent development did not occur in gametophytes. A similar association was established with the sporophytic phase of Iridaea cordata, indicating that the life history phase / carrageenan type ofthe host may determine susceptibility to colonisation by A. operculata. Other species in the Gigartinaceae and species with A, carrageenan were not tested. Although no cross-infection experiments were carried out with putative hosts for 'Chlorochytrium' and 'Codiolum' in my study, susceptibility of colonisation between agarophytes and carrageenophytes and between different phases of a single host species was determined by comparing endophyte densities. To shed light on the host specificity of Acrosiphonia's sporophyte in southwestern British Columbia this study (1) determines the degree of host specificity for 'Chlorochytrium' and 'Codiolum' by examination of the foliose Rhodophyte species and crustose species present at Burrard Inlet, Sooke and Bamfield; (2) considers the role of structural and chemical characters of the host versus host availability / chance settlement for determining host specificity and (3) speculates on the natural dynamics of'Chlorochytrium' and 'Codiolum' over the range of their hosts. M A T E R I A L S A N D M E T H O D S Putative host species for 'Chlorochytrium' and 'Codiolum'. Collection of foliose red algae and crustose algae for identification of hosts for Acrosiphonia''s sporophytic phase was carried out from July 1995 to November 1998. Due to time constraints while working in the low intertidal zone, difficulty in removing the thinner crusts from boulders and the fact many foliose I l l red algal species were present in low numbers, no specific protocol was adhered to for collection of putative hosts. Bladed red algal species were collected haphazardly in the low to high intertidal zone (0.2 - 5.1m above zero tidal level) of the study sites at Burrard Inlet, Sooke and Bamfield whenever time permitted (approximately two or three times a year) and were representative of species present. Table 4.2 provides sampling dates at the three sites. The number of blades collected reflected the seasonal abundance of particular species. Mazzaella linearis (Setchell & Gardner) Fredericq and Mazzaella cornucopiae were collected outside the study site at Bamfield; M. linearis in the low intertidal zone in an area of higher wave exposure and M. cornucopiae in the high intertidal zone of a less wave-exposed site. At Sooke Mazzaella sanguinea (Setchell & Gardner) Hommersand was found as drift. Data from collections by students working under my guidance are included in this chapter: in May 1997 Jan Le Moux collected all foliose red algal species found at Burrard Inlet and Sooke, and in 1998 Dana Small's Honours project enabled a more thorough sampling of putative hosts for 'Chlorochytrium'. For the period February 1998 to November 1998 we focused sampling effort on the Sooke site (since the greatest densities of 'Chlorochytrium' in Mazzaella splendens were found there in 1997, Chapter 2). Sampling at Sooke was more intensive (eight occasions) than at Burrard Inlet or Bamfield (Table 4.2). A l l red algal blades were identified to species where possible, wrapped in newspaper and frozen. Examination for the presence of 'Chlorochytrium' involved thawing the blades and using magnification. For the species Chondracanthus exasperatus (Harvey & Bailey) Hughey and Mastocarpus papillatus the thickness ofthe blades prevented detection of endophytic cells with a dissecting microscope. Rather than engage in labour intensive cross-sectioning of blades for 'Chlorochytrium' detection, blades were bleached in boiling water for a few seconds. Blades became semi-transparent (the red algal phycobilin pigments are more soluble in water than chlorophyll), enabling them to be examined microscopically. Identity of the green unicells endophytic in the foliose red algal species collected was primarily based on morphological characters. Since high densities of cells were found only within Schizymenia pacifica, ITS sequencing was only employed for establishing the identity of endophytes in Schizymenia. The ITS1, ITS2 and 5.8s regions of endophytes from three isolates of Schizymenia were sequenced. The protocols for D N A extraction of Mazzaella splendens and 'Chlorochytrium' tissue (rather than the method requiring isolation of cells from host) and amplification and sequencing of'Chlorochytrium' from M. splendens blades (Chapter 2), were 112 Table 4.2 Collection dates for putative host species for 'Chlorochytrium' at Burrard Inlet, Sooke and Bamfield and their relative abundance from (+) low to (++++) high numbers. A (-) indicates the species may have been present but was not collected. Burrard Inlet Species July 95 Nov. 95 Jan. 96 Mar. 96 Jun 96 May 97 Apr. 98 July 98 Mazzaella splendens ++++ +++ ++ +++ ++++ ++++ +++ ++++ Mastocarpus papillatus - - - - - ++ ++ +++ Chondracanthus exasperatus - - + ++ - + + + Polyneura latissima - ++ + - +++ +++ +++ +++ Sparlingia pertusa - - - - + + + + Constantinea subulifera + - - - ++ ++ - ++ Porphyra sp. - - - + - + - -Bamfield Species July 95 Sept.95 Apr. 96 Jun 96 Apr. 98 Aug. 98 Mazzaella splendens ++++ +++ +++ ++-H- +++ ++++ Mazzaella linearex ++++ +++ - ++++ - -Mazzaella heterocarpa ++ - - - - ++ Mazzaella cornucopiae ++++ - - +++ - +++ Mastocarpus papillatus +++ + ++ +++ ++ +++ Chondracanthus exasperatus +++ - +++ +++ +++ +++ Polyneura latissima +++ - - +++ ++ +++ Sparlingia pertusa +++ + - ++ ++ ++ Constantinea subulifera + - - + - + Schizymenia pacifica + + - + - + Sooke Species Mar. 96 July 96 May 97 Feb./ Mar. 98 Apr./ May 98 June 98 Jul/Aug 98 Nov. 98 Mazzaella splendens + ++++ +++ + ++ -H-++ ++++ + Mazzaella heterocarpa - - - - - - ++ -Mazzaella sanguinea2 - - - - - - - + Mastocarpus papillatus - +++ +++ + ++ +++ +++ -Schizymenia pacifica - + + + + + + -Polyneura latissima - - ++ - ++ ++ ++ -Palmaria palmata 4~H- ++++ ++++ +++. +++ ++++ ++++ + Porphyra sp. + - ++ + - - - -Smithora sp. - - - - +++ - - -Callophyllis firma - - - - + - - -Rhodoglossum sp. - - - - - - + -Cryptopleura sp. - - - - + - - -1 collected from very wave exposed site 2 collected as drift 113 used. 'Chlorochytrium' density calculations were carried out as stated in Chapter 2 for cells within M. splendens. However, since a number of people were involved in estimating cell densities of different blades, it was necessary to ensure estimates among workers were comparable. Thirteen patches of the crustose brown alga, Ralfsia pacifica, and 18 patches of the red crust, Hildenbrandia occidentalis, were collected randomly from the mid intertidal zone at Bamfield (1.4 - 1.9 m above zero tidal level) in October 1998. At Sooke 30 patches of Ralfsia were obtained throughout the intertidal zone (0.2 - 2.1 m above zero tidal level) in May 1998. These sampling dates were chosen to correspond to periods of high 1997 'Codiolum' densities found in 'Petrocelis' at Bamfield and Sooke (Chapter 2). Only 'Petrocelis' was collected from Burrard Inlet. Crusts were removed by scraping patches of approximately 5 x 5 mm with a razor blade; these crusts were then air dried and later hydrated for detection of'Codiolum' cells. The same procedures described in Chapter 2 were carried out for 'Codiolum' detection and density calculation. Due to very low densities of endophytic cells, D N A analysis was not attempted; instead identification depended on cell morphology. 'Chlorochytrium' affinity for life history phase. Non-reproductive Mazzaella splendens blades, examined for 'Chlorochytrium', were analysed with resorcinol reagent (Dyck et al. 1985, Garbary & DeWreede 1988, Shaughnessy & DeWreede 1991) to determine their life history phase. A 5 mm diameter disk was removed from each M. splendens blade using a single-hole paper punch; disks were then air dried and tested with resorcinol. A red colour develops if K-type carrageenans are present, indicating the blade is gametophytic, whereas the tetrasporophyte remains colourless, indicating the presence of A.-type carrageenans. A control of known reproductive phase was included in all resorcinol tests. Data analysis. 'Chlorochytrium' densities within each host were compared with densities within Mazzaella splendens to determine if the host is as susceptible to colonisation as is M. splendens. Likewise, 'Codiolum' densities in Hildenbrandia and in Ralfsia were compared with densities in 'Petrocelis'. Independent samples t tests were performed using SPSS 9.0 for Windows (1999) for each host / M. splendens or 'Petrocelis' pair, as well as to determine whether 'Chlorochytrium' densities differed significantly among tetrasporophytic and gametophytic Mazzaella splendens blades. In several tests the assumption of equal variances was violated 114 (Levene Statistic), with transformations of the data failing to reduce heterogeneity to nonsignificant levels. A l l data, however, satisfied the Kolmogorov-Smirnov Test for normality, and t tests are sufficiently robust when sample sizes are large (> 30) and nearly equal (Zar 1996). In only a few cases, where foliose red algal species were scarce, sample sizes were small (e.g. n = 5, n = 8) and unequal to M. splendens sample size (approximately 30). Small sample size decreases the power of the test (Zar 1996), i.e. reduces the probability of detecting a difference between means (Type II error). Yet, non-parametric tests have an even greater probability of committing a Type II error, and thus were not considered for analysis. RESULTS 'Chlorochytrium' hosts. Sixteen foliose red algal species, in addition to Mazzaella splendens, were examined for the presence of 'Chlorochytrium'. Table 4.3 lists these species and collection sites; relative abundance of each species is illustrated in Table 4.2. Green endophytic cells were found in four of the 16 species: Mazzaella heterocarpa, Mazzaella sanguinea, Schizymenia pacifica and Sparlingia pertusa Saunders [= Rhodymenia pertusa (Postels & Ruprecht) Agardh]. No endophytes were present in blades of Mastocarpus papillatus, Constantinea subulifera, Palmaria mollis or Porphyra sp.; all species which have been reported as hosts for 'Chlorochytrium' (Table 4.1). Identity of endophytes within foliose red algae. A l l endophytes were identified to be 'Chlorochytrium inclusum' by morphological comparisons with 'Chlorochytrium' cells from Mazzaella splendens blades. Those cells found within Schizymenia pacifica from Sooke were present in high enough densities to conduct D N A analysis for confirmation of their identity. A l l three endophyte isolates from Schizymenia sequenced corresponded to 'Chlorochytrium' isolate sequences (sequences established in Chapter 1). Two of the endophyte isolates from Schizymenia were 100% identical to 'Chlorochytrium' isolates of the Alaskan Acrosiphonia sp. genotype. This genotype made up more than 50% of the 'Chlorochytrium' isolates sequenced from Sooke. The third endophyte isolate was identical to the 'Chlorochytrium' sequence corresponding to an Acrosiphonia arcta sequence. This genotype comprised approximately 30% of the 'Chlorochytrium' isolates from Sooke. 115 Table 4.3. 'Chlorochytrium' presence in putative host species. Note the much lower sample sizes of all species compared to Mazzaella splendens. Species Order Site collected Endophyte Present n Mazzaella splendens Gigartinales Burrard Inlet, Sooke, Bamfield + 1840 Mazzaella linearis Gigartinales Bamfield - 34 Mazzaella heterocarpa Gigartinales Sooke, Bamfield + 13 Mazzaella cornucopiae Gigartinales Bamfield - 67 Mazzaella sanguinea[ Gigartinales Sooke + 1 Schizymenia pacifica Gigartinales Sooke, Bamfield + 40 Chondracanthus exasperatus Gigartinales Burrard Inlet, Bamfield - 68 Mastocarpus papillatus Gigartinales Burrard Inlet, Sooke, Bamfield - 282 Rhodoglossum sp. Gigartinales Sooke - 3 Callophyllis firma Cryptonemiales Sooke - 3 Constantinea subulifera Cryptonemiales Burrard Inlet, Bamfield - 64 Palmaria palmata Rhodymeniales Sooke - 123 Sparlingia pertusa Rhodymeniales Burrard Inlet, Bamfield + 56 Smithora sp. Bangiales Sooke - 6002 Porphyra sp. Bangiales Burrard Inlet, Sooke - 28 Polyneura latissima Ceramiales Burrard Inlet, Sooke, Bamfield - 104 Cryptopleura sp. Ceramiales Sooke - 4 collected as drift Smithora blades epiphytic on 4 blades of eelgrass Mean densities of 'Chlorochytrium'. Table 4.4. provides mean densities of 'Chlorochytrium' for the different red algal species. Sparlingia pertusa was collected at both Bamfield and Burrard Inlet on several occasions (Table 4.2), but endophytes were only found in Bamfield blades collected in July 1995. The mean density of 0.02 cells cm"2 per blade (n = 16) is not significantly lower (p = 0.0156, t = 1.452 ) than Bamfield mean 'Chlorochytrium' density 116 Table 4.4 Mean densities o f Chlorochytrium'. Note the lower sample sizes of species compared to Mazzaella splendens. Species Site collected Date collected Endophyte density (# cm"2) n Schizymenia pacifica Sooke May '97 19.9 5 Mazzaella splendens Sooke May '97 22.6 30 Schizymenia pacifica Sooke May '98 47.5 5 Mazzaella splendens Sooke May '98 29.4 30 Mazzaella heterocarpa Sooke July '98 4.8 8 Mazzaella splendens Sooke July '98 10.6 30 Sparlingia pertusa Bamfield July '95 0.02 16 Mazzaella splendens Bamfield July '97/'98 0.1/4.2 30 Mazzaella sanguinea Sooke Nov. '98 1.2 1 Mazzaella splendens Sooke Nov. '98 2.7 30 (n = 30) of 0.1 cells cm"2 within Mazzaella splendens for July 1997, but significantly lower (p = 0.010, t = 2.742) than 4.2 cells cm"2 (n = 32) for July 1998. A single large blade (40 cm in length) of Mazzaella sanguinea collected as drift at Sooke in November 1998 was colonised by 2 2 1.2 cells cm" ; 2.7 cells cm" (n = 30) were found within M. splendens blades sampled at Sooke on the same day. Mazzaella heterocarpa collections from Sooke (n = 8) in July 1998 revealed colonisation by 'Chlorochytrium'. No significant difference (p = 0.523, t = 0.646) was found between mean densities of 'Chlorochytrium' cells within M. heterocarpa (4.8 cells cm"2) and within M. splendens (10.6 cells cm" , n = 30) collected at the same time. Endophyte densities in Schizymenia pacifica collected at Sooke in May 1997 (19.9 cells cm"2, n = 5) were not significantly different (p = 0.897, t = 0.131) from May 1997 'Chlorochytrium' densities in M. splendens (22.6 cells cm" , n = 27). Since more intense sampling was carried out in 1998 (i.e. Schizymenia blades were collected in all seasons) and endophytes were found on more than one collection day, it was possible to compare seasonal abundance of endophytes within M. splendens and within Schizymenia: 'Chlorochytrium' abundance peaked in late April / early May and density declined 117 toward zero by autumn for both hosts. 'Chlorochytrium' densities in Schizymenia in May 1998 (47.5 cells cm" , n= 5) and in July 1998 (19.4 cells cm" , n = 12) did not differ significantly (p = 0.593, / = 0.561; p = 0.527, t = 0.642 respectively) from densities in M. splendens (29.4 cells cm" , n = 23; 10.6 cells cm" , n = 30 respectively). High variance in the data and small sample sizes could have prevented detection of any significance, but since the similarity of the seasonal trends is clearly visible in the graph (Fig. 4.1), there is reason to believe that seasonal abundance of'Chlorochytrium' did not differ among M. splendens and Schizymenia. The small sample sizes for Schizymenia precluded endophyte density comparisons with Mazzaella for all sampling dates. 'Codiolum' hosts. Green unicells were found attached to and embedded within the filaments of the red algal crust, Hildenbrandia occidentalism and the brown algal crust, Ralfsia pacifica. The unicells were identified as 'Codiolum' by morphological comparison with the green endophytes found in 'Petrocelis'. Due to very low densities (0-75 cells in 1 cm 2 patches), D N A extraction, amplification and sequencing for endophyte identification was not feasible. The mean densities of 'Codiolum' in Hildenbrandia (24 cells cm"2, n = 18) and Ralfsia (12 cells cm"2, n = 13) from Bamfield were significantly lower (p = 0.037, t = 2.247) than in 'Petrocelis' (199 cells cm"2, n = 20) from Bamfield collected in fall 1997. Likewise mean 'Codiolum' density in Ralfsia (51 cells cm"2, n = 30) was significantly lower (p = 0, t = 4.479) than 'Codiolum' density in 'Petrocelis' (11,760 cells cm"2, n = 27) for Sooke samples collected in late spring 1997. 'Chlorochytrium' colonisation of gametophytic vs. tetrasporophytic phase. 'Chlorochytrium' densities in gametophytic Mazzaella splendens did not differ significantly (Table 4.5) from densities in tetrasporophytic M. splendens at Sooke and Burrard Inlet respectively. Tetrasporophytic M. splendens blades sampled at Bamfield, however, were colonised by significantly higher numbers (Table 4.5) of 'Chlorochytrium' than were gametophytic blades. When data from all three sites were combined 'Chlorochytrium' densities of tetrasporophytic and gametophytic M. splendens blades did not differ significantly (Table 4.5). 118 Figure 4.1 'Chlorochytrium' density in Mazzaella splendens and in Schizymenia pacifica at Sooke for 1998. Data are means ± S.E. from cells counted in 30 M. splendens blades and in 2-12 S. pacifica blades. 119 Table 4.5 'Chlorochytrium' densities in tetrasporophytic and gametophytic Mazzaella splendens blades. Results of independent samples r-tests performed for Burrard Inlet, Sooke and Bamfield and all three sites combined. The significant t value (*) indicates a statistically significant difference among endophyte densities in tetrasporophytic and gametophytic blades. Burrard Inlet tetrasporophyte 247 gametophyte 288 Mean S.D. S.E. 0.05529 0.3347 0.02130 1.820 0.069 0.1182 0.4619 0.02722 Sooke tetrasporophyte 275 10.7810 36.7495 2.2161 0.300 gametophyte 319 9.7425 46.2328 2.5885 0.765 Bamfield tetrasporophyte 235 4.9049 23.5178 1.5341 gametophyte 220 1.1781 4.7117 0.3177 2.379 0.018 A l l 3 sites tetrasporophyte 757 5.4572 26.0892 0.9482 gametophyte 827 4.1126 29.1369 1.0132 0.969 0.333 DISCUSSION The results from this, including Small's (1998) and Le Moux's (1997) studies, combined with previous reports (Table. 4.1), suggest a wide range of hosts for Acrosiphonia's sporophytic phase. This study, however, also suggests greater affinity for colonisation of 'Petrocelis', Mazzaella splendens, Schizymenia pacifica and Mazzaella heterocarpa than other foliose red and crustose algae present at Burrard Inlet, Sooke and Bamfield. Three foliose red algae, Mazzaella heterocarpa, Mazzaella sanguinea and Sparlingia pertusa, and the two crustose algae, Ralfsia pacifica and Hildenbrandia occidentalis, are newly reported hosts for 'Chlorochytrium' and 'Codiolum'. This is the first study to report 'Codiolum' colonisation of a brown alga (= Ralfsia), which prompts the question of whether 120 'Chlorochytrium' inhabits other bladed, e.g. brown, algae. 'Chlorochytrium' has, however, never been reported in brown algae. Laminaria saccharina, the only foliose alga present at Kitsilano Beach, Vancouver, was intensely sampled (n = 45) in May, the time Acrosiphonia is fertile (Chapter 2), and no 'Chlorochytrium' cells were detected. Clearly 'Chlorochytrium' and 'Codiolum' are not host specific for a particular species, family or even taxonomic order; Sparlingia pertusa, unlike the other hosts in this study, is in the Rhodymeniales. In addition, i f the identification of 'Chlorochytrium' in other studies is accepted (I have not examined type specimens nor did other authors base identifications on molecular work), the range of hosts for 'Chlorochytrium' extends to a number of species within the Cryptonemiales (Table 4.1). For host algae collected from outside the northeast Pacific, it is possible that 'Chlorochytrium' is the sporophyte of Spongomorpha (closely related genus to Acrosiphonia not found on our coast), and thus host specificities of Spongomorpha's sporophyte would not necessarily be shared by Acrosiphonia''s sporophyte. A number of host species for 'Chlorochytrium' reported in the literature, and collected in this study, were not found to be colonised by endophytic cells. The absence of'Chlorochytrium' in Mastocarpus papillatus was surprising, since Hudson (1974) found M. papillatus blades with 'Chlorochytrium' in the Puget Sound area, Washington. A single blade of M. papillatus colonised by 'Chlorochytrium' was , however, collected by C A . Borden in the vicinity of the study site at Burrard Inlet. The endophytes were identified by me to be 'Chlorochytrium'. 'Chlorochytrium' is also present in the sample of Constantinea subulifera from Sooke deposited in the U B C Herbarium in 1968. Unfortunately, Constantinea was not detected at Sooke during this study. Furthermore, 'Chlorochytrium' densities in other genera at both Bamfield and Burrard Inlet, where Constantinea was present (but in relatively low abundance), were much lower than at Sooke. The report of Palmaria mollis as a host was from Trondhjemsfijordes, Norway. In the only specimen (UBC herbarium specimen) from the northeast Pacific (Prince William Sound, Alaska) documented to be colonised by 'Chlorochytrium' I could not detect any endophytes, suggesting 'Chlorochytrium' may have been misidentified. It is interesting to note that a spherical green unicell, Halochlorococcum porphyra (Setchell & Gardner) West, Smith & McBride (= Chlorochytrium porphyrae Setchell & Gardner), is endophytic, embedded in the cell wall of some species of Porphyra from British Columbia to California (Setchell & Gardner 1920, Scagel 1966, Abbott & Hollenberg 1976). Reproductive and cytological characteristics 121 suggested to West et al. (1988) that the endophyte be placed in the genus Halochlorococcum (Chlorococcales), delineating it from 'Chlorochytrium1. The absence of'Chlorochytrium' in many ofthe foliose red algal species collected from Burrard Inlet, Sooke and Bamfield does not rule out the possibility that they are hosts. There are a number of reasons 'Chlorochytrium' many not have been found in the representatives I examined of these species; for example, (1) not all species were represented at Sooke where 'Chlorochytrium' densities were much higher than at Burrard Inlet and Bamfield, (2) some species were present in relatively low abundance during the time Acrosiphonia is fertile, translating to very low sample sizes and (3) some species may have a very patchy distribution and thus colonisation by 'Chlorochytrium' is less likely than for a species such as Mazzaella splendens, which is present throughout an area. Essentially, i f algal species which are not readily available as hosts in a given area are found absent of endophytes, it is difficult to conclude they are incapable of being colonised by Acrosiphonia's sporophyte. Laboratory cross-infection experiments would be valuable in determining a species' capability as a host. Based on the limited data of this study Mazzaella splendens, Schizymenia pacifica, Mazzaella heterocarpa and Mazzaella sanguinea were equally well colonised by 'Chlorochytrium'. 'Chlorochytrium' densities were not significantly different within Schizymenia and M. heterocarpa blades paired with M. splendens for a particular site and date; M. sanguinea is the exception since only a single blade was available for comparing 'Chlorochytrium' densities. Although 'Chlorochytrium' density in Bamfield Sparlingia pertusa was not significantly lower when compared to Bamfield M. splendens, dates of collections did not coincide, i.e. densities of 'Chlorochytrium' in M. splendens were not available for comparison in the year endophytes were found in S. pertusa. Furthermore, S. pertusa was not found to be colonised by endophytes in either of the years M. splendens was shown to be colonised. It is therefore very difficult to conclude anything about S. pertusa's status as a host. 'Petrocelis' was shown to be colonised by a significantly higher number of 'Codiolum' cells than Hildenbrandia and Ralfsia, indicating it is preferentially colonised. If some algal species are really not penetrated by 'Chlorochytrium' and 'Codiolum', or at least are less preferentially colonised, what determines which species are better hosts than others? One explanation for differential colonisation of algal blades and crusts by 'Chlorochytrium' and 'Codiolum' is variable structural composition among host species. 122 Upon examination of filaments among the three algal crusts, 'Petrocelis' filaments appear easier to penetrate than the more closely appressed filaments oi Ralfsia and Hildenbrandia. Similarly, degree of host specificity among foliose red algae may be determined by how tightly cells are compacted. 'Chlorochytrium' cells were generally located in the cortex layer of Mazzaella species and Schizymenia pacifica, all of whose cortex is comprised of small relatively loosely packed cells which become progressively smaller towards the surface (Abbott & Hollenberg 1976). The medulla consists of loosely interwoven filaments. This does not, however, address how 'Chlorochytrium' actually penetrates the surface of blades. Other species from the northeast Pacific similarly constructed include Mastocarpus papillatus, Constantinea subulifera, Chondracanthus exasperatus, Halymenia sp., Kallymenia sp., Weeksia sp., all but Chondracanthus found colonised by 'Chlorochytrium'. The absence of reports of 'Chlorochytrium' within Chondracanthus is somewhat surprising. However, the blades are very thick, such that unless the bleaching technique used in this study is applied, the chance of detecting 'Chlorochytrium' cells is low. Furthermore, no Chondracanthus was available at Sooke , where (in this study) 'Chlorochytrium' densities were highest. The cortex of Palmaria mollis and Sparlingia pertusa is only a few cells thick and the medulla is composed of large, compacted cuboidal cells, suggesting endophytic cells may have difficulty penetrating below the thin cortex (Fig. 4.2). None of the 123 P. palmata blades collected from Sooke were found to be inhabited by 'Chlorochytrium', and of the less than 10% of S. pertusa blades collected which contained 'Chlorochytrium', the latter was present in very low densities (0.01-0.1 cells cm"2 per blade). The representatives of the Bangiales, Porphyra sp. and Smithora sp., are structurally very different from the other foliose red algae. Parenchymatous in construction, a filamentous medulla is lacking, and cells of uniform size are densely packed. 'Chlorochytrium' cells have never been found in these genera, except that some Porphyra species are reported to be colonised by Halochlorococcum porphyrae. The surfaces of many macroalgae are believed to be covered by a layer of material called the cuticle (Craigie 1990). This layer is chemically and structurally distinct from the cell wall, and is formed from protein-rich cellular secretions that presumably move through the cell wall to the thallus surface. Few studies (Hanic & Craigie 1969, Gerwick & Lang 1977, Pedersen et al. 1981) have, however, successfully characterised the cuticle of an algal species structurally and chemically. It is feasible that host cuticle thickness and / or chemical composition may play a role in ease of penetration by the endophyte. B Figure 4.2 A comparison of blade cross-sections: Sparlingia pertusa (A) is characterised by a medulla comprised of large, compacted cuboidal cells whereas Mazzaella splendens (B) exhibits a medulla consisting of loosely interwoven filaments. 124 A second mechanism for host specificity considers variable chemical composition among host species. Experimental evidence (Correa & McLachlan 1991, Sanchez et al. 1996) strongly suggests that the green endophytes Acrochaete operculata (endophytic in Chondrus crispus and Iridaea cordata) and Endophyton ramosum (the causative agent of green patch disease in Mazzaella laminarioides) discriminate between hosts with different cell wall composition. Host specificity for both endophytes is suspected to be mediated by the presence of carrageenan in host cell walls. Cell wall composition of the algal species examined as hosts in my study is summarised in Table 4.6. The species are differentiated into carrageenophytes, agarophytes and carragar (agar / carrageenan hybrid) producing algae. Where no data were available for a particular species (studies tend to centre on species that contain large quantities ofthe commercially important polysaccharides), the cell wall composition was based on that ofthe taxonomic family the species is found in. From the table it is clear that the foliose species colonised by 'Chlorochytrium' in my study do not all share the same cell wall composition. However, those host species whose 'Chlorochytrium' densities were highest and not significantly different from Mazzaella splendens, were primarily carrageenophytes; Schizymenia being a producer of carragar. 'Codiolum's higher affinity for 'Petrocelis' than for Hildenbrandia can not be interpreted in terms of carrageenan presence, since the cell walls of neither Hildenbrandia nor any other members ofthe Hildenbrandiaceae have been examined. Although the presence of carrageenan may not dictate which species can or can not function as hosts for 'Chlorochytrium' (I have already suggested that 'Chlorochytrium' has a wide range of hosts from different taxonomic orders), the data suggest that carrageenophytes / carragar producing hosts are more readily penetrated by the endophyte. Furthermore, it may be that interactions between cell wall chemical composition and structural characteristics dictate differential endophyte success in colonising different host species. Andrews et al. (1979) suggestion that algal life history phases might be differentially susceptible to colonisation by other organisms, and Correa and McLachlan's (1991) confirmation by experimental evidence (Acrochaete operculata infected blades of host species of the same life history phase and containing X,-type carrageenans), was not supported by my study. Among host species with heteromorphic life history phases (suspected to be chemically different), neither gametophyte nor sporophyte seemed to be consistently more colonised. Mastocarpus papillatus, the foliose gametophyte (K-type carrageenan-containing), may be less colonised than 'Petrocelis', the crustose tetrasporophyte (A,-type carrageenan-containing); 125 Table 4.6 Cell wall composition of algal species examined for the presence of endophytes. Note all are members of the Rhodophyta, except Ralfsia pacifica, a brown alga. Species carrageenan / Agar Reference Endophyte present Mazzaella spp. KA carrageenans G) X/Q/a carrageenans (S) McCandless etal. 1975, 1983; Whyte etal. 1984, Correa-Diaz et al. 1990, Chopin etal. 1999 + Schizymenia pacifica carragars Whyte etal. 1984, Chopin etal. 1999 + Mastocarpus papillatus KA carrageenans Whyte etal. 1984, Correa-Diaz et al. 1990 + ( but not in my study) 'Petrocelis franciscana' X/Z/n carrageenans DiNinnoera/. 1979, McCandless etal. 1983, Whyte etal. 1984, Correa-Diaz et al. 1990 + Chondracanthus exasperatus KA carrageenans G) £, carrageenans (S) McCandless etal. 1983, Correa-Diaz et al. 1990 Rhodoglossum sp. K carrageenans (G), X carrageenans (S) McCandless et al. 1983 Callophyllis sp. X/Q/a carrageenans, carragasr Chopin etal. 1999 Constantinea subulifera carragars Chopin etal. 1999 + (but not in my study) Palmaria mollis agar Usov & Klochkova 1992 Sparlingia pertusa agar Usov & Klochkova 1992 + Porphyra sp. agar Usov & Klochkova 1992 Polyneura latissima agar Craigie 1990 Cryptopleura sp. agar Usov & Klochkova 1992 Smithora sp. agar Rees & Conway 1962 Hildenbrandia occidentalis unknown Hildenbrandiaceae unstudied + Ralfsia pacifica absent + G = gametophyte S = sporophyte Schizymenia,on the other hand, recently established as a gametophyte alternating with the crust Haematocelis (DeCew et al. 1992) showed very high colonisation by 'Chlorochytrium'. The crustose phase, Haematocelis, has, unfortunately, not been examined for endophytes. This is also the case for Farlowia's [bladed gametophyte found to be colonised by 'Chlorochytrium' in 126 Japan (Miyaji & Kurogi, 1976)] crustose tetrasporophytic phase. Of course, where phases are both structurally and chemically different, it is difficult to isolate whether host specificity is determined by interactions between the chemical and structural composition ofthe host or by only one property. For host algal species of the Gigartinaceae where the isomorphic generations, although structurally similar, have different cell wall compositions (McCandles et al, 1983), 'Chlorochytrium' did not appear to show a higher affinity for either the A.-type carrageenan-containing tetrasporophyte or K-type carrageenan-containing gametophyte. When endophyte densities in tetrasporophytic and gametophytic blades of Mazzaella splendens obtained from all three study sites were compared, no significant difference was detected. However, examining densities site by site, 'Chlorochytrium' cells were found to be present in significantly higher numbers in tetrasporophytes than in gametophytes at Bamfield; no difference was found for Burrard Inlet or Sooke. This discrepancy may be site-specific, e.g. tetrasporophytic blades at Bamfield may be structurally different from those at Burrard Inlet or Sooke. Correa and McLachlan (1988) suggested tetrasporophytic blades in the Gigartinaceae have a thinner cuticle than gametophytic blades. If this were true at Bamfield, it could explain differential colonisation between the life history phases. Host availability may be a factor in which algal species are colonised by 'Chlorochytrium' and 'Codiolum1. The endophytes may have evolved to colonise hosts present in abundance (whether by structural or chemical recognition), or chance settlement may dictate colonisation, i.e. i f algal hosts are less abundant than others during the period Acrosiphonia is fertile, the endophyte wil l have less chance for colonisation. The hypothesis of chance settlement determining an endophyte's success in inhabiting its host, is weakened, however, because Schizymenia pacifica at Sooke is both densely colonised by 'Chlorochytrium' and very low in abundance (Table 4.2). Furthermore, Ralfsia is as abundant as 'Petrocelis' at Sooke, yet 'Codiolum' densities were significantly lower in Ralfsia. The hypothesis that a requirement of a host is its availability long enough for the endophyte to reach maturity, is also not supported here. Mazzaella heterocarpa, found to be densely colonised by 'Chlorochytrium' in July at Sooke, disappears from the intertidal zone by late summer / early fall (pers. observ.), long before 'Chlorochytrium' has been observed to be fertile (Chapter 2). Since Schizymenia pacifica was the only host species collected over a one year period, and which was consistently found to be colonised by endophytes, it is difficult to speculate on 127 the natural dynamics of'Chlorochytrium' and 'Codiolum' for their range of hosts. Nonetheless, the similar patterns of seasonal abundance for 'Chlorochytrium' endophytic in Schizymenia and in Mazzaella splendens (Fig. 4.1) do suggest similar life histories for 'Chlorochytrium' and 'Codiolum' within hosts which are present year round. Nonetheless, more intensive and frequent sampling of putative hosts would be necessary to attempt generalising about the life history of Acrosiphonia's sporophyte for its range of hosts. Based on evidence presented in this chapter, Acrosiphonia's unicellular sporophyte employs a wide range of hosts. The evidence also suggests that 'Codiolum' has a greater affinity for 'Petrocelis' than for Hildenbrandia and Ralfsia, and 'Chlorochytrium' colonises some foliose red algal species more readily than others. Clearly, laboratory cross-infection experiments accompanying field observations of endophyte colonisation would greatly improve our understanding of host specificity of'Chlorochytrium' and 'Codiolum'. Both structural characteristics and cell wall composition of the host may play a role in differential colonisation. Availability of the host species does not appear to be an important factor in host specificity, as seen by high densities of'Chlorochytrium' in both Schizymenia (present in low abundance year round) and Mazzaella heterocarpa (a seasonal species absent late summer / fall until winter, R. Scrosati pers. comm.). Endophyte survival in M. heterocarpa, which disappears prior to 'Chlorochytrium' fertility, may be possible through 'Chlorochytrium' survival in drift blades and in the basal crust of M. heterocarpa (as was suggested for non-fertile endophytes in M. splendens blades dislodged by winter storms). In general, these results further support a bet-hedging strategy for Acrosiphonia's sporophytic phase. 'Chlorochytrium' and 'Codiolum' may have evolved to preferentially colonise and synchronise their life histories with hosts which are abundant and available long enough for the endophyte to reach maturity (M. splendens and 'Petrocelis', Chapter 3), but also seem capable of colonising a range of less 'suitable' hosts to compensate for the absence of primary hosts. In addition, some sporophytes of Acrosiphonia may survive epilithically or planktonically. A l l of these mechanisms for endophyte survival are significant since conditions which affect host availability (already mentioned in Chapter 3) are variable, e.g. herbivore pressure may fluctuate or winter storms vary in intensity. 128 REFERENCES Abbott, I. A . & Hollenberg, G. J. 1976. Marine algae of California. Stanford University Press, 827 pp. Andrews, J. H. , Vadas, R. L., Wheeler, W. N . , Neushul, M . , Woessner, J., Kohlmeyer, J., Dodds, J. A. , Kazama, F. Y . & North, W. J. 1979. Pathology of seaweeds: current status and future prospects. Experientia 35:429-50. Bird, C. J., Garbary, D. & Hansen, G. I. 1981. Observations on Ptilothamnionopsis lejolisea (Farl.) Dex. (Ceramiaceae, Rhodophyta) in northwestern North America. Syesis. 14:109-13. Boney, A . D. 1980. Post-attachment responses of monospores of some endophytic Audouinella spp. (Nemiales: Florideophyceae). Nova Hedwigia 33:499-507. Chihara, M . 1969. Culture study of Chlorochytrium inclusum from the northeast Pacific. Phycologia 8:127-33. Chopin, T., Kerin, B. F. & Mazerolle, R. 1999. Phycocolloid chemistry as a taxonomic indicator of phytogeny in the Gigartinales, Rhodophyceae: A review and current developments using Fourier transform infrared diffuse reflectance spectroscopy. Phycological Research. 47:167-88. Conn, F. 1872. Uber parasitische Algen. Beitr. Biol. Pflanzen, 1 (Abt. 2):87-106,1 pi. Correa, J. A. , Nielsen, R., Grund, D. W. & McLachlan, J. 1987. Endophytic algae of Irish moss (Chondrus crispus Stackh.). Hydrobiologia 151/152:223-8. Correa, J. A . & McLachlan, J. L. 1988. Variation in the fine structure ofthe cuticle of Chondrus crispus Stackhouse (Rhodophyta). Br. Phycol. J. 23:285-6. Correa, J. A . & McLachlan, J. L. 1991. Endophytic algae of Chondrus crispus (Rhodophyta). III. Host specificity. J. Phycol. 27:448-59. Correa-Diaz, F., Aguilar-Rosas, R. & Aguilar-Rosas, L. .E. 1990. Infrared analysis of eleven carrageenophytes from Baja California, Mexico. Hydrobiologia 204/205:609-14. Craigie, J. S. 1990. Cell walls. In Cole, K. M . & Sheath, R. (eds.) The Biology of Red Algae. Cambridge University Press, New York, pp. 221-57. DeCew, T. C , Silva, P. G., & West, J. A . 1992. Culture studies on the relationship between Schizymenia and Haematocelis (Gigartinales, Rhodophyceae) from the Pacific Coast of North America. J. Phycol. 28:558-66. Dickson, L. G. & Waaland, J. R. 1985. Porphyra nereocystis: a dual-daylength seaweed. Planta (Berl.) 165:548-53. 129 DiNinno, V. , McCandless, E. L. & Bell, R. A. 1979. Pyruvic acid derivative of carrageenan from a marine red alga (Petrocelis species). Carbohydr. Res. 71.C1-4. Dyck, L. DeWreede, R. E. & Garbary, D. 1985. Life history phases in Iridaea cordata (Gigartinaceae): relative abundance and distribution from British Columbia to California. Jap. J. Phycol. 33:225-32. Ellertsdottir, E. & Peters, A. F. 1997. High prevalence of infection by endophytic brown algae in populations of Laminaria spp. (Phaeophyceae). Mar. Ecol. Progr. Ser. 146:135-43. Freeman, E. M . 1899. Observations on Chlorochytrium. Minnesota Botanical Studies, Ser.2, part 2, pp. 175-90 Minneapolis. Gerwick, W. & Lang. N . 1977. Structural, chemical and ecological studies on iridescence in Iridaea (Rhodophyta). J. Phycol. 13:121-7. Garbary, D. 1979. A revised species concept for endophytic and endozoic members of the Acrochaetiaceae (Rhodophyta). Bot. Notiser. 132:451-5. Garbary, D. & DeWreede, R. E. 1988. Life history phases in natural populations of Gigartinaceae (Rhodophyta): quantification using resorcinol. In: Experimental Phycology: A Laboratory Manual (Ed. C. S. Loban, D. J. Chapman & B. P Kremer), Cambridge Univ. Press, Cambridge, M A , pp. 174-8. Goff, L. J. 1982. The biology of parasitic red algae. Prog. Phycol. Res. 1:289-369. Gonzalez, M . A . & Goff, L. J. 1989. The red algal epiphytes Microcladia coulteri and M. californica (Rhodophyceae, Ceramiaceae). II. Basiphyte specificity. J. Phycol. 25:558-67. Hanic, L. A. & Craigie, J. S. 1969. Studies on the algal cuticle. J. Phycol. 5:89-102. Hansen, G. I. 1986. A newly discovered host of the seagrass epiphyte Smithora naiadum (Bangiophyceae, Rhodophyta). Can. J. Bot. 64:900-1. Hudson, M . L. 1974. Field, culture and ultrastructural studies on the marine green alga Acrosiphonia in the Puget Sound region. Ph.D. Thesis, University of Washington, 206 pp. lima, M . & Tatewaki, M . 1987. On the life history and host specificity of Blastophysa rhizopus (Codiales, Cheatosiphonaceae), an endophytic green alga from Muroran in laboratory cultures. Jpn. J. Phycol. 35:241-50. Kjellman, F. R. 1883. The algae of the Arctic sea. K. Svensk. -Akad. Handl. 20:320-1. Kornmann, P. 1964. Zur Biologie von Spongomorpha aeruginosa (Linnaeus) van den Hoek. Helg. Wiss. Meeresunters. 11:200-8. 130 Le Moux, J. 1997. Study ofthe host specificity of two algal endophytes, "Codiolum petrocelidis" and "Chlorochytrium inclusum". Internship report, Institut national agronomique Paris-Grignon / University of British Columbia, 23 pp. McCandless, E. L. , Craigie, J. S. & Hansen, J. E. 1975. Carrageenans of gametangial and tetrasporangial stages of Iridaea cordata (Gigartinaceae). Can. J. Bot. 53:2315-18. McCandless, E. L. , West, J. A. & Guiry, M . D. 1983. Carrageenan patterns in the Gigartinaceae. Biochem. Syst. Biol. 11:175-82. Miyaji, K . & Kurogi, M . 1976. On the development of zoospores of the green alga Chlorochytrium inclusum from eastern Hokkaido. Bull. Jap. Soc. Phycol. 24:121-9. Pedersen, M . , Roomans, G. M . & Hofsten, A. V . 1981. Bromine in the cuticle of Polysiphonia nigrescens: localization and content.Phycol. 17:105-8. Peters, A. F. 1991. Field and culture studies of Streblonema macrocystis sp. nov. (Ectocarpales, Phaeophyceae) from Chile, a sexual endophyte of giant kelp. Phycol. 30:365-77. Printz, H . 1926. Die Algenflora des Trondhjemsfijordes, Skr. Norska Vidensk. Akad. Oslo, Mat. Nat., 5. Rees, D. A . & Conway, E. 1962. The structure and biosynthesis of porphyran: a comparison of some samples. Biochem. J. 84:411-16. Sanchez, P., Correa, J. A . & Garcia-Reina, G. 1996. Host specificity of Endophyton ramosum (Chlorophyta), the causative agent of green patch disease in Mazzaella laminarioides (Rhodophyta). Eur. J. Phycol. 31:173-9. Scagel, R. F. 1966. Marine algae of British Columbia & northern Washington, Part I: Chlorophyceae. Nat. Mus. Can. Bull. No. 207, Biol. Ser. No. 74, 257 pp. Setchell, W. A. & Gardner, N . L. 1920. The marine algae of the Pacific Coast of North America. Part IL Chlorophyceae. Univ. Calif. Publ. Bot. pp. 149-50. Shaughnessy, F. J. & DeWreede, R. E. 1991. Reliability of the resorcinol method for identifying isomorphic phass in the Gigartinaceae (Rhodophyta). J. Appl. Phycol. 3:121-7. Small, D. M . 1999. Identification and host specificity of the green algal endophyte 'Chlorochytrium'. Honours Thesis, University of British Columbia, 31 pp. SPSS version 9.0 for Windows. 1999. SPSS Inc. Chicago, IL. Tarn, C. E., Cole, K . M . & Garbary, D. J. 1986. In situ and in vitro studies on the endophytic red algae Audouinella porphyrae and A vaga (Acrochaetiales). Can. J. Bot. 65:532-8. 131 Usov, A. I. & Klochkova, N . G. 1992. Polysaccharides of algae. 45. Polysaccharide composition of red seaweeds from Kamchatka coastal waters (Northwestern Pacific) studied by reductive hydrolysis of biomass. Bot. Mar. 35:371-8. West, J. A. , Smith, C. M . , & McBride, D. L. 1988. Observations on the marine unicellular endophyte Chlorochytrium porphyrae (Chlorophyceae). Bot. Mar. 31:299-305. Whyte, J. N . , Foreman, R. E. & DeWreede, R. E. 1984. Phycocolloid screening of British Columbia red algae. Hydrobiologia 116/117:537-41. Zar, J. H . 1996. Biostatistical Analysis, 3rd Ed. Prentice Hall, New Jersey, 662 pp. 132 G E N E R A L C O N C L U S I O N S This thesis has demonstrated the value of combining molecular studies with field work to expand our understanding of complex algal life histories and endophyte / host associations. This approach was successful in the identification of the alternate life history phase of Acrosiphonia and in revealing the natural dynamics of the life cycle in southern British Columbia. The relationship of Acrosiphonia's gametophyte and sporophyte had never been examined in nature; even fundamental studies on distribution, seasonal abundance and reproductive phenology of both phases were lacking. A unicellular sporophyte, found to be endophytic in a wide range of crustose and foliose red algae for a duration of < one year, was shown to alternate with a free-living filamentous gametophyte that is abundant in the rocky intertidal zone in spring and summer. Host availability and specificity were identified as important factors in the survivorship of the sporophytic endophytes. In addition, examination of the evolutionary relevance of Acrosiphonia's complex life history led to speculation that heteromorphy may have evolved as a bet-hedging strategy. D N A sequencing proved to be a powerful tool in the identification of the spherical 'Chlorochytrium' cells endophytic in Mazzaella splendens and the stalked 'Codiolum' cells found in 'Petrocelis'. D N A sequences of endophyte / Acrosiphonia pairs provided conclusive evidence that the unicells were the sporophytes of Acrosiphonia and not of other genera in the Codiolales reported to exhibit 'Codiolum' phases in their life histories. Lack of information in the ITS sequence data did not permit resolution of infrageneric relationships within Acrosiphonia. The morphological species, A. arcta and A. coalita did, however, show consistent ITS sequence variation. A number of interesting questions regarding Acrosiphonia and endophyte genotypes were generated in the study. These are detailed in the discussion of Chapter 1. Regardless of the number of Acrosiphonia and endophyte genotypes found, the sequence data showed that the morphologically different endophytes, 'Chlorochytrium' and 'Codiolum', can be produced by the same Acrosiphonia species, A. arcta. This supports Kornmann's hypothesis that 'Chlorochytrium' and 'Codiolum' represent alternate phenotypes of the same sporophyte. The sporophyte of A coalita was, however, only found to be associated with 'Codiolum', suggesting it may be host specific in southern British Columbia, only colonising crusts. Host specificity examined in Chapter 4 focused on the range of foliose (red) and crustose hosts for Acrosiphonia's sporophyte at Burrard Inlet, Sooke and Bamfield. Host specificity of 133 sporophytes of individual Acrosiphonia "species" was not further examined. Two general conclusions were drawn: firstly that the endophytic sporophytes colonise a wide range of hosts, and secondly that 'Codiolum' demonstrates a greater affinity for 'Petrocelis' than for other coexisting crustose algae, and 'Chlorochytrium' colonises some foliose red algal species more readily than others. Foliose brown and green algae have never been reported colonised by 'Chlorochytrium'. Endophyte density data from Chapter 4 suggested that variable structural composition among host species, i.e. how closely appressed crust filaments are and the degree of compactness ofthe cells ofthe cortex and medulla of foliose red algae, may be important for ease of penetration by the endophyte. Host cell wall composition is also suspected to play a role in preferential colonisation. Carrageenophytes and carragar producing hosts were colonised by higher densities of'Chlorochytrium' than were agarophytes. The hypothesis that host life history phases might be differentially susceptible to colonisation by endophytes was not supported in my study; among host species with an alternation of generations neither gametophytes nor sporophytes seemed to be consistently more colonised. The relationship between the unicellular endophytes and filamentous Acrosiphonia plants was established in nature. Acrosiphonia arcta and A. coalita, identified both morphologically and by ITS sequences, comprised the majority of Acrosiphonia plants in southwestern British Columbia, abundant in the low to mid intertidal zone throughout the spring and summer. Acrosiphonia's gametophyte became reproductive almost immediately after establishment, and endophyte colonisation of its host occurred subsequently, one to three months after Acrosiphonia appeared in the intertidal zone. Colonisation may occur as two major events, corresponding to two Acrosiphonia pulses, or continuously over the spring and summer. Although counter-intuitive, 'Chlorochytrium' colonisation was shown to occur earlier than 'Codiolum' colonisation. This was unexpected since A. coalita, associated only with 'Codiolum', was observed to become fertile before A. arcta. The endophytic sporophytes remain within their hosts, even long after Acrosiphonia has disappeared from the intertidal zone, maturing during the winter and releasing zoospores in winter and spring. Clearly the two morphologically distinct phases of Acrosiphonia differ with respect to tolerance of environmental conditions. The success in colonisation and survival of Acrosiphonia's endophytic sporophyte is largely determined by host availability. Mazzaella splendens, found consistently to be densely colonised by 'Chlorochytrium' in spring and summer, is the dominant foliose red alga at all study sites. Likewise, 'Petrocelis franciscana', abundantly colonised by 'Codiolum', is a conspicuous 134 crust in the intertidal zone along the southern coast of British Columbia. However, seasonal fluctuations in abundance were shown to occur for both hosts: a decrease in Mazzaella density was noted in the winter, primarily due to winter storms dislodging blades; loss of tissue of 'Petrocelis' crusts also occurred in (fall) winter and is attributed to a combination of herbivory, adverse environmental conditions such as rainfall, water temperature and light levels and tissue decay from senescence. Release of endophyte zoospores appears to coincide with the decrease in host abundance, such that the endophytes spend < one year in their host. In the case of 'Chlorochytrium', M. splendens blade dislodgement prior to endophyte maturity may be compensated by the survival of'Chlorochytrium' in drift blades, endophytes persisting in perennial basal crusts of Mazzaella not dislodged and an epilithic or planktonic existence for 'Chlorochytrium'. 'Codiolum' survivorship may be affected by the ability of cells present in 'Petrocelis' tissue consumed by limpets to survive partial digestion and the survival of cells epilithically or planktonically. Acrosiphonia's success in southern British Columbia has been indicated by gametophytic filamentous plants alternating with endophytic sporophytes at three environmentally different sites. Although variable seasonality and abundance of both phases is exhibited at different sites (see Chapter 2 discussion for details), this study suggests adaptability oi Acrosiphonia to a range of habitats. I suggest that Acrosiphonia's success may be attributed to its complex heteromorphic life history, which evolved as a bet-hedging strategy: (1) two morphologically different phases have adapted to a seasonally variable environment (no evidence is available suggesting an adaptation to grazing pressure); (2) the endophytic sporophyte phase (at least of Acrosiphonia arcta) successfully colonises both foliose red algae and crustose algae; (3) 'Chlorochytrium' and 'Codiolum' appear to have evolved to synchronise their duration as endophytes with host availability and (4) 'Chlorochytrium' and 'Codiolum', although showing higher affinity for some hosts, seem to exhibit relatively low host specificity and may exist epilithically or planktonically. Suggested Future Research As is generally the case in scientific investigations, many questions remain unanswered in this study of the life history oi Acrosiphonia. The major areas where additional research could potentially clarify aspects of the relationship oi Acrosiphonia's sporophyte and gametophyte in southwestern British Columbia are highlighted below. 135 Firstly, there is great need for a molecular study to resolve the taxonomic confusion of the genus. The ITS sequence data from this study lacked sufficient variation for resolution among Acrosiphonia species. A multiallelic population level approach or selection of a more variable gene, e.g. the even faster evolving IGS (intergenic spacer) region ofthe nuclear ribosomal D N A cistron, could enable the number of Acrosiphonia species and their corresponding endophytic sporophytes in British Columbia to be determined. If different Acrosiphonia species exhibit variable seasonal abundance patterns, reproductive phenology or host specificity (colonising only crusts of foliose algae), the life histories of individual species could be established. Another aspect of interest which commands further attention is the significance ofthe role the endophytic sporophytes play in Acrosiphonia's life history. Since 'Chlorochytrium' and 'Codiolum' become reproductive at a time when host availability has been shown to decrease, are other factors, yet undiscovered, involved? Perhaps the unicellular sporophytes can survive epilithically (or planktonically) as do those of Urospora, and as is suggested by free-living 'Chlorochytrium' and 'Codiolum' in culture and the presence of filamentous Acrosiphonia plants in a habitat devoid of hosts. Perhaps the relative contribution of sexual reproduction is low compared to asexual or vegetative propagation for maintenance of Acrosiphonia populations. As yet, it remains unknown if Acrosiphonia rhizoid overwintering is a means of propagation, and i f filamentous gametophytic plants can give rise to haploid zoospores in nature that survive the winter. Some of these questions could be experimentally tested through the use of sterilised settling plates placed in the intertidal zone or in large outdoor tanks, as described in Chapter 3 for testing the ability of'Chlorochytrium' to become fertile, despite mortality of its host. The feasibility of survival of'Chlorochytrium' in drift host blades also needs further investigation. An area of study where nothing is known is the mechanism for entry of'Chlorochytrium' and 'Codiolum' into their hosts and the exit of zoospores. Does host penetration occur by mechanical means, e.g. where surfaces are wounded or the cuticle of bladed algae is disrupted by growth, or by enzymatic hydrolysis or by both? Scanning electron microscopy, as well as experimental scarification of host blade surfaces are two techniques which may shed light on how Acrosiphonia's endophytic sporophytes get in and out of their hosts. With regard to the range of hosts employed by 'Chlorochytrium' and 'Codiolum' in southern British Columbia, more intensive sampling of putative hosts accompanied by laboratory cross-infection experiments, 136 would contribute to increased understanding of host specificity of Acrosiphonia's endophytic sporophytes. Lastly, herbivory as a factor in Acrosiphonia's life history along British Columbia's coast remains to be investigated. As mentioned in this thesis, herbivore pressure is speculated to be an evolutionary force in the heteromorphy of algae. Nothing is known of differential susceptibility of Acrosiphonia, Mazzaella splendens and 'Petrocelis' to grazing. Furthermore, the role of herbivory in reducing 'Petrocelis' crust abundance (as suggested by the photographic time series and abundance of limpets in this study) requires additional study. Feeding experiments and manipulative grazer exclusion experiments may be useful. In conclusion, this research identified the alternate life history phase of Acrosiphonia, and established the natural dynamics of the life cycle in southern British Columbia, through a combination of molecular and field studies. The value of utilising molecular tools to solve ecological questions has been made evident. The thesis contributes not only to our understanding of complex algal life histories and endophyte / host associations, but also addresses the adaptive significance of heteromorphic life cycles among algae. Increased knowledge of algal life histories in nature is vital for the cultivation or harvesting of economically important algae, the control / eradication of detrimentally invasive seaweeds and our comprehension of marine intertidal communities, biodiversity and ultimately conservation. 137 * NOTE: Alignments for isolate sequences in phylogenetic analysis (Chapter 1) have been entered in TreeBASE. APPENDIX A: Acrosiphonia percent cover means. Numbers in parentheses are sampling days from start. Burrard Inlet Sampling Date n Mean S.D. S.E. March 5/97 (100) 30 0.673 1.853 0.338 March 28/97 (123) 30 2.383 3.486 0.637 April 9/97 (135) 30 9.483 16.782 3.06 April 26/97 (152) 30 12.741 13.777 2.515 May 8/97 (164) 30 7.458 7.841 1.432 May 23/97 (179) 30 1.792 2.890 0.528 April 28/98 (519) 30 26.533 21.603 3.944 May 12/98 (533) 30 10.667 13.498 2.464 June 9/98 (563) 30 4.275 9.359 1.709 Sooke Sampling Date n Mean S.D. S.E. March 7/97 (118) 60 7.150 13.725 1.772 March 29/97 (138) 60 8.921 16.482 2.128 April 10/97 (160) 90 10.108 15.126 1.594 April 28/97 (178) 90 5.247 11.986 1.263 May 25/97 (205) 90 4.806 11.333 1.195 June 24/97 (235) 90 3.683 9.327 0.983 July 19/97 (260) 90 2.094 6.148 0.648 August 19/97 (291) 90 6.322 18.019 1.899 May 1/98 (545) 90 2.294 4.055 0.427 May 14/98 (559) 90 1.653 3.749 0.395 June 11/98 (577) 90 3.0 10.162 1.071 Bamfield Sampling Date n Mean S.D. S.E. May 24/97 (209) 60 4.112 10.828 1.398 June 22/97 (238) 60 0.867 2.327 0.3001 July 20/97 (266) 30 0.117 0.639 0.117 May 26/98 (576) 30 8.333 18.613 3.340 June 27/98 (608) 30 8.390 12.669 2.313 July 21/98 (632) 30 2.9333 6.061 1.107 APPENDIX B: 'Chlorochytrium'Density from start. 138 Means. Numbers in parentheses are sampling days Burrard Inlet Sampling Date n Mean S.D. S.E. April 9/97 (136) 30 0 0 0 April 26/97 (152) 30 0 0 0 May 23/97 (180) 33 0.0046 0.0164 0.0029 July 18/97 *236) 30 0 0 0 Sept. 14/97 (294) 29 0.0018 0.0069 0.0013 April 28/98 (519) 30 0 0 0 May 12/98 (534) 28 0.0370 0.0809 0.0153 June 9/98 (564) 29 0.6640 1.1850 0.2200 July 9/98 (594) 29 0.6550 0.9310 0.1730 Aug. 8/98 (624) 31 0.2160 0.3100 0.0557 Nov. 3/98 (711) 31 0.0491 0.0690 0.0124 Sooke Sampling Date n Mean S.D. S.E. March 29/97 (138) 27 0 0 0 April 10/97 (160) 29 12.160 11.6730 2.1680 April 28/97 (178) 27 22.571 43.8800 8.4450 May 25/97 (205) 27 23.578 35.0690 6.7490 June 24/97 (235) 28 6.304 8.3810 1.5840 July 19/97 (260) 25 5.788 11.6650 2.3330 August 19/97 (291) 29 23.3810 71.2710 13.2350 Sept. 16/97 (319) 30 2.3410 3.7110 0.6780 Oct. 17/97 (350) 39 3.3530 5.4870 0.879 Nov. 30/97 (394 32 0.4160 1.8740 0.3310 Jan. 11/98 (435) 31 1.3230 3.7200 0.6680 Feb. 25/98 (481) 33 0.1070 0.2240 0.0390 April 1/98 (516) 30 1.5160 5.2470 0.9580 April 16/98 (531) 29 9.1310 26.5590 4.9320 May 1/98 (546) 27 53.0290 146.3050 28.157 May 14/98 (559) 23 29.4420 73.7240 15.373 June 11/98 (577) 33 22.4070 48.6160 8.463 July 23/98 (620) 30 10.6110 45.6520 8.334 August 20/98 (648) 30 5.2220 17.0210 3.107 Nov. 7/98 (725) 28 2.6790 6.6390 1.254 Bamfield Sampling Date n Mean S.D. S .E . Oct. 27/96 (1) 26 7.2180 18.4110 3.6110 Dec. 9/96 (43) 32 9.3330 33.4260 5.9090 Feb. 5/97 (101) 32 0.1750 0.5190 0.0918 May 24/97 (209) 25 0 0 0 June 22/97 (238) 22 0.0020 0.0080 0.0017 July 20/97 (266) 32 0.1020 0.3120 0.05520 Oct. 17/97 (355) 32 0.6390 1.8390 0.3250 Dec. 2/97 (401) 32 0.1250 0.4310 0.0762 Feb.28/98 (489) 28 0 0 0 March 26/98 (515) 32 0 0 0 April 25/98 (545) 24 0 0 0 May 26/98 (576) 32 0.9850 3.6190 0.6400 June 27/98 (608) 29 6.7650 19.9040 3.6960 July 21/98 (632) 30 4.2540 8.4590 1.5440 Sept. 6/98 (679) 29 4.2770 16.1210 2.9940 Oct. 6/98 (710) 29 14.2360 47.6950 8.8570 140 A P P E N D I X C : 'Codiolum' Density Means. Numbers in parentheses are sampling days from start. Burrard Inlet Sampling Date n Mean S.D. S .E. Nov. 24/96 (1) 23 23523.910 21242.994 4429.470 Jan. 10/97 (48) 24 7965.833 5319.667 1085.877 March 5/97 (101) 30 926.333 2350.968 429.226 March 28/97 (124) 30 80.833 129.768 23.692 April 9/97 (136) 30 545.167 1303.724 238.026 April 26/97 (153) 30 576.667 522.961 95.479 May 23/97 (180) 27 11597.590 14691.292 2827.340 June 22/97 (210) 26 16510.000 16023.279 3142.423 July 18/97 (236) 30 11514.170 11357.357 2073.560 August 17/97 (266) 26 8048.462 13277.291 2603.891 Sept. 14/97 (294) 30 4968.333 4303.546 785.716 Oct. 18/97 (328) 26 3535.192 4882.857 957.607 Nov. 14/97 (355) 26 1843.846 5356.893 1050.573 Dec. 15/97 (386) 30 8.333 21.105 3.853 Jan. 28/98 (430) 26 1437.308 3531.957 692.674 Feb. 24/98 (457) 15 96.667 258.579 66.765 March 30/98 (491) 29 11.207 25.518 4.739 April 28/98 (520) 18 924.444 1663.357 392.057 May 12/98 (534) 25 8939.000 9937.591 1987.518 June 9/98 (564) 26 33094.23 30372.782 5956.593 July 9/98 (594) 28 93333.040 84676.282 16002.310 August 8/98 (624) 30 99111.000 82123.760 14993.680 Nov. 3/98 (711) 29 15602.610 38725.265 1562.825 Sooke Sampling Date n Mean S.D. S.E. Nov. 11/96 (1) 29 4699.138 4866.868 903.755 Jan. 8/97 (59) 25 1011.600 1417.323 283.466 March 7/97 (118) 29 276.724 507.793 94.295 March 29/97 (138) 29 37.931 56.149 10.427 April 10/97 (160) 27 783.333 1080.074 207.860 April 28/97 (178) 27 4329.815 3385.810 651.599 May 25/97 (205) 26 12038.080 13773.900 2701.284 June 24/97 (235) 29 17027.070 10837.888 2012.545 July 19/97 (260) 29 13376.550 7418.535 1377.587 August 19/97 (291) 29 14903.620 13403.218 2488.915 Sept. 16/97 (319) 28 12915.000 9376.810 1772.050 Oct. 17/97 (350) 28 13091.790 8870.571 1676.380 Nov. 30/97 (394 29 10166.210 8925.600 1657.442 Jan. 11/98 (435) 22 6196.364 8771.494 1870.089 Feb.25/98 (481) 29 903.103 809.151 150.256 April 1/98 (516) 29 462.069 389.346 72.300 April 16/98 (531) 29 1323.621 1896.518 352.175 May 1/98 (546) 29 661.379 621.078 115.331 May 14/98 (559) 27 8568.704 7859.414 1512.545 June 11/98 (577) 28 5601.786 6548.875 1237.621 July 23/98 (620) 29 23031.720 12569.741 2334.142 August 20/98 (648) 29 20197.070 17634.758 3274.692 Nov. 7/98 (725) 29 7341.724 10601.194 1968.592 Bamfield Sampling Date n Mean S.D. S.E. Oct. 27/96 (1) 30 491.667 746.640 136.317 Dec. 9/96 (43) 23 559.783 836.651 174.454 Feb. 5/97 (101) 10 140.000 295.851 93.556 May 24/97 (209) 22 32.954 101.590 21.659 June 22/97 (238) 30 86.667 178.926 32.667 July 20/97 (266) 30 134.167 236.942 43.259 Oct. 17/97 (355) 20 198.750 371.384 83.044 Dec. 2/97 (401) 30 255.833 403.684 73.702 Feb. 28/98 (489) 9 8.333 17.678 5.893 March 26/98 (515) 7 0 0 0 April 25/98 (545) 25 4.000 9.354 1.871 May 26/98 (576) 26 16.346 78.403 15.376 June 27/98 (608) 28 5.357 12.467 2.356 July 21/98 (632) 30 33.333 99.856 18.231 Sept. 6/98 (679) 30 109.167 213.396 38.961 Oct. 6/98 (710) 26 2947.500 2654.369 520.564 142 APPENDIX D: Surface area (cm2) of individual 'Petrocelis' crusts at Burrard Inlet for one year. Aug '97 Oct '97 Dec '97 Feb '98 March '98 May '98 June '98 August '98 2.396 2.9956 1.83016 3.3407 1.6899 11.424 6.991 10.196 2.65 4.3704 5.3954 0.1742 6.89 5.0889 10.408 7.0534 3.129 15.635 13.024 11.006 2.5585 7.423 4.375 8.859 2.312 6.743 3.5905 7.0098 3.0686 6.782 4.397 13.732 1.953 5.222 11.1214 1.4475 2.096 14.8008 4.545 10.172 1.97 7.57 0.32677 1.607 1.8166 5.4889 15.995 1.606 3.444 9.3018 4.366 2.1678 2.765 11.01 4.672 7.708 1.7067 42.367 3.3225 6.963 3.492 2.504 18.1125 5.486 12.83 4.435 3.3406 3.778 6.499 10.908 7.47 10.6 2.9267 8.372 5.896 3.804 7.1093 11.082 16.053 5.875 9.4325 5.412 0.7069 3.688 0 0 6.95 8.69 2.232859 0.79056 3.936 2.241 0 0 6.502 0 20.488 17.0685 1.7374 2.708 0 0.8103 0 11.948 3.9236 10.152 5.752 0 3.0938 0 3.09766 3.416 3.071 0.8754 0 0 0 7.96 23.632 2.2705 0.6103 0 0 0 4.617 7.577 2.4733 5.8465 0 0 0 2.9686 4.906 10.16 0 0 0 0 16.199 2.1215 15.57 0 0 2.741 8.468 1.207 0 18.635 13.347 2.068 0 6.446 4.111 4.85 5.247 1.078 0.43 3.669 1.3027 11.486 15.908 0 1.551 0 n = 24 n = 26 n = 26 n = 18 n = 12 n = 19 n = 21 n = 18 CO I CD > o C O OT I l 3 co OT oo ON ON 00 CO I c -3 cu 3 e O M o o GO t3 C O OT 1 >» CO CN CO co iri CM in CO cd co co o |r^ -CM co co csi co cq CN C D in C N co CM OT CO CO | C 0 CO CO co C M C M C M C O C O OT C O CO C O C O iri m m C O CD CM 1^ -CM C O C O ! C M co | o CM CM OT 00 iri CO co CM in C M CM CM o CM O CM CO CM 143 CO OT 1 C L < m in m o o CM in C\i co CM C O OT 1 u. in I in CM CO co CM CM (U o o "S 13 T3 T3 C+1 o co OT 1 O , 0 Q OT O CO CM co co I iri in CM m CM 00 OT 1 C CO OT in iri C M C O OT OT in co o co co CM CM co 1^  OT 1 O O co O CM in co I CM I OT C O C O co CM cu L H CO CO o C+H I H , GO s p* < IS-OT i CL CD CO co co o co C O OT m CM co o Ico in co CM OT OT 00 O CO CO 1^  CM O C O CM in co co OT C O in co 00 OT C N in co CM CM CO o o co 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089864/manifest

Comment

Related Items