Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Biosystematics of the endemic Hawaiian species of Lysimachia (Primulaceae) Marr, Kendrick Lloyd 1995

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

Item Metadata

Download

Media
831-ubc_1995-060160.pdf [ 3.87MB ]
Metadata
JSON: 831-1.0088322.json
JSON-LD: 831-1.0088322-ld.json
RDF/XML (Pretty): 831-1.0088322-rdf.xml
RDF/JSON: 831-1.0088322-rdf.json
Turtle: 831-1.0088322-turtle.txt
N-Triples: 831-1.0088322-rdf-ntriples.txt
Original Record: 831-1.0088322-source.json
Full Text
831-1.0088322-fulltext.txt
Citation
831-1.0088322.ris

Full Text

BIOSYSTEMATICS OF THE ENDEMIC HAWAIIAN SPECIES OF LYSIMACHIA (PRIMULACEAE) by KENDRICK LLOYD MARR B.A., The University of Colorado, 1981 M.S., The University of Hawaii, 1989 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Botany)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA April, ©  1995  Kendrick Lloyd Marr  ______________________  ____  In presenting this  thesis  in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department  of  The University of British Columbia Vancouver, Canada Date  DE-6 (2/88)  abstract  The endemic Hawaiian species of Lysimachia were examined using a combination of analyses of metric and non—metric morphological characters, studies.  allozyme variaton and crossing  The purpose was to produce a taxonomic revision and  to investigate the degree of genetic divergence and type of reproductive isolation that has accompanied speciation. Principal components analysis of metric characters produced an ordination of OTU’s,  that was used as a first step toward  understanding the range of variation.  However, greater  reliance was placed on non—metric characters for the taxonomic revision.  Sixteen species and four subspecies,  including  three previously undescribed species are now recognized. Allozyme variation was analyzed in 15 taxa from 48 populations.  Genetic identities ranged from 0.71-1.00,  indicating high genetic similarity and supporting the hypothesis of a monophyletic origin.  These values are  intermediate to those of other insular genera.  There is a low  correlation between morphological variation and allozyme variation.  Total genetic diversity, Ht, within species,  ranges from 0.02-0.22.  Genetic diversity within populations  does not decrease in an entirely linear manner from the oldest to the youngest island.  However, there is a step—wise  decrease in genetic diversity among taxa within an island from the oldest through to the youngest island.  ii  All interspecific crosses that were attempted produced fertile seeds.  Pollen stainability was quite variable in all  species and in the F 1 hybrids of most species pairs.  The  reduced stainability of the parents involved in the crosses makes it difficult to interpret the significance of reduced stainability of some hybrids.  Approximately 80% of self This appears to be due to  pollinations within flowers failed. protogyny,  not to self-incompatibility.  Adaptive radiation in Lysimachia has resulted in species that have diverged in corolla pigmentation, shape,  and  ecological preferences.  leaf size and  Speciation has occurred  without the types of divergence between species that often accompanies speciation on continents,  i.e.,  internal, post—  zygotic barriers to reproduction and low genetic identities. This pattern of morphological and genetic variation is similar to that seen in other insular groups.  iii  Table of Contents Abstract List of Tables List of Figures Acknowledgements Dedication  Page ii vi ix Xi xii  Chapter 1 1.1 1.2 1.3 1.4 1.5  Introduction 1 Overview of insular evolution 1 Affinities of the Hawaiian flora 3 Dissertation objectives 3 Previous studies of Hawaiian Lysiniachia 5 Geographical affinities of endemic Hawaiian Lysimachia 7 Chapter 2 Analysis of morphological variation and Taxonomic Revision of the Endemic Hawaiian Lysimachia 12 2.1 Introduction 12 2.1.1 Taxonomic position of the Endemic Hawaiian Lysimachia 12 2.1.2 Objectives 16 2.2 Methods and Materials 17 2.2.1 Source of specimens for measurements. .17 2.2.2 Metric measurements 18 2.2.2.1 Analyses of morphological data 25 2.2.3 Non—metric observations 26 2.3 Results 28 2.3.1 Principal Components Analysis 28 2.3.2 Distribution of non-metric characters 38 2.4. Discussion 42 2.4.1 Taxonomic conceptss 42 2.4.2 Taxonomic treatment 42 2.4.2.1 Taxonomic revision of L. filifolia s.1 43 2.4.2.2 Taxonomic revision of the remyi s.1./L. hillebrandii s.1. . complex 44 2.4.3 Summary of taxonomic revision 49 2.5. Key and descriptions of endemic Hawaiian Lysimachia 54 Chapter 3 Allozyme diversity in endemic Hawaiian Lysimachia 125 3.1 Introduction 125 3.1.1 Objectives of allozyme analysis 133 3.2 Materials and Methods 133 3.2.1 Sample collection 133 3.2.2 Electrophoretic procedures 138 3.2.3 Data Analysis 139 3.2.3.1 Analysis of allozyme diversity. .140 3.2.3.2 Principal Components Analysis.. .141 iv  3.3  3.4  Results . 143 143 3.3.1 Electrophoretic patterns Allozyme variation within 3.3.1.1 147 populations Allozyme variation among and 3.3.1.2 154 within species 163 Principal Components Analyses 3.3.2 Principal components analysis of 3.3.2.1 163 floral characters Principal components analysis of 3.3.2.2 166 vegetative characters Principal components analysis 3.3.2.3 166 using allele frequencies Comparison of morphological and 3.3.2.4 166 allozyme results 170 Discussion... Allozyme variability within 3.4.1 species of Hawaiian Lysimachia .170 3.4.2 Allozyme variability among 172 species of Lysimachia 3.4.3 Comparison of allozyme and 180 morphological variation 182 Taxonomic implications 3.4.4 183 3.4.5 Summary  Chapter 4 Fertility of artificial interspecific hybrids and breeding behavior of the endemic Hawaiian 184 Lysimachia 184 Introduction 4.1 186 Materials and Methods 4.2 186 4.2.1 Artificial crosses 190 Pollen stainability 4.2.2 191 Vacuolar flavonoids 4.2.3 192 4.3 Results 4.3.1 Fruit-set in Interspecific crosses.. .192 192 4.3.2 Fruit-set in seif-pollinations 195 4.3.3 Morphology of hybrids 201 4.3.4 Pollen stainability of species 202 4.3.5 Pollen stainability of hybrids 207 4.4 Discussion 211 summary 4.4.2 Chapter 5 Summary Findings of this research 5.1 5.2 Areas for future research  .213 .213 214 215  References  v  List of Tables Table  Description  Page  2.1  Collection locations of populations of endemic Hawaiian Lysimachia sampled for morphometric analysis.  19  2.2  Principal components analysis of Hawaiian Lysimachia using only vegetative characters.  30  2.3  Principal components analysis of Hawaiian Lysimachia using floral characters.  30  2.4  Principal components analysis of Hawaiian Lvsixnachia using vegetative and floral characters.  33  2.5  Principal components analysis of Hawaiian Lysimachia using vegetative and calyx characters.  35  2.6  Distribution of selected metric and non—metric characters among groups of populations previously classified as L. hillebrandii fl.j.  39  2.7  Distribution of selected metric and non—metric characters among groups of populations previously classified as L. remyi s.l.  40  3.1  Collection localities for populations of species of endemic Hawaiian Lysimachia sampled for allozyme analysis.  134  3.2  Geographic and taxonomic distribution of the highest frequency alleles for loci at which the allele with the highest frequency is different from the most common allele, for that locus, in taxa of Hawaiian Lysimachia.  145  3.3  Genetic variability in 48 populations of species and hybrid swarms of endemic Hawaiian Lysimachia.  148  3.4  Nei’s genetic diversity statistics calculated for 15 taxa of endemic Hawaiian Lysimachia.  153  3.5  Genetic diversity statistics of endemic Hawaiian Lysimachia for each island.  155  vi  List of Tables Table  Description  Page  3.6  Intraspecific variation in Nei’s Genetic Identities (I) for species of endemic Hawaiian Lysimachia.  156  3.7  Genetic identities (I) for species of endemic Hawaiian Lysimachia.  158  3.8  Nei’s genetic identities (I) within and among islands for endemic Hawaiian species of Lysimachia.  162  3.9  Principal components analysis of floral characters of endemic Hawaiian Lysimachia averaged over 35 populations.  165  3.10  Principal components analysis of vegetative characters of endemic Hawaiian Lysimachia averaged over 35 populations.  165  3.11  Principal components analysis of endemic Hawaiian Lysimachia based on allele frequencies from allozyme analysis of 35 populations.  167  3.12  Summary table of Pearson correlation matrix comparing principal component scores from allozyme, floral, and vegetative data.  169  3.13  Comparison of Nei’s genetic identities among species of island genera.  175  4.1  Localities and collection numbers of seeds of species of endemic Hawaiian Lysimachia used for artificial hybridizations.  187  4.2  Percent fruit-set of artificial crosses among species of endemic Hawaiian Lysimachia.  193  4.3  Percent fruit—set of seif—pollinations of species of endemic Hawaiian Lysimachia, intrafloral selfs and intraplant selfs.  196  4.4  Corolla pigmentation of Fl hybrids between species of endemic Hawaiian Lysimachia.  198  4.5  Pollen stainability in cotton blue of species of endemic Hawaiian Lysimachia.  203  vii  List of Tables Table  Description  Page  4.6  Summary table of pollen stainability in cotton blue of F 1 hybrids between species of endemic Hawaiian Lysimachia.  204  A3.l  Table of allele frequencies of Hawaiian Lysimachia populations.  225  A4.l  Pollen stainability in cotton blue of species of endemic Hawaiian Lysimachia.  237  A4.2  Pollen stainability in cotton blue of F hybrids between species of endemic Hawaiian Lysimachia.  242  viii  List of Figures Figure  Description  Page  2.1  Plot of principal components 1 and 2 using vegetative measurements.  29  2.2  Plot of principal components 1 and 2 using floral measurements.  31  2.3  Plot of principal components 1 and 2 using floral and vegetative measurements.  32  2.4  Plot of principal components 1 and 2 using vegetative and calyx measurements.  34  2.5  Plot of principal components 1 and 2 using vegetative and calyx measurements of Lysimachia filifolia ..],.  37  2.6  Plot of principal components 1 and 2 using vegetative and calyx measurements of Lysimachia hillebrandii s.1. and Lysimachia remyi s.1.  41  2.7.  Distribution of endemic Hawaiian species of Lysimachia on the island of Kauai.  50  2.8.  Distribution of endemic Hawaiian species of Lysimachia on the island of Oahu.  51  2.9.  Distribution of endemic Hawaiian species of Lysimachia on the islands of Lanai and Molokai.  52  2.10.  Distribution of endemic Hawaiian species of Lysimachia on the island of Maui.  53  2.11  Lysimachia iniki.  82  2.12  Lysimachia pendens.  95  2.13  Lysimachia scopulensis.  119  3.1  Representative leaves of endemic Hawaiian Lysimachia by island.  129  3.2  Genetic variation by island for populations of endemic Hawaiian Lysimachia are compared to average values for species of various geographical distributions.  152  ix  List of Figures Figure  Description  Page  3.3  UPGMA phenogram derived from Nei’s genetic identities of 48 populations of 17 species of endemic Hawaiian Lysimachia.  164  3.4  Comparison of principal components analysis of vegetative, reproductive and allozyme data of endemic Hawaiian Lysimachia.  168  4.1.  Leaves and flowers of parents and hybrid of L. filifolia (#11149.1) (female parent) X L. glutinosa (#255.1).  197  4.2.  Leaves and flowers of the cross L. remyi subsp. remyi (#410.101) (female parent) X L. kalalauensis (#271.1).  197  4.3.  Leaves and flowers of two siblings of the cross L. glutinosa (#255.8) (female parent) X . remyi subsp. remyi (#359.1).  200  4.4.  Leaves and flowers of three siblings of the cross j. glutinosa (#255.2) (female parent) X L. kalalauensis (#257.1).  200  4.5  Pollen from the cross . remyi subsp. (female parent) X j.. mauritiana.  206  x  remyi  Acknowledgements I am grateful to my supervisor Dr. Bruce Bohm for encouraging me to pursue biosystematic research and for giving me the freedom to choose a genus for investigation and for his careful editing of this manuscript. The other members of my supervisory committee, Drs. Jack Maze, Wilf Schofield and Gerald Straley have all made useful suggestions and provided encouragement. I especially appreciate their comments on the earlier drafts of this dissertation. Dr. Fred Ganders generously provided laboratory space for the electrophoretic work and several useful discussions regarding this aspect of the research. I also acknowledge the assistance of Drs. Thomas Wells, Stewart Schultz and Andre Desrocher in training me to do electrophoresis. Financial and/or logistical support was generously provided by the Hawaiian Plant Conservation Center, Haleakala National Park, the Cooperative Parks Studies Unit, University of British Columbia graduate student travel grant and the American Society of Plant Taxonomists. This research would not have been possible without the enthusiastic support of many field botanists in Hawaii. To all of these I extend my warmest “Mahalo”: Tim Flynn, Dave Lorence, Clyde Imada, Patty Welton, Art Medeiros, Paul Higashino, Ed Misaki, Bob Hobdy and especially Ken Wood and Steve Penman. Thanks to the personnel of Haleakala National Park, and the Nature Conservancy of Hawaii for permission to collect specimens on the lands they administer. I appreciate the many hours that Bob Kantimyer and Alan Reid spent in caring for the plants in cultivation in the greenhouse. Dr. Ann Dusing, Classics Department, University of British Columbia, provided the Latin diagnoses for the three new taxa. Leslie Bohm drew the illustrations for the newly described species. Dr. Fred Ganders and Dr. Helen Kennedy contributed immensely to the rewriting of the taxonomic revision and the key to species. I appreciate their careful attention to the nomenclatural details. Cheryl’s understanding, encouragement and love helped me to maintain a connection with the “real world”, my sanity, and a sense of humor during the final months of the writing by scattering magical moments of Joy amidst the long hours of writing and revising.  xi  Dedication  This dissertation is dedicated to the memory my father, Dr. John W. Marr (1914-1989),  a kind and gentle man,  and plant ecologist, who treated all people with dignity.  You taught me how to love, What it took, what it took, You never said too much, But still you showed the way, And I knew,  (From the song,  from watching you.  “Everthing I Own”, written by David Gates  for his father.)  And to my mother, Ruby and sister Linda, their love and encouragement.  xii  for all of  1  Chapter 1 Introduction  1.1  Overview of insular evolution Oceanic islands are volcanic in origin and have never  been connected to continents; their significance is seen in observations of the biota of the Galapagos Islands, which were important in helping Charles Darwin to formulate his ideas concerning both the fact and the mechanisms of evolution.  It  is this isolation from source areas of propagule dispersal that is perhaps the most important difference between evolutionary processes on continents and those on islands (Hubbell,  1968).  Darwin, however,  apparently did not realize  that the separation of two populations that had once belonged to the same gene pooi could result in the formation of new species, preferring instead to invoke natural selection as the primary force of evolution (Carson,  1987),  although Darwin’s  reliance on geographic variation does rely on a tacit assumption of isolation. Two additional features of oceanic islands that influence evolutionary processes are their relatively recent origin and the close proximity of a diversity of habitats in a small area  (Crawford et al.,  l987b).  Because habitats of an  individual island are generally much younger than continental ones, the events that lead to speciation are usually more recent and it is easier to discern the processes responsible for divergence among species  (Carson,  1987).  Internal  2  reproductive barriers among island plants in general appear to be lacking and it is usually easy to produce artificial F 1 and later generation hybrids as well as backcrosses.  This makes  it possible to examine the nature of reproductive isolation between species.  Furthermore from the frequency of various  traits among the progeny of such crosses,  it may be possible  to determine the genetic basis of morphological differences. After the passage of time, the descendants of founding populations may diverge into several species.  This process is  frequently referred to as “adaptive radiation,” defined by Hubbell  (1968)  as “the separation of the descendants of one  ancestral stock into numerous species,  adapted to live in a  variety of new situations by changes in form, ecological tolerances and requirements.”  function,  Cariquist  and  (1974) has  presented some of the most important trends in insular evolution in his 24 “Principles of dispersal and evolution.” Those that are most relevant to the present study are summarized as follows: events, most taxa, coastal areas,  (1)  Because of the rarity of dispersal  especially those that grow away from  are probably established at a single time and  there is no further genetic input from the ancestral species. (2)  Founding populations must overcome the restriction of  genetic material that is a product of the small size of the initial population if effects such as inbreeding are to be countered. mechanisms.  This often results in selection for outcrossing (3)  Adaptive radiation is inevitable where a  small number of founding taxa encounter a broad range of  3  ecosystems.  (4)  oceanic islands, stature.  (5)  New growth forms evolve among plants on especially a tendency toward increased  Pollination relationships correspond to,  change with respect to,  and  availability of insects and other  pollination agents on islands.  (6)  Some mutations that would  be lethal or disadvantageous in continental environments have a more neutral value in the less competitive environment of an oceanic island.  1.2  Affinities of the Hawaiian flora The Hawaiian Islands are perhaps the most isolated  oceanic islands in the world. California,  They lie about 2,000 miles from  3,400 miles from Japan,  450 miles to the nearest  small island and 850 miles south to the next island chain. Prior to the arrival of people, propagules that gave rise to the native flora arrived by floating in sea currents, natural rafts, birds,  on  attached to the feathers or in the gut of  or were carried by the wind  (Cariquist,  1974).  The  approximately 270 native Hawaiian angiosperm genera have affinities with all parts of the Pacific basin: Indo—Pacific,  18.3% are American,  16.5% Austral,  40.1% are 2.6% Boreal,  12.5% pantropic and cosmopolitan and 10.3% are obscure in origin Fosberg  1.3  (1983).  Dissertation objectives The endemic species of Lysimachia were chosen for this  study of insular evolution with two objectives in mind:  first  4  to produce a taxonomic revision; second, to attempt to understand the evolutionary processes that resulted in the divergence among species.  The need for a taxonomic revision  was evident from the discrepancy in the number of species recognized by two recent treatments.  St. John (1987)  published diagnoses of 44 new endemic species and four new combinations,  to add to the 10 previously described.  et al.  treated the genus as consisting of 10 endemic  (1990)  Wagner,  species,  but noted that a thorough biosysternatic study was  needed.  The second objective had two parts: one, to arrive at  an estimate of the level of genetic divergence that has accompanied adaptive radiation in Lysimachia as measured by electrophoretically detectable allozyme variation; and two, to evaluate the nature of reproductive isolating mechanisms among species. The results of this study are presented in three chapters.  In Chapter Two,  several areas relating to the  classification of the group are discussed including taxonomic history, a multivariate analysis of morphological characters, which provides insight into patterns of variaton,  and a  taxonomic revision at the end of the chapter based primarily upon qualitative non—metric characters.  The non—metric  characters proved to be more useful in grouping populations together based on the possession of shared combinations of characters.  In Chapter Three the results of allozyme analysis  are presented.  These results were used to estimate the level  of genetic variation within,  and divergence among populations  5  of the same species as well as between species.  The nature of  reproductive isolation among species was evaluated by making artificial crosses in the greenhouse and recording the pollen stainability (as an estimate of fertility)  of F 1 hybrids.  These results are presented in Chapter Four.  1.4  Previous studies of Hawaiian Lysimachia  Prior to the present research, there had been no investigation of any aspect of the biology of Lysimachia apart from their alpha taxonomy.  Even this information was  presented in a somewhat disorganized manner.  Gray (1862)  published the first description of the Hawaiian species. Hillebrand  (1888), Heller (1897)  and Knuth (1905)  provided  keys and descriptions as additional species were discovered. After 1905,  several more species were described but there were  no further comprehensive treatments of the Hawaiian species until Wagner et al.  (1990)  produced a revised and expanded  treatment in the Manual of the Flowering Plants of Hawai’i. At the time of his death in 1991, Hawaiian flora,  long-time student of the  Dr. Harold St. John had a manuscript in  preparation that included a key and lengthy description of 72 species.  This manuscript was never published; however, a copy  was loaned to me by the Bishop Museum in Honolulu.  Many of  the specimens that St. John named in this paper were collected as many as 60—95 years preceeding his manuscript. cases,  In some  St. John applied different names to specimens that bore  the same collection number.  In other cases,  specimens that  6  were collected several years apart, but from the identical location, were given different specific epithets. these serious deficiencies,  I found his treatment of the  taxonomic history to be especially useful. a second,  Despite  The page proofs of  not professionally published manuscript written in  1983 by Otto and Isa Degener, was also made available to me: “Plants of Hawaii National Parks Illustrative of Plants and Customs  South Seas”.  This, the third edition, was  never formally published, but was effectively published because copies were sent to many botanical institutions worldwide.  The only other analysis of Hawaiian Lysimachia are  chromosome counts of one species by Skottsberg (1953) two species by Carr (1978) (1947), Huynh (1970)  and of  and pollen analyses by Selling  and Bennell and Hu  (1983).  Nothing was known about the reproductive compatibility system of Hawaiian Lysimachia,  or of the existence of  mechanisms that might promote outcrossing such as dichogamy or heterostyly.  I have not found any information regarding  pollination of Lysimachia, apart from a note on a herbarium label that the flowers were more fragrant at night (this agrees with my observations of plants grown in the greenhouse).  Zimmerman  (1978)  lists three insect species  whose host plants are different species of Lyismachia, but no pollinators are mentioned. As far as I have been able to determine,  there have been  two previous attempts to grow Hawaiian Lysimachia under artificial conditions.  This is documented in an exchange of  7  letters written between 1938-1940, that are stapled to an herbarium specimen (Degener 17,672 at NY) Waianae Mtns.  collected from the  One set of correspondence is between Hawaiian  plant collector Otto Degener and English plant anatomist J.H. Priestly,  and concerns an attempt to cultivate this species in  a glasshouse from cuttings sent by Degener.  Dr. Priestly was  evidently interested in this species because it reportedly lacked secondary phloem and he wanted to investigate its potential suitability for physiological studies on translocation of plant foods. Leonard Croizat at Harvard.  Cuttings were also sent to There is no further record of the  success or failure of these attempts at cultivation.  1.5  Geographical affinities of endemic Hawaiian Lysimachia The combination of characters found in the endemic  Hawaiian Lysimachia is not found in any of the other approximately 180 species distributed worldwide.  This makes  it difficult to speculate on their geographical origin or sister group.  Hawaiian Lysimachia are characterized by their  shrubby habit, having alternate leaves, regularly dehiscent capsules,  axillary flowers, connate filaments adnate to the  base of the corolla, tetracolporate pollen and basifixed anthers.  The most common corolla color is red, that of one  species is green and another is white. 2n=72  (Carr,  1978)  Chromosome numbers of  are known for two of the endemic species.  Lysiinachia glutinosa Rock (mistakenly identified as L. kalalauensis Skottsb.)  and L. hillebrandii J.  D. Hook.  ex A.  8  Gray.  There are no counts for any other species.  indigenous j. mauritiana Lam. has 2n=20  (Carr,  The  1978).  A monophyletic origin for the endemic Hawaiian species (excluding the widespread lines of evidence.  .  mauritiana)  is supported by two  First, the pollen of the seven endemic  Hawaiian species that have been observed is tetracolporate, whereas that of non—Hawaiian species is tricolporate (Huynh, 1970,  1971; Bennel and Hu,  1983).  The most parsiminous  explanantion for this is that it was a character that evolved once, presumably from an ancestor with tricolporate pollen. Second, despite the remarkable range of morphological variation among the endemic species, they nevertheless have more in common with each other than they do with any non— Hawaiian species because all are woody, have a variable (5—10) number of floral parts, six)—merous)  (non—Hawaiian species are five (rarely  and most Hawaiian Lysimachia have a reddish  corolla, while extra—Hawaiian species generally have a yellowish or white corolla. Handel-Mazetti Champ.  (1928)  considered that L.  (a south China species)  shrubby red—flowered  (i.e.  “is nearly allied to the  endemic Hawaiian)  However, pollen studies of Huynh  (1970,  this nor could Chen and Hu (1979) support such a relationship. similarities,  alpestris  1971)  species.” did not support  find any characters to  Based on some pollen  Bennell and Hu (1983) very tentatively  speculated that the Hawaiian species were derived from section Alternifoliae  (Subgen. Lysimachia)  through  .  mauritiana.  9  Lysimachia mauritiana however is distinct from the endemic species in being a somewhat fleshy perennial herb, having 2n=20, distinct filaments that are versatile and an irregularly dehiscent capsule.  This species has always been  placed in a different subgenus from the endemic species. The single North American west coast species of Lysimachia,  L. thyrsiflora L.  (2n=40,  54)  is an herbaceous  species with opposite or whorled leaves and yellow flowers that occur in short, dense, pedunculate racemes in the leaf axils  (Hickman,  1993).  Such a plant is unlikely to be similar  to the ancestor of the Hawaiian species. Wagner et al. manuscript)  (1990)  and St. John (unpublished  suggested that Malesia  (the Malay Peninsula, and  all islands north of Australia and eastward to the easternmost Solomon Islands  (Carlquist,  for the Hawaiian Lysimachia. Hemsl., j. Baudo,  decurrens Forst.  L mauritiana Lam.,  .  1974))  Eight species, j. f..,  .  (Van Steenis,  capillipes  laponica Thunb., L.  montana  peduncularis Wall ex. Kurz and j. the Malesian area  is a likely source region  (Reinw.)  Bakh,  .  sikokiana Miq., are found in  1972).  All are herbaceous  with axillary infloresences and alternate leaves. yellow corollas,  laxa  except for L. decurrens Forst.  They have  f. which is  either white or red and L. mauritiana which is white or pink. St. John comm.)  (unpublished manuscript)  suggest that L.  laxa may be the species most closely  related to the Hawaiian species. i.  laxa and possibly L.  and Hu Chi Ming (personal  Of the species listed above,  capillipes, both in subgenus  10  Idiophyton,  are the only ones that have in common with the  Hawaiian species basifixed anthers with lateral dehiscence, and filaments connate and adnate to the corolla. leaf veins, lacking in  Marginal  which are present in all Hawaiian species, .  are  Ligneous stems are found in some  capillipes.  species of subgenus Idiophyton (Chen and Hu,  1979),  further  support for a connection between a species similar to L.  laxa  and the Hawaiian species. Based on corolla pigmentation alone, i. L.  mauritiana and  decurrens have more in common with the endemic Hawaiian  species.  However,  these two species differ in having  versatile anthers and the filaments are not connate at the base. Chromosome numbers are available for approximately 35 non—Hawaiian species,  but are not especially useful in  elucidating the extra—Hawaiian ancestor. are not available for j.  laxa or  .  Unfortunately counts  capillipes.  At least 19  different sporophytic numbers have been reported: 20, 108,  24, 112  28,  30,  32,  (Ornduff,  34,  36,  1967,  40,  1968;  42,  60,  Federov,  84,  92,  1981,  Johnson,  The count of 2n=36 is for j.  1990).  1985,  98,  1974; Moore,  1977; Goldblatt,  1984,  2n=16, 100,  18, 102,  1974,  1988; Goldblatt and nummularia,  a  yellow—flowered species with opposite leaves native to Europe. Its geographical distribution and morphological characters make it highly unlikely that this species is closely related to the Hawaiian ones,  despite the fact that a single polyploid  event would yield the same chromosome number as the Hawaiian  11  species. al.  (1986)  Base numbers of x=5,6,7 have been suggested by 1(0 et and Tanaka and Hizume  To date,  (1980).  one can only speculate on the extra—Hawaiian  ancestor of the Hawaiian Lysimachia.  A more accurate  identification of the ancestor of the Hawaiian species could provide information useful for classification,  as well as  providing insights into other processes involved in speciation and evolution,  such as changes in breeding system, pollinator  syndrome and ecological adaptation.  12  chapter 2 Analysis of morphological variation and taxonomic revision of the endemic Hawaiian Lysimachia  2.1  Introduction  2.1.1.  Taxonomic position of endemic Hawaiian Lysimachia  Lysimachia is one of the largest genera of the Primulaceae,  consisting of approximately 180 species of  upright or sprawling perennial or annual herbs, subshrubs.  shrubs or  The center of diversity is in southwest China,  where there are 122  (110 endemic)  species  (Chen and Hu,  1979).  The remaining species grow in temperate areas of the northern hemisphere,  the Southeast Asian tropics,  and Australia (Bennell and Hu,  1983).  South America, Africa  Lysimachia and  Anagallis are considered by Hutchinson (1969)  to be the most  primitive members of the family because of the contorted corolla lobes and the presence in some species of staminodes that alternate with the stamens.  The presence of secretory  cells in Lysimachia is an otherwise uncommon feature in the Primulaceae.  This may be an indication that this genus is  most closely related to another Priinulalean family, Myrsinaceae (Cronquist,  1981),  the  thus also placing Lysimachia at  or near the phylogenetic base of the family. Lysimachia has been divided into six subgenera and 18 sections, (1928).  based primarily on the work of Handel—Mazzetti In a taxonomic revision of the Chinese species, this  13  work included a subgeneric classification of the entire genus and emphasized floral structure, particularly the androecium. The subgeneric classification was modified somewhat by Chen and Hu (1979). Two subgenera occur in the Hawaiian islands. Palladia  (Moench)  Hand.-Mazz.  is represented by a single  indigenous coastal species, Lysimachia mauritiana. Lysimachiopsis  Subgenus  Subgenus  (Heller) Hand.—Mazz., consists exclusively of  species endemic to the Hawaiian islands.  Only the endemic  species were examined in the present study.  These have always  been considered to belong to a single subgroup within the genus  (Handel—Mazzetti,  1928), but there has been some  disagreement regarding their affinities and subgeneric position.  Heller  urceolate,  reddish corollas of the endemic Hawaiian species  (1897)  concluded that the shrubby habit and  (characters otherwise not found in the genus, at least among the species known at that time) were sufficiently distinctive to create a separate genus, (1905)  Lysimachiopsis Heller.  Knuth  did not regard the shrubby habit as a sufficient basis  for the creation of a separate genus and returned the Hawaiian species to Lysimachia.  Knuth  (1905)  did, however,  create  section Fruticosae Knuth solely for the endemic Hawaiian species. Klatt  Handel-Mazzetti  (1928),  combined section Cilicina  (a prior name for Fruticosae Knuth, published in 1866,  which Handel-Mazzetti cited as a synonym)  with section  Rosulatae Champ. to form subgenus Lysimachiopsis (Heller) Handel—Mazzetti.  Rosulatae contains one Asian species,  .  14  alpestris, a yellow—flowered prostrate herb.  Synonymy of  Cilicina with Fruticosae was unwarranted because Fruticosae was typified to consist only of the Hawaiian species.  Whereas  Cilicina Klatt appears to have been a heterogeneous group as Ray (1956)  included only part of Cilicina Klatt within  subgenus Lysimachia. Studies of pollen morphology from representatives of all sections shed further light on the distinctness of the endemic Hawaiian species.  Seven endemic Hawaiian species were  examined and all had tetracolporate pollen, whereas non— Hawaiian species possessed tricolporate grains, although some species may occasionally form tetracolporate grains 1970,1971; Bennell and Hu,  1983).  Pollen grains of Hawaiian  species were the largest in the genus, forbesii Rock  (Huynh,  the pollen type of L. Hawaiian species.  1970,1971).  (Huynh,  especially those of j.  Huynh  (1970)  observed that  alpestris differed from the endemic  Based on this observation Huynh (1970)  split subgenus Lysimachiopsis along sectional lines and elevated the sections to subgeneric status:  subgen.  Sandwicensia Huynh for section Fruticosae and subgen. Nullicaulis Huynh for section Rosulate.  Sandwicensia was an  unnecessary name because Lysimachiopsis was already validly published at the subgerieric level to include only the Hawaiian species.  Although subgenus Sandwicensia was mistakenly  retained by Bennel and Hu  (1983), the endemic Hawaiian  Lysimachia are properly classified in subgenus Lysimachiopsis (Heller)  Handel-Mazzetti.  15  Wagner et al. extinct species  (j.  (1990), recognized nine extant and one forbesii Rock)  the Hawaiian Islands. (A.Gray)  Hillebr.,  .  of Lvsiinachia endemic to  The extant species are j. daphnoides filifolia C.N. Forbes & Lydgate, j.  glutinosa Rock, L. hillebrandii Hook. kalalauensis Skottsb, j. Knuth)  f. ex A. Gray, i.  lydgatei Hillebr.,  St. John, L. remyi Hillebr., and  .  .  maxima  venosa  (R.  (Wawra)  St.  John. According to the taxonomic treatment of Wagner et al. (1990),  L. hillebrandii is distributed on Kauai, Oahu, Molokai  and Maui,  and L.  remyi on Molokai and Maui alone.  However,  using a combination of foliar and corolla characters, plants from Molokai fit the description for closely than they do that of  .  remyi,  fit the description of j.. remyi not circumscribed by Wagner et al. remyi as well as L. variation.  hillebrandii more  .  and plants from Maui hillebrandii.  As  (1990), j.. hillebrandii and J..  filifolia encompass a broad range of  These names are retained in the revised taxonomic  treatment presented here, however not in the same sense as in Wagner et al.  (1990).  In other words,  several additional taxa  are recognized either at the specific or the subspecific rank within the concept of these species as proposed by Wagner et al.  (1990).  In order to avoid confusion, whenever these three  species are discussed in the broadest sense, they are referred to as s.l.  .  hillebrandii  4.,  j.  remyi  and j..  filifolia  16  Species are found on all of the major islands except Kahoolawe, Niihau and Hawaii.  Species of Lysimachia occur in  a wide variety of these vegetation types including montane bogs, waterfall spray zones,  subalpine mesic shrublands,  montane dry and wet forests and lowland mesic shrublands, ranging in elevation from 250 Gagne and Cuddihy  in  to 2300  in  as classified by  (1990).  Hawaiian Lysimachia differ from each other in leaf and calyx lobe size and shape, phyllotaxy, pigmentation,  corolla size,  and the extent of pubescence.  a reddish-purple corolla,  shape and  Most species have  but that of j. cilutinosa is white to  cream and L. kalalauensis is green with a reddish base and veins  (as was L.  shrubs. of L.  forbesii).  All species are perennial woody  Most are scandent or even upright, with the exception  filifolia  which is pendulous.  The hermaphrodite  flowers are sweetly fragrant, especially in the evening,  and  are presumably insect pollinated.  2.1.2.  Objectives  In the tentative taxonomic treatment of the most recent flora of the Hawaiian islands, Wagner et al.  (1990)  observe  that Lysimachia is “greatly in need of careful monographic work,  especially the Lysimachia hillebrandii-L.  on the younger islands”  remyi complex  (i.e. Maui and Molokai) which  “virtually form a broad continuum of variation”,  Contributing  to the nomenclatural uncertainty and taxonomic confusion is the fact that St. John (1987)  described 44 new endemic  17  species, bringing to 54 the total number of names published. Many of these are referable to either I.!. hillebrandii j. remvi .i.  or  according to the classification of Wagner et al.  (1990) The purpose of the present analysis of morphological variation was to evaluate the tentative taxonomic treatment of Wagner et al.,  (1990)  and to determine whether or not there  are breaks in the continuous range of variation that had been described, s.1.  especially in the j.. hillebrandii  complex.  analysis  remyi  A multivariate tool, principal components  (PCA), was used as the first step in a quantitative  description of the variation in order to provide a framework for organizing taxonomic decisions.  ifowever,  ultimately, the  distribution of variation in non—metric, qualitative characters made the greatest contribution to the revised classification.  2.2  Methods and Materials  2.2.1.  Source of specimens for measurements  Field collections were made in the summers of 1990, and 1992.  1991  Priority was placed on seeking populations that  included the full range of variation,  as understood from  examination of herbarium specimens, without necessarily following the system of classification of Wagner et al. (1990).  The single known population of T. venosa is almost  inaccessible and was not visited. made of the type localities of  .  Unsuccessful searches were forbesii  (last collected in  18  1934)  as well as several other taxa  (sensu St.  John,  1987)  that have not been collected for 50-80 years. The Operational Taxonomic Units were individual plants.  (OTU’s)  in this study  Collection locations and population  codes are presented in Table 2.1.  Because the islands of Maui  and Oahu both resulted from two geographically and chronologically separate volcanic events,  collection  designations are subdivided into the Waianae Mtns. Mtns.  for Oahu,  size varied,  and East Maui and West Maui for Maui.  Sample  but an attempt was made to collect at least 10  flowering specimens from each population. populations,  and Koolau  no plants were flowering,  of fewer than ten individuals.  In some  still others consisted  A number of populations that  were represented only by herbarium specimens had a sample size of fewer than ten,  2.2.2.  sometimes one.  Metric measurements  Measurements were made on plants grown in the greenhouse from seed or cuttings,  pressed specimens from collections done  for this study,  and specimens borrowed from the following  herbaria:  F,  BISH,  GH, MASS, MO,  NY,  (abbreviations from Holmgrem et al.,  PTBG,  RSA,  US,  and W  1990).  The choice of characters measured reflects to some extent the observation of Wagner et al. leaf shape,  (1990)  that leaf spacing,  calyx and corolla size and shape and pedicel  length are the characters that show the greatest variation in  Kokee; Honopu Trail. Kokee; Edge of Honopu Valley. Kokee; Awaawapuhi Trail. Kokee; Rim of Kalalau Valley. Kokee; North of Makaha Valley Road. Kokee; Upper Nualolo Valley, Branch. Kokee; Maile Flats Trail. Kokee; Waialae Ridge. On ridgetop SW of Limahuli Falls. SW of Kulanaililia.  1230 1230 1110 1140 1230 1015 1080 955 1110 650 615  Kauai Kauai Kauai Kauai Kauai Kauai Kauai Kauai Kauai Kauai Kauai  GKALL GHONO KHONO KAAPU KKALL KMAKA KNUAP KKOHU KWAIA OLIMA OWAIN  glutinosa  glutinosa  kalalauensis  kalalauensis  kalalauensis  hillebrandii  j.  L. kalalauensis  kalalauensis  L.  L. kalalauensis  kalalauensis  L.  j.  L.  L. hillebrandii  Table 2.1 continued on next page.  L.  .  Wainiha Pali;  Kokee; Road between Kalalau and Puu 0 Kila Lookouts. Kokee; Below Kalalau Lookout.  1260  Kauai  GKALR  L. glutinosa  SE  Kokee; Kalua-Puhi trail.  1300  Kauai  GKALP  L. glutinsoa  .  Location  Altitude (meters)  Island  Pop. Code  Species  Table 2.1 Collection locations of populations of endemic Hawaiian Lysimachia sampled for Populations marked with an asterisk are represented by herbaria morphometric analysis. Specific epithets and species codes (in parenthesis) are sensu Wagner et specimens only. al. (1990).  Head of N fork of Wailua River. Head of N fork of Wailua River. Head of N fork of Wailua River. Head of N fork of Wailua River. Koolau Mtns.; Waiahole Ditch Trail. Koolau Mtns.; Waiahole Ditch Trail. Koolau Mtns.; Waiahole Ditch Trail. Waianae Mtns.; Ridge SE of Puu Kaua. Waianae Mtns.; Ridge E of Puu Kaua. Waianae Mtns.; Kapuna Gulch.  615 615 615 615 250 250 250 710 800 540  Kauai Kauai Kauai Kauai Oahu Oahu Oahu Oahu Oahu Oahu  FKTWO FKTHR FKFOU FKFIV FOONE FOTWO FOTHR HPUKS HPUKE HKAPU  filifolia  filifolia  filifolia  filifolia  filifolia  filifolia  filifolia  L.  L.  L.  L.  L.  L.  L. hillebrandii  i. hillebrandii  i. hillebrandii  Table 2.1 continued on next page.  .  Head of N fork of Wailua River.  615  Kauai  FKONE  filifolia  L.  Kokee; Alakai Swamp.  1230  Kauai  DBIGB  i.. daphnoides  1230  Kauai  DSECO  daphnoides  L.  Kokee; Below and E of Kalalau Lookout, North-South ridge. Kokee; Alakai Swamp Trail.  800  Kauai  NSPTR  pendens  Location  Altitude (meters)  Island  Pop. Code  L.  Species  Collection locations of populations of endemic Hawaiian Lysimachia continued. Table 2.1 Specific epithets are sensu Wagner et al. (1990). sampled for morphometric analysis.  0  Waianae Mtns.; Kamaileunu Ridge in Makaha Valley; No of small peak N of Puu Kawiwi. Waianae Mtns.; N fork of Palwai Gulch. Waianae Mtns.; Ridge between Lualualei and Nanakuli Valleys. lao Valley; Black Gorge. lao Valley; Nakalaloa Stream. Hanaula. Hanaula. Hanaula; rim of Ukeiuehame canyon. Manawainui Plant Preserve. Lihau,  Lihau sumiuit.  800 740 830 340 500 1050 1075 1170 860 1130 1260 985  Oahu Oahu Oahu W. Maui W. Maui W. Maui W. Maui W. Maui W. Maui W. Maui W. Maui W. Maui  HWAIK HPALI HLUAL WMBLG WNNAK WMMN2 WNMN3  WNUKE WMNPP WMLIL WMLIS WMHPL  hillebrandii  hillebrandii  hillebrandii  remyi  remvi  remyi  remyi  remyi  L.  L.  L.  L. remvi  remvi  j.  J.. remyi  i.e. remyi  Table 2.1 continued on next page.  .  .  .  L.  Halepohaku,  ridge SW of summit.  ridge west of summit.  Waianae Mtns.; Pahole Gulch.  540  Qahu  HPAHO  L. hillebrandii  .  Location  Altitude (meters)  Island  Pop. Code  Species  Collection locations of populations of endemic Hawaiian Lysimachia continued. Table 2.1 Specific epithets are sensu Wagner et al. (1990). sampled for morphometric analysis.  I-.’  E. E. Maui E.  WMHEL EMPAL EMHPA EMKUI EMKUK EMLWA EMKAW EMKPR EMKAE EMWAI EMKOT EMKOB  remyi  remyi  remyi  remvi  J.. remvi  remyi  J.. remyi  remvi  remyi  remyi  remyi  L. remyi  Maui  Maui  E. Maui  Maui  Maui  E. Maui  E. Maui  E.  E. Maui  E. Maui  E.  E. Maui  W. Maui  Table 2.1 continued on next page.  .  j.  .  j.  j..  j.  .  .  .  W. Maui  WMHPY  lydgatei  .  j.  W. Maui  Island  WMHPS  Pop. Code  remyi  Species  2920  1970  2030  1845  2300  1785  2155  2280  2215  2090  2000  1230  1000  1050  Altitude (meters)  E of summit.  ridge SW of summit.  rim of Kipahulu Valley.  below spring.  Koolau Gap,  Koolau Gap,  below treeline.  treeline.  Koolau Gap, Waikau Cabin site.  Kaupu Gap below Paliku.  Kalapawili Ridge; rim of Kipahulu.  W Kaupo Gap,  Above Lake Waianapanapa.  Kuiki,  Ridge N of Kuiki.  Ridge SE of Paliku.  Ridge SE of Paliku.  Helu,  Halepohaku,  Halepohaku summit.  Location  continued. Table 2.1 Collection locations of populations of endemic Hawaiian Lysimachia sampled for morphometric analysis. Specific epithets are sensu Wagner et al. (1990).  Es.) Es)  Makakupaia; ridge S of Onini gulch. Makakupaia; bottom of Onini Gulch. SE of Puu Kaeo.  Waikolu;  Small ridge between Kauanakakai and and Kupaia Gulches. Pelekunu; N of Ohialele.  970 850 1050 920 950  Molokai Molokai Molokai Molokai Molokai Maui  NMAKA MONIN MWAIK MKAUN MMAXI  hillebrandii  hillebrandii  L. hillebrandii  hillebrandii  Oahu Oahu Oahu Oahu Oahu  FORBE* PUNAL* HALAW* KALIB* MOANA*  forbesii  L. hillebrandii  L. hillebrandii  hillebrandii  hillebrandii  L.  Table 2.1 continued on next page.  .  .  E.  EMKAL  hillebrandii  j..  L. maxima  j.  .  .  unknown  400-500  unknown  unknown  600-710  1510  W of Puu Kolekole,  1040  Molokai  NICOLE  hillebrandii  L.  920  Molokai  MKAWE  L. hillebrandii  lower Kaupo Gap.  Koolau Mtns.; Moana Valley Pali.  Koolau Mtns.; Kalihi Pali.  Koolau Mtns.; Middle Halawa Ridge.  Koolau Mtns.; Punaluu Valley/Castle Trail. Koolau Mtns.; Castle Trail.  Haleakala;  S of road.  West fork Kawela Gulch.  Kipahulu valley.  1000  E. Maui  EMKIP  L.  remyi  Location  Altitude (meters)  Island  Pop. Code  Species  Collection locations of populations of endemic Hawaiian Lysimachia continued. Table 2.1 Specific epithets are sensu Wagner et al. (1990). sampled for morphometric analysis.  Oahu Oahu Oahu Qahu Lanai Kauai Kauai Kauai Oahu Oahu Oahu Oahu Oahu Oahu  WAIHO* KOOLY* NUUAN* KULIO* LANAI* OLOKE* HANAP* KAHIL* PALAW* HAPAP* KALEN* KAALA* KANEH* WAIAW*  L. hillebrandii  L. hillebrandii  L. hillebrandii  L. hillebrandii  L. hillebrandii  filifolia  L. kalalauensis  L. glutinosa  L. hillebrandii  L. hillebraridii  L. hillebrandii  L. hillebrandii  L. hillebrandii  L. hillebrandii  L.  Island  Pop. Code  Species  800  830  770-1140  760-900  800  740  unknown  unknown  unknown  unknown  unknown  unknown  710  unknown  Altitude (meters)  Waianae Mtns.; Puu Kanehoa, near summit. Koolau Mtns.; Waiawa.  Waianae Mtns.; Mt.Kaala, N slope.  Waianae Mtns.; Puu Kalena.  Waianae Mtns.; Puu Hapapa.  Waianae Mtns.; S. Palawi Gulch.  Kahili Ridge.  Hanapepe Valley.  Olokele Valley.  Koolau Mtns.; Olympus-Waimanalo Pali. Koolau Mtns.; Lateral ridge in Nuuanu Valley. Koolau Mtns.; Puu-O-Kona (Crest of Kuliouou Valley). Mtns. at east end.  Koolau Mtns.; Waihole Gulch.  Location  Collection locations of populations of endemic Hawaiian Lysimachia continued. Table 2.1 Specific epithets are sensu Wagner et al. (1990). sampled for morphometric analysis.  25  L.  hillebrandii s.l.  on three leaves,  From each OTU, measurements were taken  (representative of the most frequent form)  and whenever possible,  one to four flowers.  measurements included leaf length widest point  (LLPUBE),  leaf width at the  petiole length (PETLM), upper leaf  (LEAFWM),  surface pubescence  (LEAFLM),  Vegetative  (ULPUBE),  lower leaf surface pubescence  stem pubescence (STEMPUBE), minimum internode length  (MINNTRND), maximum internode length  (MAXNTRND),  standard deviation of the internode lengths  and the  (STDNTRND).  This  last measurement assesses the evenness of distribution of leaves along a stem and is based on six adjacent internodes measured from the mid—portion of mature stems. measurements included calyx length (CALWIDM),  calyx pubescence  (CORLENM),  corolla width  style length pubescence  (STYLENM),  (PEDPUBE).  (CALLENM), calyx width  (CALPUBE),  (CORWIDM),  Floral  corolla length  filament length  pedicel length  (FILLENM),  (PEDLENM), and pedicel  The high density of hairs of the most  pubescent plants precluded an actual count.  A value of “one”  was assigned to glabrous surfaces and “five” to the most pubescent.  For each OTU an average of the above characters  was used in all further analyses.  2.2.2.1.  Analysis of metric data  The inconsistent occurrence of flowers and fruits indicated that more than one PCA would be needed to describe the data adequately. PCA.  Thus,  The first data set  three data sets were analyzed by  (1036 OTU’s)  included vegetative  26  characters of all specimens.  The specimens in the remaining  data sets are subsets of the specimens in the first. second data set flowers.  (405 OTU’s)  The  comprised specimens that bore  PCA of this data set used two different combinations  of characters:  1)  third data set  (636 OTU’s)  floral alone;  vegetative and floral.  2)  The  comprised OTU’s of taxa that were  not distinct from each other based on the ordination of OTU’s in data sets one and two.  By eliminating distinct taxa,  subtle relationships among the remaining taxa may be revealed. These belong to the complex, j.  .  filifolia  hillebraridii  .],.  and j.  remyi  lydgatei.  ••  PCA of this data  set used all vegetative and calyx characters. The statistical package SYSTAT  (Wilkinson,  for the principal components analyses  (PCA).  1990)  was used  Ordination of  the scores from the first two PCA axes was visualized using SYGRAPH (Wilkinson,  2.2.3  1990).  Non—metric observations A number of qualitative, non—metric characters differed  among populations of the j. complex.  hillebrandii  Observations were made of the following characters:  shape of leaf blade, veins pellucid or not, areoles,  remvi  leaf color,  calyx lobe size,  prominence of  color and shape, pedicel  color and position and presence of wax on leaves and stems. The angle of divergence between the primary vein and the secondary vein was measured on plants from 10—15 leaves of each group.  Most of these characters are non—continuous and  27  to include them in the PCA through arbitrarily assigned values would mean violating assumptions of the method. detract from their importance,  however.  This does not  In many cases,  characters such as color and shape of organs contribute more significantly to our perceptions of differences among species than do quantitative differences.  28  2.3  Results  2.3.1.  Principal Components Analysis  Using only vegetative characters, in Figure 2.1.  Lysimachia forbesii,  lesser degree, j.  filifolia  some groups are evident  i. glutinosa, and to a kalalauensis,  j.  maxima form somewhat distinct clusters. hillebrandii the plot.  .i. and  .  renwi s.1.  and j.  OTU’s of L.  are distributed throughout  The first component had high loadings for LEAFLM,  LEAFWN and PETLM,  the second component had high loadings for  ULPUBE, LLPUBE and STDNTRND  (Table 2.2).  The first two PCA  axes represented 66% of the variation. Non—flowering specimens were removed, two,  and the PCA was repeated.  floral characters alone  The ordination of OTU’s using  (Figure 2.2)  vegetative characters alone  leaving data set  differs from that using  (Figure 2.1)  in that OTU’s of the  same species form more distinct clusters in the former than in the latter.  Lysimachia daphnoides,  L.  glutinosa,  forbesii,  L.  and to a lesser extent L. maxima and some OTU’s of L. filifolia,  are distinct.  Again, there is considerable overlap  in the ordination of OTU’s of hillebrandii  ..],.  .  remyi .j.  and  L.  All characters except CALPUBE and PEDPUBE  had high loadings on the first PCA axis,  CALPUBE and PEDPUBE  had high loadings on the second PCA axis  (Table 2.3).  The  first two PCA axes represented 75.1% of the variation. The ordination of OTU’s based on PCA of both vegetative and floral characters of data set two  (Figure 2.3),  was  29  4 3  •  •  2  8  B  1 U  Mx K  H HHX 4X  •tK“Ø1H4  HH  N R  K  K  1<  K  FiMH11H  0  x  K  •  HNX  :R.  -  0  B  flI•It  8  MB 888  —  •  r B  K  :.x  B  K  K  B  H  B  —1  K —  HWI1fHHH H 1 pq  HH  G G%  -3  H  H.  G  -2  .  X..4(  G  H  Ba.  H  .  B  GG  a  B  -2  0  2  4  6  8  pci  Figure 2.1 Plot of principal components 1 and 2 using vegetative measurements. Each symbol represents a single OTU. Abbreviations: B=L. forbesii, D=L. daphnoides, F=L. filifolia s.i., G=L. glutinosa, H=. hillebrandii si., K=L. kalalauensis, R=L. remyi s.i., T=L. lydgatei, X=L. maxima.  30  Table 2.2. Principal components analysis of Hawaiian Lysimachia using only vegetative characters. Measurements taken from 1036 OTU’s. A plot of OTU’s in relation to the first two principal components is shown in Figure 2.1. Variable STEMPUBE ULPUBE LLPUBE LEAFLM LEAFWM PETLM MINNTRND MAXNTRND STDNTRND Eigenvalues % Variance Cum. Var.  Component Loadings 1 2  3  0.206 0.274 0.222 0.853 0.882 0.806 0.672 0.668 0.442  0.815 0.712 0.841 —0.335 —0.275 —0.249 —0.303 0.284 0.438  0.286 0.345 0.292 0.150 0.122 0.320 0.112 —0.670 —0.767  3.415 37.9 37.9  2.494 27.7 65.6  1.476 16.4 82.0  Table 2.3. Principal components analysis of Hawaiian Lysimachia using floral characters. Measurements taken from 405 OTU’s. A plot of OTU’s in relation to the first two principal components is shown in Figure 2.2. Variable  Component Loadings 1 2  CALPUBE PEDPUBE PEDLENM CORLENM CORWIDM CALLENM CALWIDM STYLENM FILLENM  0.109 -0.039 0.696 0.965 0.866 0.854 0.838 0.851 0.868  0.862 0.893 -0.126 -0.038 -0.077 0.116 -0.258 0.221 0.070  Eigenvalues % Variance Cum. Var.  5.087 56.2 56.2  1.697 18.9 75.1  31  4 •  H  •  B  .  3-.B •  B  x  •  2  F  R  :  ‘U  p  o  ‘H  x •  1  B’  H  F,  H  :  —  RHp H  0  F  B  HR  F  HR  1  :  B R  ‘R .R  R  )  H  H  HHH  hRRRRHR  HH  .  H  H  H  HHH  H  H.H H  HH H  R  RH  fr  H  H  H  H  H.  HH HH  H  H  —  —  R:H • •  8 H  H  B . .  ‘S .  S.  S S  .  S S  °  -2  B  R R  FF B  B.  :  X  .  .H RHRHR  R-H  H  —  :RR  0 H  BR  RB  —  .  •  0  U  .  -2  -1  0  1  2  3  4  pci  Figure 2.2. Plot of principal components 1 and 2 using floral measurements. Each symbol represents a single OTU. Abbreviations: B=L. forbesii, D=L. daphnoides, F=L. filifolia s.1., G=L. glutinosa, H=L. hillebrandii s.1., K=j. kalalauensis, R=. remyi s.1., T=. lydcratei, X=L. maxima.  32  4 3  fir  —  P  p  •  -  -  P P A  -  2  -  P  -  H •  aU-  ft H.j  PR  fl R 8 P  I •  :  -  H BB R  HH HH  -  -  Rp  •  0  :  a  -  H  0  oL o H -o -  P  tl_  0 K  K  B  -  -  -2  HHHH  •H  * pp;pr$ H  -2  H0 H  0  H  H :*;HH  F  —1  -  :H 8 p  RRPR •H P p PHR P HP 1 j P  -  -1  UHH  0  h 1  Ga  a%:a  a  a Ga  2  -  a  •  3  a  4  5  6  pci  Figure 2.3. Plot of principal components 1 and 2 using floral and vegetative measurements. Each symbol represents a single OTU. Abbreviations: B=L. forbesii, D=L. daphnoides, F=L. filifolia s.1., G=L. glutinosa, H=L. hillebrandii s.1., K=L. kalalauensis, R=L. remvi s.1., T=L. lydgatei, X=. maxima.  33  Table 2.4. Principal components analysis of Hawaiian Lysimachia using vegetative and floral characters. Measurements taken from 405 OTU’s. A plot of OTU’s in relation to the first two principal components is shown in Figure 2.3. Variable  Component Loadings 1 2  3  4  STEMPUBE ULPUBE LLPUBE CALPUBE PEDPUBE PEDLENM CQRLENM CORWIDM CALLENM CALWIDM STYLENM FILLENM LEAFLM LEAFWM PETLM MINNTRND MAXNTRND STDNTRND  0.094 0.209 0.096 0.132 —0.054 0.704 0.921 0.761 0.875 0.786 0.819 0.783 0.880 0.864 0.661 0.595 0.415 0.221  0.832 0.681 0.849 0.683 0.783 —0.070 —0.093 —0.092 0.040 —0.215 0.110 —0.013 —0.183 —0.046 —0.085 —0.162 0.387 0.467  0.051 0.119 0.058 0.205 0.341 —0.119 0.202 0.297 0.138 0.133 0.236 0.332 —0.172 —0.203 —0.169 —0.216 —0.770 —0.714  0.064 0.069 0.023 0.184 0.113 —0.194 —0.181 —0.396 0.077 —0.306 —0.407 —0.231 0.227 0.269 0.614 0.505 —0.214 —0.420  Eigenvalues % Variance % Cum. Var.  7.207 40.0 40.0  3.474 19.3 59.3  1.765 9.8 69.1  1.418 7.9 80.0  34  4 3  ,  I  I  I  I  H  H H..  H  —  -  -  H  •  H. :H  H  2  H H  :  H  H  H  H  H H  H  1  U  H  RH 141  .H  —  H  H  H  H  H  H  x H  H  R  H  x x  c:N 0  R  x  HU  RH RH RH  F  FI  P  H RH  P  P  PT  I’tK  —  x  TI  H I  r  -3  -  R I  -4  x  H  -1 -2  x x x  -2  -1  0  1  -  2  3  4  5  pci  Figure 2.4. Plot of principal components 1 and 2 using vegetative and calyx measurements. Each symbol represents a single OTU. filifolia Abbreviations: F=. .i., HL. hillebrandii s.1., R=L. remvi .]., X=i. maxima, T=. lydgatei.  35  Table 2.5. Principal components analysis of Hawaiian Lysimachia using vegetative and calyx characters. Measurements taken from 636 OTU’s. A plot of OTU’s in relation to the first two principal components is shown in Figure 2.4. Variable STEMPUBE ULPUBE LLPUBE CALPUBE CALLENN CALWIDM LEAFLM LEAFWM PETLM MINNTRND MAXNTRND STDNTRND Eigenvalues % Variance % Cum. Var.  Component Loadings 1 2  3  4  0.548 0.511 0.567 0.416 0.586 0.433 0.733 0.739 0.529 0.338 0.753 0.686  —0.649 —0.602 —0.661 —0.620 —0.103 0.164 0.431 0.418 0.301 0.472 0.299 0.168  0.038 0.159 0.089 0.322 —0.540 —0.496 0.088 0.251 0.650 0.470 —0.272 —0.443  0.095 0.071 0.048 —0.162 0.287 0.580 0.238 0.050 —0.018 0.181 —0.479 —0.540  4.101 34.2 34.2  2.435 20.3 54.5  1.661 13.8 68.3  1.076 9.0 77.3  36  similar to that using floral characters alone 2.2 and 2.3),  however,  L.  (compare Figures  daphnoides was not as distinct,  .  the first PCA axis there was less overlap between  ..],.  hillebrandii  and j..  remyi  than there was using  vegetative or floral characters alone. of OTU’s of  .  filifolia  There were two groups  evident in Figure 2.3 that did  not form as close a cluster as in Figure 2.2. distinct from all other OTU’s, remyi s•1. CALLENM,  STYLENM,  One group is  .  the second overlaps with  On the first PCA axis PEDLENM,  CALWIDM,  On  FILLENM,  CORLENM,  LEAFLM,  MINNTRND had high loadings; ULPUBE, LLPUBE,  LEAFWM,  CORWIDM, PETLM,  CALPUBE,  PEDPUBE had high loadings on the second PCA axis  and  and  (Table 2.4).  The first two PCA axes represented 59% of the variation. For data set three, kalalauensis,  OTU’s of L.  L. daphnoides and L.  glutinosa,  L.  forbesii were removed from  data set one because they were distinct from the other taxa in either quantitative vegetative or floral characters,  as well  as possessing unique suites of non—metric characters.  The PCA  was repeated and the resulting scores plotted in Figure 2.4. Many OTU’s of  .  other species. OTU’s of L.  lydgatei and L. maxima were distinct from the Although there was substantial overlap between  remyi s.1.  and L.  hillebrandii .i.,  different areas of the scatterplot. LEAFLM,  LEAFWM,  PETLM,  loadings; ULPUBE,  they occupy  On the first PCA axis  STDNTRND and MAXNTRND had high  LLPUBE,  CALPUBE and STEMPUBE had high  loadings on the second PCA axis  (Table 2.5).  represented by the first two PCA axes is 54%.  The variation  37  1  I  I  o  C  o 0  0  0  o  o 0 0  cN  0 -1  -  <  K  -2 -2.0  K  K K  K  K  K  I  I  I  -1.5  -10  -0.5  0.0  pci  Figure 2.5. Plot of principal components 1 and 2 using vegetative and calyx measurements of . filifolia Each symbol represents a single OTU. Abbreviations: K=OTU’s from Kauai, O=OTU’s from Oahu. OTU labelled “K” closest to those labelled “0”, came from Olokele Valley, the remaining “K” OTU’s came from the headwaters of the Wailua River.  38  The high density of OTU’s whose ordination is similar to that of L. 2.4.  filifolia s.1. obscures their position in Figure  When plotted alone however, two groups are clearly  evident (Figure 2.5).  2.3.2  Distribution of non—metric characters Qualitative morphological characters and selected  quantitative characters defined five groups of populations in  L. hillebrandii .],.  (Table 2.6), and three groups in  .i (Table 2.7).  In Figure 2.6, OTU’s of  and L. remyi s.1.  are plotted according to the group  designations from Tables 2.6 and 2.7. included with Group G of  . remvi  . hillebrandii .j.  Lysimachia lydgatei is  . remvi •.],• because the  geographical distribution and morphology of this species are very similar to those of OTU’s of  . remyi ••  Group E 45—55 12—24 light green elliptic yes no no no  35_550  5—8 green lanceolate 12—16 green erect no montane wet Oahu  Group D 50—65 16—24 light green elliptic no yes yes no  35_450  4—5 green ovate 11—13 green erect no lowland mesic Oahu  Group C 60—80 25—36 light green elliptic no no no no 35_400 4—5.5 dark red lanceolate 6—8.5 dark red pendulous no lowland wet ICauai  Group B 45—60 8—11 light green narrowly elliptic oblance. yes/no no no no 15—35° 5—8 green lanceolate 11.5—14 green erect no lowland mesic Molokai  55—65 8—11 dark green oblance. no no no yes 20—35° 4—5 green ovate 10—11 green erect yes lowland mesic Kauai  Leaf length (mm) Leaf width (mm) Upper leaf color  Leaves pubescent Tertiary veins pellucid Leaf areoles prominent Leaf apex recurved Angle between primary and secondary vein Calyx lobe length(mm) Calyx lobe color Calyx lobe shape Corolla lobe length(nuu) Pedicel color Pedicel position Stems pulverulent Habitat  *  Island  Population Group: A= NSPTR; B= MKAWE, MKOLE, MNAKA, MONIN, MWAIK, MKAUN; C= OLIMA, OWAIN; D= HPUKS, HPUKE, HKAPU, HPAHO, HWAIK, HPALI, HLUAL; E= MOANA, WAIHO, KOOLY, See Table 2.1 for NUUAN, KULIA, PALAW, WAHIA, WAIAW, PUNAL, HALAW, KALIB. population locations.  Leaf shape  Character  Group A*  Distribution of selected metric and non-metric characters among groups of Table 2.6. populations previously classified as L. hillebrandii  *  Island  (mm)  8-10 montane mesic Maui  Maui  Maui  8-13 montane mesic  5.5-7 green lanceolate alternate! whor led 10-11 montane wet  4-7 green lanceolate alternate  5—6 green lanceolate— alternate  20—30 4—6 light green linear— oblance. no no no no 20-30°  30—45 8—14 dark green ovate  20—55 2—17 dark green linear— elliptic no no no no 45-60° yes yes no no 25-30°  Group H  Group G  Group F*  Population Group: F=WMHPL, WMMN2, WMMN3, WMUKE, WMMPP, WNLIL, WMIJIS, WMHPS, WMHPY, WMHEL; G=EMKOT, ENKOB, ENKAL; H=WMNAK, WMBLG, EMHPA, EMKAW, EMLWA, EMKPR, EMKUI, See Table 2.1 for population locations. EMPAL.  Corolla lobe length Habitat  Tertiary veins pellucid Leaf margins revolute Leaf areoles prominent Leaf apex recurved Angle between upper secondary veins and primary vein Calyx lobe length (mm) Calyx lobe color Calyx lobe shape Phyllotaxy  Leaf shape  Leaf length (mm) Leaf width (mm) Upper leaf color  Character  Distribution of selected metric and non—metric characters among groups of Table 2.7. populations previously classified as L. remyi s.1.  0  41  A  2 V  1  -.  .I,•..•V•  5c -.•yo Qv  0  0  -1 r)  X0dD -  :  0  00  -  0  -  .  •  -4  -2  -1  .  .  1  0  0•  2  3  4  pci  Figure 2.6. Plot of principal components 1 and 2 using vegetative and calyx measurements of j. hillebrandii s.l• and L. remyi s.i. Each symbol represents a single OTU belonging to group designations for hillebrandii s.l. = Group A; 0 = = Group B; = Group C; V=Group D; ( Group E) or L. remy .i. ( 4= Group F; G= Group G; = Group H; X = Group G X Group F). .  42  Discussion  2.4.  2.4.1  Taxonomic concepts Categories recognized here as species are based entirely  upon morphological characters. highly fertile (Chapter 4),  Interspecific F 1 hybrids are  and crossability cannot be used as  criteria to delineate species.  Species are defined here as  interbreeding groups of populations that share the same unique suite of morphological characters.  These combinations of  characters are not found in any other species and are taken as evidence that each species is monophyletic and reproductively isolated from every other species.  Subspecies are considered  to consist of geographically discrete populations within the overall distribution of the species; they differ in vegetative,  2.4.2  but not floral characters.  Taxonomic Treatment The results from the multivariate analysis provided the  starting point for the taxonomic revision.  However, relying  on PCA alone to make taxonomic decisions can be misleading because qualitative characters may distinguish taxa that are not differentiated in an ordination.  Although PCA was used as  the first step, non—metric, morphological characters ulti  ultimately contributed more significantly to the ta  revision than did the results of PCA. Based on the PCA, al.  (1990)  the species as classified by Wagner et  fall into three categories.  The first category  includes those that are clearly distinct: j. glutinosa, L.  43  kalalauensis, L.  daphnoides, and L.  floral morphology (Figure 2.1)  (Figure 2.2)  forbesii. Differences in  and vegetative morphology  define these species relatively well.  In the  second category are L. maxima and to a lesser degree L. lydgatei, whose ordinations also form clusters that set them apart from other groups of OTU’s category are  In a third  . remvi .j., . hillebrandii .j and J..  filifolia s.i. ordination,  (Figure 2.4).  In spite of occupying common areas in an  . remyi  and j.. hillebrandii  have  qualitative features that separate these taxa into groups of populations,  each of which differs in a number of characters  from all other groups  (Tables 2.6 and 2.7); these groups  require taxonomic clarification.  In the case of .  filifolia  .]., a taxonomic revision is suggested because PCA separated this taxon into two clusters of OTU’s  2.4.2.1.  (Figure 2.5).  Taxonomic revision of L. filifolia s.i.  Wagner et al.  (1990)  acknowledged that the classification  as one species, of narrow leaved plants from the headwaters of the Wailua River, on Kauai,  and upper Olokele Valley (the type location)  and from Waiahole Gulch on Qahu as j..  filifolia  s.1., may have been a somewhat artificial grouping.  The  single OTU from Olokele clusters more closely with those from Waiahole River.  (Figure 2.5)  than it does with those from the Wailua  Leaves of OTU’s from the Wailua River are 2—4 mm wide  and pubescent, whereas leaves of plants from Waiahole Gulch and Olokele are 0.8-1.2 mm wide and glabrous.  St. John  44  (unpublished manuscript) viscid leaves.  noted that the type specimen had  This character would distinguish it from  plants at the other two locations of j. however,  filifolia  viscid leaves are not mentioned in the type  description nor have I been able to detect glands on my examination of the type specimen.  L. filifolia (sensu Wagner et al., 1990)  classification of would include L.  Thus a revised  filifolia C.N.  Valley and Waiahole Gulch,  Forbes & Lydgate,  and the new species  .  from Olokele pendens Marr  from the Wailua River.  Taxonomic revision of the L. hillebrandii  2.4.2.2.  •.],.  remyi  complex  From Tables 2.6 and 2.7 and Figure 2.6,  it is apparent  that there are suites of characters that separate populations of the L. groups.  hillebrandii .j.-L. remvi  ••  complex into eight  The differences among these groups are regarded here  to be sufficient for the recognition of six species and three subspecies within the complex.  .  hillebrandii  •]•—. remyi  In the following discussion, the characters used to  distinguish among these taxa are highlighted. Three groups of populations are distinguishable within L. remyi s.1.  The dimensions and shape of the calyx lobes,  corolla lobes,  and pedicels are variable within these groups  and some combinations of these characters are unique to one group.  However,  observations of plants in the greenhouse  indicate that based on these floral characters alone it is  45  often impossible to determine to which group a particular plant belongs.  There are however, vegetative characters that  are specific to each group,  and based on these, a plant from  any population on Maui can be assigned correctly to the appropriate group.  Because these groups differ consistently  in vegetative characters,  but not floral characters, they are  treated here as three subspecies of  . remvi.  Populations that belong to Group F are found only on West Maui.  Perhaps the most useful character that distinguishes  Group F from Groups G and H is the angle of divergence of the secondary veins.  In Group F and in L.  lydgatei, this angle  varies from 20_300 for the lower veins, but increases to 40— 600 for the upper veins, whereas the angle of divergence for leaves of Group G and H is the same for all veins and varies from 15_400.  Leaf size,  useful in many cases.  shape, pubescence and color are also  In populations of Groups G and H these  characters are much less variable. are seen in populations of Group F.  Two patterns of variation Plants from some  populations differ markedly from each other in having leaves that range from linear and glabrous to elliptic and densely tomentose, approaching what could be classified as j. lydgatei, while in other populations these characters are more uniform. of L.  In some populations most plants fit the description  lydgatei,  but also include plants that are nearly  glabrous and with narrower leaves that, therefore, would be classified as  . remyi ..],.  One interpretation of the pattern  of variation is that morphologically diverse populations have  46  resulted from hybridization between  .  lydgatei and a taxon  that is nearly glabrous, with narrow leaves and a small calyx. Supporting this interpretation is the fact that on leeward Maui summits narrow—leaved plants grow on the more windswept, sparsely vegetated aspects, while L.  lydgatei is found on the  adjacent, protected slopes under a low canopy forest.  This  suggests that ecological factors may contribute at least to a difference in phenotypic expression.  It is unclear however,  whether or not this is controlled by specific genes.  In the  “common garden” of the greenhouse, with an admittedly small sample size, progeny that were grown from the seed of plants with narrow leaves also have narrow leaves. of L.  However, progeny  lydgatei were morphologically variable and include  plants with nearly glabrous,  narrow leaves,  with broader tomentose leaves.  as well as plants  This observation,  and the fact  that most populations from West Maui are morphologically diverse,  argues in favor of treating these populations as  belonging to a single phenotypically diverse taxon, includes OTU’s previously classified as j. Populations of Group F and L.  that  lvdgatei.  lydgatei are here reclassified  as j. remvi subsp. renwi. The remaining populations of Groups G and H. these.  .  remyi  belong to  A number of vegetative characters separate  OTU’s of Group G have ovate leaves that are dark  green, with pellucid veins and revolute margins. Group H are linear to oblanceolate,  Leaves of  are light green and not  only are the veins not pellucid, they are often nearly  47  obscured by a thick cuticle.  Plants of populations that  belong to Group G fit the description of j. John,  caliginis St.  and are here reclassified as I. remyi subsp. caliginis  (St. John)  Marr comb. nov.  Plants of Group H fit the  description of j. kipahuluensis St. John and are here reclassified as i. remyi subsp. kipahuluensis  (St. John) Marr  comb. nov. Populations of Oahu,  and Molokai.  .  hillebrandii  are found on Kauai,  The five groups of populations in Table  2.6 can be broadly subdivided further into populations of OTU’s with narrowly elliptic to oblanceolate leaves 8-li mm wide  (Groups A and B)  wide (Groups C,  and those with elliptic leaves 12-36 mm  D and E).  In addition, the angle of  divergence between the primary and secondary veins in the leaves is 15—35° in Groups A and B, and E.  and 35—55° in Groups C, D,  Although Groups A and B share similar leaf shapes and  dimensions,  they differ in a number of features  (Table 2.6).  Plants of Group A are glabrous, have dark green leaves, often with the apex recurved, waxy deposits on the young leaves and shoots, a thick cuticle that nearly obscures the veins and ovate calyx lobes.  Plants of Group B are pubescent or  glabrate, have light green leaves,  lack waxy deposits, have  lanceolate calyx lobes, and visible veins.  Group A was  unknown prior to this study and is here designated as L. scopulensis Marr. remyi subsp.  Populations of Group B are segregated as j..  subherbacea  (St. John) Marr.  48  Although there is some overlap in the ordination of OTU’s of Group C,  D,  and E  (Figure 2.6),  there are combinations of  characters that are unique to each  (Table 2.6).  Leaves of  Group D are distinct from all other groups because the areoles are well defined based on prominent tertiary and quaternary veins.  In other taxa, the quaternary,  veins are obscured.  and often the tertiary,  Calyx lobes of Group D are broadly ovate,  whereas those of Groups C and E are lanceolate. of Group D are longer than those of Group C,  Corolla lobes  and often are  shorter than those of Group E.  Characters that distinguish  Group C are linear calyx lobes,  dark red pigmentation for the  entire length of the calyx and pedicel, pedicel.  and the pendulous  The calyx and pedicel of some OTU’S of several other  taxa are occasionally pigmented as well, full length.  but never for the  OTU’s of Group D have an ovate,  the pedicel is green and upright.  Leaves,  green calyx and  stems and calyx  lobes of OTU’s from Group E are moderately to densely pubescent and the calyx lobes are lanceolate to narrowly ovate.  By comparison,  glabrous.  leaves and stems of Groups C and D are  The location and description of OTU’s of Group C  fit that of the type specimen of Likewise,  .  ovoidea St.  OTU’s of Group D match the description and location  of the type specimen of j.. waianaeensis St. Group E fit the type description of ex A.  John.  Gray,  .  John.  OTU’s of  hillebrandii Hook.  f.  a name which is retained only for plants from the  Koolau and southern Waianae Mtns,  and one location on Molokai.  49  2.4.3.  Summary of Taxonomic Revision  A total of 16 species,  one with four subspecies are  recognized in the following key and species descriptions. distribution of species by island is as follows: Kauai, species  (Figure 2.7); Oahu,  three species subspecies  five species  The  12  (Figure 2.8); Molokai  (Figure 2.9); Maui one species and three  (Figure 2.10).  that island except for j.  All species on Kauai are endemic to filifolia and j. hillebrandii.  Three of the five species on Qahu are endemic to Oahu, with one occurring also on Molokai and the second with the same subspecies of  .  renwi as occurs on Molokai.  One of the three  species on Molokai is restricted to that island. subspecies of L.  remyi is restricted to West Maui,  One the other  two subspecies that are found on Maui occur on both East and West Maui. The taxonomic conclusions for most taxa are based on the results presented in this chapter with the following exceptions.  Lysimachia iniki Marr sp. nov. was not included  because material of this new taxon became available as the study was ending.  Lysimachia venosa was not included based on  the lack of specimens.  Lysimachia haupuensis St. John and L.  kahiliensis St. John, are treated here as distinct species but would have been classified as L. hillebrandii .1. of Wagner et al.  (1990).  by their type specimens.  in the key  These species are represented only  o it-’ o  •  )  ).J•  c:Q  --1t a) CD  CD H- I-I  CD  •-. CD  II  II  1%)  it-’  0 it-, •II  I—J I-’ Cl) H-• ctrt I—Il I—’ H H- < HCD 0 0(fl i—’O a) r1 H- Cl) —. H  D)O)  ••  0  II  z  iiit-I-t) a)Q CD  H-  CD  it-I  0o H-H  (Da)  il—-a’ it-I  HH  /  S  lIP)  -a  Cn  a’.  HCD I-P) C) H- I—a H CD a’ CD  dna) H-  0  A II Cfl dl) it-I  —.  H  ii C) P) t-’ H P) •  (fl0  CDC)Z ct  Cfl  H-  U) CD --CD .CnCn  iiH-I--  CnP’ z  it-I_.  ii 0  09  II  c) I.’. LQ  rII) 0 CD  t1o ; ‘1 i-i QCD rt (J) t,-,.  • II  IH  F-i-  f-i.  •  rt  -  II 0 z  P1  frt)  I—lCD CD  o  II- CD  I-.. Ui I-’• tn  p)  cn i-• I-.. 1) CD tlCn Ø)CD 00 CD I—.  rn  10 IIb Cn I-,. p.) ) I-,. (PP.) CD  tnz I-,. U)rt CD I— Ui  0  P.)  0 > C  0  tt)  19  Pu....  Point  I.  ii  •  •  LANAI  I II  KAHO 0 LA WE  10  MOLOKAl  MOLOKA’I  s.  CO.tOU* STIAW*LtOO6 1*11  LI  LA NA I  AND  MOLOKA’I  Distribution of endemic Hawaiian species of Lysimachia on the islands of Figure 2.9. remvi subsp. subherbacea; Lanai and Molokai: = . hillebrandii; = L. maxima; •= L. .= j. renwi subsp. remvi.  LANA’I  LI  PinI  (-TI  H101oionio.  Pont  WA it VA  MAUI  Koijiki Hood  KAHO ‘0 LA WE  AND  MAUI  Distribution of endemic Hawaiian species of Lysimachia on the island of Figure 2.10. Maui;•= L. remyi subsp. remyi;= . remyi subsp. kipahuluensis;•= . remyi subsp. caliginis.  P.powai  I.k.I&. Point  c-fl  54  2.5.  Key and descriptions of endemic hawaiian Lysimachia  Lvsimachia subgen.  Lysimachiopsis  (Heller)  Handel—Mazzetti,  Notes, Royal Botanic Garden,  Edinburgh 16:121-122,  Lysimachiopsis Heller, Minn.  Bot.  Studies 1:875,  Lysimachia section Fruticosae Knuth, 237  (Heft 22):  Huynh,  )  Hook.  Erect,  1970,  nom.  ex A.  Leaves entire,  5(6—10)—merous,  persistent,  urceolate, purple,  (as  sometimes  opposite, whorled,  or  Flowers hypogynous,  in terminal racemes,  corymbs,  or solitary in the leaf axils; calyx imbricate or  panicles, valvate,  hillebrandii  or prostrate perennial herbs,  alternate, usually glandular—dotted. actinomorphic,  (lectotype  Gray.  ascending,  subshrubs or shrubs.  IV.  Sandwicensia  illeg.-TYPE  Lysimachia hillebrandii var. f.  1897;  Das Pflanzenreich.  1905; Lysimachia subgen.  Candollea 25:288—289,  here chosen): var.  309—312,  1928.  deeply parted; corolla rotate to  deeply parted, tube very short, yellow, white,  reddish—purple,  or green,  the lobes contorted in bud;  staminal filaments slightly to nearly completely adnate to corolla,  often more or less basally connate; anthers basifixed  or versatile,  opening by apical pores or longitudinal slits;  ovary superior,  placentation free—central; capsule 5—10—valved  or irregularly dehiscent, oblong,  ovoid to globose; seeds numerous,  orbiculate or angular, testa crustaceous.  55  Key to endemic Hawaiian species. la.  Corolla green on upper portion,  dark red at the base,  petal margins erose; calyx lobes narrowly lanceolate, usually more than 12 mm long lb.  (2)  Corolla creamy white or reddish-purple, petal margins entire; calyx lobes ovate to linear, usually less than 12 mm long,  or if longer than 13 mm,  then usually at least 5  mm wide and ovate  2a.  Leaves  (3)  (80—)l60—200(—260)  mm long,  (20—)55—65(—95) mm  wide; upper and lower surface densely tomentose,  the  hairs multicellular; calyx and corolla lobes more than 20 mm long. 2b.  Leaves  Oahu  3. j.  (50—)60—80(—lOO)  mm long,  forbesii  (l5—)25—32(—45) mm  wide; upper and lower surface glabrous or glabrate, the hairs unicellular; calyx and corolla lobes less than 20 mm long.  3a.  Kauai  Corolla white,  9.  (15-)l9—24(—30)  light green above, (50—)80—l20(—l60)  Corolla red, lB(-20)  glabrous,  mm long; entire plant usually viscid, Kauai  4.  .  or base red and upper 5-7 mm white,  mm long;  leaves dull,  glabrous or pubescent,  crlutinosa (5-)6-  light or dark green,  (l5-)20-80(-lOO)  not viscid or if viscid, hirtellous  kalalauensis  leaves shiny,  slightly lighter below,  especially young shoots. 3b.  mm long;  j..  mm long; plants  then viscid-hirsute to viscid (4)  56  4a.  Leaves less than 2 mm wide, or if 2—4 mm wide,  then the  leaves narrowly lanceolate and the stems pendulous, hanging from cliffs 4b.  (5)  Leaves more than 2 mm wide; stems upright,  not hanging  from cliffs  5a.  (7)  Leaves 2—4 mm wide, narrowly lanceolate.  Kauai. 12.  5b.  L. pendens  Leaves less than 2 mm wide, usually 1 mm wide,  filiform. (6)  6a.  Plants pendulous, hanging from wet cliffs.  Kauai, Oahu.  16. 6.  Plants upright,  streamsides and bogs. 13b.  7a.  Leaves,  .  L.  filifolia  East Maui  remvi subsp. kipahuluensis  stems and pedicels densely viscid—hirsute to  viscid—hirtellous;  leaves sometimes glabrate but the  margins remaining viscid-hirsute 7b.  (8)  Leaves stems and pedicels glabrous or pubescent but not viscid  8a.  Leaves  (25-)35—38(-42)  (9)  mm wide,  broadly ovate to  orbicular, cupped upwards; secondary veins prominent; upper portion of corolla cream to white.  Kauai 7.  8b.  Leaves  (7—)lO—13(-19) mm wide,  .  iniki  oblanceolate to oblong;  57  secondary veins obscure; upper portion of corolla salmon pink to dark red.  9a.  Kauai  Calyx lobes lanceolate,  1.  13-16 mm long,  leaves obovate (50—)75-80(—l00) glabrous. 9b.  daphnoides  4—6 mm wide;  mm long; plants  Kauai  15.  Calyx lobes ovate to lanceolate, wide,  .  3-11 mm long,  .  venosa  1.5-4 mm  or if greater than 4 mm wide, then ovate;  leaves  various (l5-)20-60(-95) mm long; plants glabrous or pubescent  (10)  ba. Leaves whorled,  3—4 per node, more than 20 mm wide  obovate or elliptic; stems rusty tomentose, glabrate with age. lob.  Molokai  10. j. maxima  Leaves alternate, wide,  or if whorled, then less than 20 mm  shape various; stems glabrous or pubescent.  ha. Leaves elliptic, 16 mm wide,  ..  ovate or obovate and usually more than  apex usually rounded,  often abruptly  acuminate hib. Leaves linear,  (11)  (12) lanceolate or oblanceolate, and mostly  less than 16 mm wide,  if elliptic, then less than 16 mm  wide or densely rusty tomentose, apex attenuate, acuminate or sometimes acute  (14)  12a. Pedicel and calyx lobes dark red for entire length; pedicel pendulous; calyx lobes linear,  1.5—2 mm wide;  58  petioles  (6—)7.5—12(—15)  mm long.  Kauai . 11. L. ovoidea  12b.  Pedicel and calyx lobes entirely or mostly green; pedicel erect; calyx lobes lanceolate to ovate, wide; petioles  (2—)3—8(—15)  obscure); calyx lobes lanceolate, calyx lobes,  (13)  mm long  l3a. Leaves with poorly defined areoles  pedicel,  (tertiary veins (4-)5—8(—9) mm long;  stem and lower leaf surface Kauai,  densely dark brown pilose to tomentose.  Oahu,  6. L. hillebrandii  Molokai. 13b.  2—5 mm  Leaves with well defined areoles,  (tertiary and higher  order veins prominent); calyx lobes usually broadly ovate,  (3-)4—5(—8) mm long; plants nearly glabrous. L. waianaeensis  16.  Oahu  14a. Leaves with prominent glands when dried,  elliptic. Kauai. 5. L. haupuensis  l4b. Leaves without internal glands that are prominent when dried,  (15)  shape various  15a. Young leaves and stems pulverulent (powdery appearance due to fine wax crystals), otherwise glabrous;  leaves  dark green, the tips recurved; calyx lobes glabrous, broadly ovate. 15b.  Leaves and stems not pulverulent, pubescent;  14.  Kauai  .  scopulensis  slightly to densely  leaves light or dark green,  the tips flat;  59  calyx lobes glabrous or pubescent,  lanceolate,  occasionally ovate  (16)  16a. Angle of divergence of secondary leaf veins acute, nearly parallel to primary vein; long;  leaves narrowly obovate,  mm wide.  Kauai  calyx lobes 9-11 mm  35-50 mm long, 8.  L.  9-12(-15) kahiliensis  16b. Angle of divergence of secondary leaf veins more obtuse; calyx lobes usually less than 10 mm long, 10 mm,  if longer than  then the leaves more than 50 mm long,  various.  Oahu, Molokai, Maui  leaves 13.  .  remyi  60  1.  Lysimachia daphnoides  Isi.  285,  (A.  Gray)  Hillebr.,  Fl. Hawaiian  Lysimachia hillebrandli Hook.f. var.  1888.  daphnoides A. Gray, Proc. Amer. Acad. Arts and Sci.  5:329,  Lysimachiopsis daphnoides (A. Gray) Heller, Minn. Bot.  1862.  Studies 1:875, Mts., U.S.  1897.—TYPE:  Sandwich  (Hawaiian)  Islands, Kauai,  Exploring Expedition under Capt. Wilkes 1838-1842,  (holotype: US!).  Lysimachia longa St. John, Kauai, 2736  Phytologia 64:46,  bog at head of Wahiawa stream,  (holotype:  1987.-TYPE:  19 Oct.,  1895, Heller  BISH! ;isotypes: BISH! ,F! ,GH! ,MASS! ,MO! ,NY!,  P[3] ! ,US!)  Erect,  upright shrubs usually less than 1 m tall,  largely from the base,  rarely above;  branching  stems green when young  becoming dark reddish—brown, densely viscid—hirtellous when young becoming densely reddish—brown tomentose, glabrous.  Leaves alternate,  (0.5—)3—7(—11)  occasionally  mm apart,  sessile  or petioles less than 1 mm long; blades oblanceolate to oblong,  thickly coriaceous,  13(-21)  mm wide,  (20—)30—35(—52)  base obtuse,  (6—)l0—  apex acute, margins slightly  revolute; upper surface light green, hirtellous,  mm long,  usually viscid—  but sometimes glabrous, the hairs white in younger  leaves becoming brown,  lower surface lighter green than upper,  densely viscid—hirtellous; base of lamina, primary vein and sometimes secondary veins red; secondary veins often obscure. Flowers solitary in leaf axils,  6—8—merous,  cainpanulate;  61  pedicels  (20-)25-32(-l10) mm long,  brown viscid—hirtellous; the base, ovate  erect, densely reddish-  calyx lobes green,  sparsely viscid—hirtellous,  (6—)7.5-ll mm long,  (7—)ll—13(—16) mm wide; anthers 2—2.5 mm long;  filaments style  Capsules 7—10 mm long.  lanceolate to narrowly  (2—)3-4(—4.5)  dark maroon to salmon pink obovate,  sometimes red at  mm wide; corolla lobes  (l3-)15-l8(-21) mm long,  (7—)8.5—11(—14)  mm long,  (8-)1O.5-13(-l4) mm long.  Seeds dark brown,  irregularly shaped,  1-1.4 mm long.  Phenology.  Flowering Jan.-Nov.  Distribution and Habitat.  Kauai.  Montane wet sedgelands of  Alakai Swamp usually restricted to low hummocks that rise slightly higher than the surrounding bog vegetation,  1230-  1400m.  Panicum,  Growing with Oreobolus,  Metrosjderos, Dicranopteris,  Rhynchospora,  Cheirodendron, Melicope, Adenophorus, Vaccinium,  Carex,  Dubautia,  Cibotium,  Styphelia, Viola,  Lycopodium and Coprosma. Comments.  The diagnostic characters of this species include  the densely viscid—hirtellous leaves,  stems and pedicels.  These characters are also found in L.  iniki.  between these two species are discussed under  The differences .  iniki.  There is some confusion regarding the origin of the collection of this species from Wahiawa Bog.  Specimens  labelled Heller 2736 were collected on different dates and from different locations. specimens at P[3],  BISH,  According to their labels, F and US were collected on 19 Oct.  62  1895 and specimens at MASS and F were collected on 12 Aug., 1895 “in and near a bog at the head of the Wahiawa”. specimen at G was collected on 14 Aug. Hanapepe River,  near the falls”.  collected on 24 Aug. Rivers”.  Heller  1895,  A  “along the  A specimen at MO was  1895 “between Hanapepe and Wahiawa  (1897),  notes that these collections came  from 800 ra, which is higher than the 480 m elevation of Wahiawa Bog,  but somewhat lower than the Alakai Swamp.  A  recent inventory of Wahiawa Bog failed to find this species (Tim Flynn,  personal comm.).  Given these considerations,  it  seems most likely that Heller 2736 did not come from Wahiawa Bog and instead came from the eastern part of the Alakai Swamp,  or no longer occurs in Waiawa Bog due to habitat  alterations. In the Hawaiian language L.  daphnoides is identified as  “lehua makanoe” or “kolokolo kuahiwi” Representative Specimens ecamined. Swamp  Warshauer 3354  (GH),  Fay 322  Hillebrand s.n.  (BISH), Hobdy 153  Lorence 6355  (PTBG), Wawna 2122  Takeuchi  (PTBG),  (BISH), Selling 2905  (BISH), Van Royen 11701  Mann 504-Mann 522  (K).  (MO,PTBG), Davis 133  (F,GH,MO,NY,PTBG,W),  Takeuchi 92b  250,251,252,  s.n.  eastern Alakai  (BISH,F,NY), Fay and Bulmer 332  (GH),  1888).  (BISH); western Alakai Swamp, Forbes  888.K (BISH,US,W), Herbst 2175 92a  KAUAI:  (“Sincocks Bog”), Penman 10631  (BISH),  (Hillebrand,  (BISH), Marr  (UBC), Lonence 5700 (W), Wawra s .n.  (MO,PTBG),  (W),  Sinclair  63  2.  Lysimachia filifolia C.N.  Papers of Bernice P.  Bishop Museum,  Lysimachiopsis filifolia Deg.,  Forbes & Lydgate, 6(3):74—75,  & I.  Deg.  391,  1983.-  1912, Lydgate 2  BISH!).  Lysimachia waiaholeensis St.  (holotype:  0.  third edition, p.  TYPE: Kauai, upper Olokele Valley, Jan.  TYPE: Oahu,  1916.  Forbes & Lydgate)  (C.N.  Plants Hawaii Nat. Parks,  (holotype:  Occasional  Waiahole gulch,  John, 250 in,  Phytologia 64(1):50,  1987.—  26 July 1926, Degener 17666  NY!).  Lysimachia funkiae St. John,  Phytologia 64(1):44,  Oahu,  24 Jan.  Waiahole gulch,  250 m,  1987.-TYPE:  1984, Funk 211  (holotype:  BISH!).  Decumbent, long,  delicate shrubs,  branching profusely up to 60 cm  reddish—brown to green,  stems red,  nearly glabrous.  pilose,  eventually glabrate;  Leaves alternate,  (1—)3—5(—9)  mm apart, petioles 0.1 mm long; blades filiform,  coriaceous,  (12—)25-40(--50)  attenuate,  mm long,  apex attenuate, upper surface glabrous,  lower surface pilose when young, secondary veins obscure.  minutely pilose, linear,  glabrate,  base  dark green,  dark green;  Flowers solitary in leaf axils,  merous; pedicels l7—27(—32)  green,  0.5-1.2 mm wide,  mm long, pendulous,  5—7—  glabrous or  occasionally red toward calyx; calyx lobes  to lanceolate, occasionally red toward base  64  (4-)5(-6) mm long, red,  (1-)1.5-2(2.5) mm wide; corolla lobes dark  lighter at tips, widely obovate,  (5.5-)6-8(-lO) mm long,  4.5—6 mm wide; filaments (2.5-)4-5 mm long, anthers 1 mm long; style 3.5-4.5(—5) mm long. Seeds dark brown,  Phenology.  Capsules ovoid,  irregularly shaped,  3.5-5 mm long.  1—1.5 mm long.  Flowering in Jan. -July.  Distribution and habitat.  Kauai.  Known from a single  collection made in 1912 from the upper part of Olokele Valley. Qahu.  Growing in waterfall spray zones, hanging from wet  cliffs with Isachne, Selaginella.  Known from only three small sub—gulches of  Waiahole Gulch, Comments.  Eragrostis, Machaerina, Bidens, and  250 m.  As the name implies, a distinguishing character of  this species is the extremely filiform leaves.  Further  collections from the type location would be most useful to verify that plants from Oahu are the same in all regards as the type specimen.  Unfortunately no collections from the type  location have been made since the initial collection. land is privately owned and access is not available.  The The  habitat is not stated either on the label on Lydgate 2 or in the type description. label,  From the description on the specimen  “far mauka, Olokele Valley”  for “toward the mountain”)  (mauka is the Hawaiian word  it is likely that the collection  came from the upper part of Olokele Valley, quite possibly in a waterfall habitat similar to the Oahu population.  The  original description states that the plant is a shrub though  65  the habit is not specified. branching and pendulous. cm wide,  Plants from Oahu are finely  The stem of the type specimen is 0.4  thicker than that of the Oahu plants, which suggests  that it may have been more upright. manuscript)  states that  Wagner et al. river  (1990,  (L. pendens)  type is viscid.  p.  .  St. John  (unpublished  filifolia has viscid leaves.  1080)  state that the Oahu and Wailua  plants are sparsely puberulent, while the  I have been unable to detect viscid leaves in  any specimens of L. was included in L.  filifolia or on those of  .  pendens, which  filifolia sensu Wagner et al.  leaves are not mentioned in the type description.  Viscid St. John’s  comment is in reference to the type specimen from Olokele only.  If plants from Olokele were viscid and upright, these  characters would distinguish them from the Oahu plants. Representative Specimens examined.  the type specimen. 90—705  (UBC).  Known only from  OAHU: Waiahole gulch Obata 90—689,990-703,  (BISH), Penman 11149  791-Mann 799  KAUAI.  (PTBG), Marr 246,247,248, Mann  66  3.  Lysimachia forbesii Rock,  Lysimachia longisepala C.N. P.  Bishop Museum.,  Forbes,  4(3):222,  Lysimachia koolauensis C.N. P.  Bishop Museum.,  (Rock)  0.  Deg.  edition, p.  391,  non Forrest  1914.  (1908).  Occasional Papers,  Deg., Plants Hawaii Nat.  2300 ft.,  1914.  Bernice  Lysimachiopsis forbesli  1983.-TYPE: Oahu,  Mtns., wet forest, s.n.  1909,  12:361,  Occasional Papers, Bernice  Forbes,  6(1): 39,  & I.  Fedde Repert.  Parks,  Koolau Range,  Sept.,  1908,  third  Punaluu  C.N. Forbes & Rock  (lectotype here designated: BISH 576726!;  isolectotype:  MO 786053.!) Nomenclatural note.  Lysimachia koolauensis C.N.  Forbes,  though published in the same year as L. forbesii, was published later than L. forbesii Rock In the original description, collected in Sept.  John,  1933).  flowering specimens  1908 and fruiting specimens collected eight  months later are cited. lectotype,  (St.  The specimen chosen here as the  bears an immature flower and is labelled as a type  but bears no date.  This specimen matches the original  description in having three flowers in the leaf axils, calyx lobes exceed the corolla lobes corolla lobes exceed the calyx),  (in older flowers the  thus it may be the specimen  upon which the type description was based. is dated Sept. description,  1908,  elev.  2300 ft.,  The isolectotype  as stated in the original  but does not show floral features as well as the  one designated as the lectotype. “Sheet no.  and the  3”.  The lectotype bears a note  I have not seen sheets no.  1 or 2.  67  Sprawling woody shrubs, with stems up to 1.5 m long, usually unbranched;  stems dark red,  glabrous when older. apart, petioles  Leaves alternate,  (23—)31-45(-62)  broadly elliptic, (20—)55—65(-95)  pilose with red hairs,  chartaceous,  mm wide,  (3—)5—23(—40)  mm  mm long; blades narrowly to (8O—)160-200(—260) mm long,  base attenuate,  apex acuminate, upper  surface dark green, glandular punctate when young, glabrous,  becoming  lower surface much lighter,  pilose,  becoming  the hairs red,  multicellular, glandular punctate throughout; primary, secondary, surface.  and tertiary veins prominent, Flowers 1—5 in leaf axils,  pedicels l6-25(—32)  nun long,  mm long,  red veins,  (6—)7—8(—9)—merous;  densely tomentose,  calyx lobes green with red veins, 22(—27)  especially on lower  pilose,  lanceolate (18-)20-  (3—)4—5(—7) mm wide; corolla lobes green with  narrowly elliptic,  long,  (6.5—)7—1O mm wide;  long,  anthers 3 mm long;  persistent in fruit.  the margins erose 20—24(-28) mm  filaments style  (9—)9.5—l3.5(—16.5)  (l6.5-)l7-l9(-24)  Capsules lO.5-12(-15)  irregularly shaped 1.5-2.5 mm long.  Phenology.  Flowering in September.  Distribution and habitat.  Koolau Range,  Oahu.  (“Pig-God”)  600-710 m.  mm  mm long,  mm long.  dark brown,  along the Castle  pendulous;  Seeds  Known only from wet forest  Trail,  Punaluu Mtns.  in the  From the collection labels it is  unclear if all collections came from a single population. Some specimens were collected “near the top of the trail overlooking the Valley”  (presumably Punaluu).  Another  68  collection comes from a “small gulch at the head of Kaluanui”. The latter is the valley immediately west of Punaluu; the trail connects the two. Comment.  This species cannot be confused with any other  Hawaiian Lysimachia.  In all characters,  species of Lysimachia in the world.  this was the largest  Among Hawaiian species it  was unique in having multicellular hairs, punctate glands on the leaves,  and often more than one flower per axil.  extant species,  kalalauensis is almost certainly the  .  closest relative,  Of the  because these are the only species that have  a green corolla with erose margins. Lysimachia forbesii was last collected in 1934 and is presumably extinct.  During the present study unsuccessful  searches were made in the area of the type location, place this species was ever found.  the only  The ecology of the area  has been significantly altered by the invasion of introduced species,  yet many native species do persist.  Representative Specimens examined. Trail,  Rock 815  (P,BISH), Degener 17688  (GH,US), Rock s.n.  (GH),  Rock 12502  (MASS,MO,NY,US), Forbes s.n. and Cooke s.n. Hosaka 38  OAHU: Koolau Mtns.  (BISH), Rock 1031  (GH,MO,NY), Rock 8839  (BISH), Degener 17689  (F), Degener 17690  (NY), Rock 376  Castle  (NY), Forbes  (BISH), Swezey s.n.  (BISH), Forbes s.n  (BISH),  (BISH).  69  4.  Lysimachia glutinosa Rock,  304,  1910.  Lysimachiopsis glutinosa  Plants Hawaii Nat. Kauai,  Bull. Torrey Bot.  Parks,  third edition,  ridge west of Haleinanu,  (holotype:  BISH,  (Rock) p.  14-26 Feb.  0.  Club 37:297-  Deg.  391,  & I.  Deg.,  1983.—TYPE:  1909, Rock 1770  photo at NY! ;isotypes: BISH! ,NY! ,P! ,US! ,W!)  Lysirnachia fayi St. John,  Phytologia 64(1):44,  1987.—TYPE:  Kauai, Mt.  3 Dec.  (holotype:  Kahili,  550 In,  1975, Fay 502  BISH!).  Lysimachia olokeleensis St. 1987.-TYPE: Kauai, (holotype:  John,  Phytologia 64(1):47—48,  Olokele Valley, Jan.  1912, Lydgate 9  BISH!).  Woody shrubs up to 2.5 In tall,  branching primarily from the  base,  often pulverulent; stems green  the entire plant viscid,  to reddish—brown. petioles  Leaves alternate,  (1—)4.5—10.5(—12)  long; blades oblanceolate to  broadly obovate or elliptic, long,  (l6-)25—35(-49)  acuminate, glabrous,  coriaceous,  mm wide,  light green,  shiny,  (23-)33-47(-51)  erect; calyx lobes green,  glabrous,  lanceolate to ovate,  (8—)10—12(—16)  apex acute to  upper surface  Flowers solitary in leaf axils,  campanulate; pedicels  mm  lower surface slightly lighter  glabrous,  nerves,  (50-)80—120(—160)  base attenuate,  sometimes abruptly acuminate,  than above. merous,  (l-)3-13(-35) nun apart,  mm long,  (5—)6—7(—8)— mm long,  the margins hyaline,  often with prominent branching (4.5—)5.5—7(—8)  mm wide;  70  corolla lobes white or cream, base,  obovate,  (15—)19—24(—30) mm long,  wide; filaments  Capsules  (6-)8-9(-10.5) mm long, white,  (9.5—)10-11(-12) mm long.  irregularly shaped,  (2—)2.5green or  Seeds dark brown,  1.2—1.7 mm long.  Flowering January-July.  Phenology.  Distribution and Habitat.  Kokee area, Ridge,  (9—)11—14(—17) mm  (6-)8-l1(-12) mm long, white, anthers  4(—5.5) mm long; style red.  sometimes reddish—purple at  1090—1290 m.  SE Kauai  for the species.  (Fay 502),  Mostly restricted to the  Kauai.  A single collection from Kahili 550 m is an unusually low elevation  Exact location of Olokele collection is  unknown, but is also outside the known principal range of the species.  Growing in Lowland Wet Forest dominated by  Metrosideros, Dianella, Dicranopteris, Perrottetia, Psychotria, Myrsine, Myrsine,  Scaevola, Coprosma,  Styphelia, Nestecris,  Ilex and Cheirodendron.  Comments.  Lysimachia.  This is the only white flowered species of Hawaiian It is also distinct in having sessile glands that  are responsible for producing a viscid surface.  “Viscidness”  may be a recessive trait as hybrids between this species and non—viscid species,  are not viscid.  A single putative hybrid (Marr 615) Kalalau Lookout,  collected below  is morphologically intermediate between this  species and j.. kalalauensis.  This individual closely  resembles artificial F 1 hybrids between these two species. Plants that appear to be hybrids between this species and j..  71  scopulensjs have also been collected below Kalalau lookout (Wood 1712,  1421)  and near Puu Ku  (Wood 2396,  Representative Specimens examined. Degener 30208,33496 21561  KAUAI:  Carlson 3713  (BISH),  (BISH,MASS,PTBG), Herbst 1001  (BISH),  Takeuchi 2473  Takeuchi Alakai 91a,91b Henrickson 4036 1616,3765  (BISH,RSA,US),  (BISH),  Lorence 6297  (BISH),  Plews 129  (PTBG),  Marr 616—Marr 619,  Degener 22334,  (BISH,US), Fosberg 41467 (BISH,G,RSA),  Cariquist 1986 Lyon 5019L  (PTBG),  Lorence 5126  Flynn 2769  Kalalau Lookout,  (F),  Wagner 5005  (BISH),  2397).  (BISH),  Lorence 6319  (PTBG),  (BISH,US),  (PTBG),  (BISH),  Marr 559-Marr 576  Herbst 2154  (GH,K,PTBG,US), 6532  (NY);  Kilohana,  (BISH,MASS,NY),  St.  Marr 254 (UBC),  (BISH),  Kilohana and Puu 0 Kila,  (F,NY),  (BISH);  Marr 255,  Ridge between Puu 0 Kila and Pihea, Carlquist 1317 Herbst 2055 1014,1715  (RSA,US),  Marr 614, (BISH);  Spence 174  Honopu Trail, (PTBG),  Gagne 546  Darwin 1126  Herbst 2374  Yuncker 3494 John 19985  (UBC);  Flynn 55  (F,MO,NY,PTBG),  Hobdy 69  (MO,PTBG),  Stern 2998  (BISH,F,NY); Kaluapuhi Tr.  Marr 260,  Stone  (BISH,MO,PTBG),  Kalahu Degener 21465 Flynn 160  (RSA),  (BISH),  Degener 22334  along road between  Marr 590-Marr 609 Lamoureux 2846  Carlquist 1793  (RSA);  Lorence 5798  road Degener 23949 (NY,P,US,W),  (W);  Kahuamaa Flat  Wilder 446  (BISH),  Neill s.n.  (MO),  (BISH),  Shear s.n.  Rock s.n.  Gustafson  (PTBG); Lehuaxnakarioi Trail,  Kokee area,  MacDaniels 810  (UBC);  (BISH,F),  (F); Road opposite Awaawapuhi trailhead,  (RSA),  Sohmer  (US),  (K).  not specific, Stern 2998  Forbes 786.K  (BISH,RSA),  Degener 17667  near  (K,M0),  72  5.  Lysimachia haupuensis St. John, Phytologia 64(1):45,  1987.—TYPE: Kauai, Haupu Range, Feb.  1927, MacDaniels 883  along base of cliff,  (holotype:  400 m,  26  BISH!).  Low branching shrub at least 50 cm tall; stems dark brown, pilose when young.  Leaves alternate,  1—10 mm apart, petioles  2 mm long; blades elliptic, coriaceous, glandular, 45(—50)  mm long,  (25-)36-  (7-)9-13(-14) mm wide, base acute, apex  acute, upper surface dark green,  shiny, glabrous,  lower  surface lighter than above, brown pilose becoming glabrous; secondary veins prominent, solitary in leaf axils, pilose,  tertiary veins obscure.  6—merous; pedicels 25 mm long, densely  erect; calyx lobes lanceolate, pilose,  wide; corolla lobes red, filaments 4 mm long,  Flowers  obovate,  10 mm long,  6 mm long,  2 ram  5 mm wide;  anthers 2 mm long; style 7 mm long.  Capsules not seen.  Distribution.  Possibly extinct.  Kauai.  Known from only the  single sheet of the type collection made in 1927, Haupu Range. Comments.  This species is difficult to classify due to the  limited collections.  A character that it shares with only j.  scopulensis, also from Kauai, internal foliar glands.  are the prominent  The leaves of L.  (upon drying)  scopulensis are much  lighter green and are linear, or narrowly lanceolate to obovate.  Lysimachia haupuensis also differs from j.  scopulensis in having longer,  lanceolate shaped calyx lobes  and lacking pulverulent leaves and stems.  Although j.  73  haupuensis was collected probably within 1-2 km of kahiliensis, made,  .  another species for which a single collection was  it differs from the latter in a number of regards  including differences in leaf shape,  leaf venation and calyx  size, thus it would be inappropriate to combine these two taxa,  although they do have similar leaf sizes.  The leaf  shape of j. haupuensis most closely resembles that of collections of L. hillebrandii from the Koolau Mtns on Oahu, yet internal glands are not prominent in dried specimens of  .  hillebrandii and leaves of this species have a thicker texture and are much lighter green than those of  .  haupuensis.  74  6.  Lysimachia hillebrandii Hook.  Acad. Arts and Sci.  hillebrandii  (as var.oL)  Arts and Sci. f.  ex A.  Deg.,  5:329,  5:329,  ex A.  Gray,  Proc. Am.  1862; Lysimachia hillebrandii var.  Hook.  f.  ex A.  Gray,  Proc. Am. Acad.  1862; Lysirnachiopsis hillebrandii  Heller.; Lysimachiopsis grayi 0.  Gray)  Plants Hawaii Nat.  Kalihi, Hillebrand 183 isolectotype:  f.  Parks P.  391,  Deg.  (Hook.  & I.  1983.—TYPE: Oahu,  (lectotype here designated: K!;  F!).  Lysimachia macclanielsii St. John, Phytologia 64(1):46, TYPE: Nov.  Oahu, 1926,  Konahuanui-Olympus Trail,  MacDaniels 89  (holotype:  edge of Pali,  6 Feb.  1793,  1912, Forbes 1747.0  Minn.  Bot Stud.  John,  Pacific Trop.  nom.  1:876,  illegit. 1897.  Bot.  Nuuanu, Hillebrand s.n.  (holotype:  Gard., Mem.  Lysimachia rubrimaculata St. John, TYPE:  Oahu,  April 1909,  BISH!).  284,  1888,  Lysimachiopsis ovata Heller,  Lysimachia ovata  (holotype:  1  1987.-TYPE:  Lysiniachia rotundifolia Hillebr., Fl. Hawaiian Isi. non Schmidt,  710 ra,  BISH;isotype: BISH!).  Lysimachia mannhl St. John, Phytologia 64(1):46, Oahu, Waiahole,  1987.-  1:270,  (Heller)  St.  1973.—TYPE: Oahu,  fragment of B at BISH!).  Phytologia 64(1):48,  1987.—  Koolau Mtns., Moanalua Valley Pali, Forbes s.n., (holotype:  BISH!).  6  75  Lysimachia russii St. John, Oahu, Waiawa  (top of Koolau),  isotype:  BISH;  Phytologia 64(1):49, Feb.  Oahu,  April 1909,  1930, Russ s.n.  Phytologia 64(1): 49,  (holotype:  (holotype:  .  —  2  BISH!;isotype: Mo!).  Phytologia 64(1):50,  TYPE: Molokai, Waiehu, Wailau Valley,  Sept.  1987.-  1912, Forbes  BISH! ; isotypes: Mo! ,P! ,W!).  Lysimachia websteri St.  John,  Oahu,  28 Mar.  S.  1987  Koolau Mtns., Kalihi Valley, Forbes 1255.0,  Lysimachia waiehuensis St. John,  559 .Mo  (holotype:  BISH!).  Lysirnachia stenophylla St. John, TYPE:  1987.-TYPE:  Palawai gulch,  Phytologia 64(1):50, 1948, Webster 1458  1987.-TYPE: (holotype:  BISH!).  Nomenclatural Note. Gray  (1862)  This species was originally published in  as one of three varieties of Lysimachia  hillebrandii Hook.  f.  ex A.  Gray.  This publication provided  descriptions of specimens collected during the U.S. Pacific Exploring Expedition of 1840. this variety of “L.  .  South  The descriptions for  hillebrandii were provided by J.D. Hooker  hillebrandii Hook.  f.  in litt.-Oahu and Maui”.  The  specific epithet suggests that a collection made by Hillebrand was the basis for the description of Hooker. however,  did not arrive in Hawaii until 1851.  Hillebrand, It is unclear  whether or not Hooker based his description on specimens  (US  76  2983211,  NY s.n.)  from the 1840 collection, or upon a specimen  sent to him later by Hillebrand. specimens cited for of Oahu, (1905)  In Hillebrand (1888),  var. are restricted to the Koolau Mtns.  “bare mountain ridges of Kalihi and Manoa.”  cited Hillebrand s.n. and Wawra 2211 and 2380,  restricted var.  typica R. Knuth,  to Oahu.  Knuth and also  However, Wawra  2211 is from the Waianae Mtns and is L. waianaeensis St. John. The specimen chosen here, Hillebrand 183,  as the lectotype was  collected from Kalihi and sent to Kew in 1865 St. John, unpublished manuscript) hillebrandii.  (according to  and was identified as L.  This sheet bears four flowering branches and  fits the description of L. hillebrandii var. C.of Gray (1862). It is here designated as the lectotype of  .  hillebrandii.  Woody shrubs up to 2 m tall; stems dark brown, densely reddish—brown tomentose at the tip, alternate,  (0.5—)3—24(—36)  glabrate.  apart, petioles  Leaves  (2—)3—7(—13) mm  long; blades narrowly to broadly elliptic or obovate, coriaceous,  (24—)36—55(—60) mm long,  (6—)12—24(—40) mm wide,  base attenuate, apex acute to rounded, often abruptly acuminate, base,  upper surface usually light green, pilose, red at  lower surface usually slightly lighter than above,  pilose to rusty tomentose, and petiole.  Flowers solitary in leaf axils,  merous; pedicels tomentose,  especially along the primary vein (6-)7—8(-9)-  (lO-)11-24(-30) mm long, densely rusty  erect; calyx lobes green,  lanceolate to narrowly ovate,  often red toward base,  (4—)5-8(-9) mm long,  (l.5-)2-  77  3.5(-5)  mm wide; corolla lobes red,  (5—)6—8(—9) mm wide; 6—8(-l2) brown,  mm long.  irregular,  Phenology.  filaments  Capsules  (4—)5—7.5(—8.5)  (6-)8-lO mm long.  mm long; style Seeds dark  1.5—2.5 mm long.  Oahu.  Koolau Mtns, wet forest at 400-710 m.  Southern Waianae Mtns.  in lowland mesic forest and cliffs 720—  growing with Dubautia,  Elaphoglossum, Peperomia, Hibiscus,  (9—)12—l6 mm long,  Flowering Sept.-Feb.  Distribution.  820 m,  obovate  Bidens,  Carex,  Hedyotis,  Styphelia,  Lysimachia waianaeensis,  Plantago,  and Silene.  Known from a single 1912 collection from Wailau, location not given.  Kauai.  Eragrostis,  Molokal. exact  Single collection, with the year  and location not given. Corninents.  Lysimachia hillebrandii resembles L. waianaeensis  in having obovate,  ovate or elliptic leaves, however the  areoles are not prominent.  It is sometimes difficult to  observe this feature from herbarium specimens. a comment on the label of a specimen April,  1932 from near Mt.  Punaluu  (Degener 4138)  collected  Kaala in the Waianae Mtns.:  leaves are “barely coriaceous, the Punaluu plants”.  Useful here is  the  thinner with showier veins than  Degener had collected specimens from  (Koolau Mtns., Oahu)  earlier that year.  Lysimachia  waianaeensis has stems, pedicels,  calyx lobes and leaves that  are glabrous, whereas those of j.  hillebrandii are pubescent;  calyx lobes of J. waianaeensis are narrowly ovate,  4—5 mm  78  long, while those of L. hillebrandii are lanceolate to broadly ovate,  5—8 mm long.  Leaf width is extremely variable in this species. leaves of most specimens are 12—40 mm wide, collections  a few scattered  (Forbes 1255.0, Forbes s.n., Degener 17669, St.  John 13004, Faurie 707) leaves,  While  6—14 mm wide.  from the Koolau Mtns. have narrower The distribution of these specimens  overlaps somewhat with that of the broader—leaved specimens and in fact,  some sheets  (Mann and Brigham 229,  BISH,  F)  bear  narrow leafed branches as well as broader leafed ones. Specimens from Oahu have leaves less than 32 mm wide,  however,  those from Molokai are up to 40 mm wide. The most recent collection of this species from the Koolau Mtns. was a 1980 collection  (Obata 434),  population evidently of a single individual. L.  from a  Prior to 1980,  hillebrandji had not been collected from the Koolau Mtns.  since 1937.  More recently collections have been made in the  Waianae Mtns.  but flowering specimens have never been  collected from there. Lysimachia hillebrandii is identified as “puahekili” in the Hawaiian language  (Hillebrand,  1888).  Representative Specimens examined.  O1HU:  Pali, Rock s.n. Trail),  (BISH),  Degener 17681  Degener 17683,  33503  Garber 249  (BISH);  Olympus—Waimanalo Punaluu  (GH,NY,US), Degener 17685 (NY); Suehiro s.n.  (Pig God  (GH,MO,NY),  (BISH); Nuuanu,  opposite King’s Falls, Mann and Brigham 229  (BISH[2],F[2],GH,  79  MASS,MO,NY[2j); Puu-O—Kona  (Crest of Kuliouou)  Obata 434  (BISH); Kalihi Valley, top of ridge, Swezey s.n. Christopherson 1270  (BISH), Forbes 229.0  (K), Hillebrand s.n. Wood 1812 622 1293  (BISH), Faurie 707  (PTBG), Penman 5404  (GH),  (BISH), Forbes 2291.0 (P); NE of Palikea  (PTBG); Palawai Gulch Wilbur  (BISH); ridge between Nanakuli and Lualualei Mann 1292, (UBC), Wood 1976, 2917  (PTBG), Wood 1977,  (PTBG,UBC); Middle Halawa Ridge, Degener 17669 Kalauao-Waimalu Ridge, St. John 13004 Palikea, Wood 1817  Koolau Mtns.), Wawna 2380 Expedition,  1838-1842,  Valley, Forbes 559 .Mo (BISH 586775).  (W),  s.n.  U.S.  (probably  South Pacific Exploring  (NY, US).  (BISH,MO, P, W).  (GH,NY);  (BISH); drainage NE of  location not stated  (PTBG);  1982  MOLOKAI: Waiehu, Wailau KAUAI: Lydgate s .n.  80  7.  Lysimachia iniki Marr sp. nov., TYPE: KAUAI: Headwaters  of N.  fork of Wailua River,  (holotype:  720 m,  PTBG!; isotype: BISH!).  30 Oct.  1992, Lorence 7270  Figure 2.11.  Lysimachiae daphnoidi affinis a qua praecipue differt apice albo corollae minoris; foliis latioribus, plus orbiculatis, sursum versis, viridibus, versum basem, non rubris.  Small woody shrubs with pendulous branches 30-150 cm long; stems green, densely hirsute. apart,  Leaves alternate, 2—5(—10) mm  sessile; blades obovate to orbicular,  coriaceous,  chartaceous when dry,  cupped upward,  (35—)37—45(—54)  (25—)35-38(—42) mm wide, base cordate,  nun long,  apex acute to rounded,  abruptly acuminate, upper and lower surfaces light green, translucent viscid—hirtellous throughout, becoming glabrate, scattered patches of red pigment on lower surface; veins pellucid.  Flowers solitary in leaf axils,  6—7—merous,  funneliform; pedicels 15-25 mm long, densely viscid hirtellous,  erect; calyx lobes green,  densely translucent viscid—hirtellous, long,  the margins hyaline, lanceolate, 8—10 mm  2-3 mm wide; corolla lobes with the upper 5-7 mm white,  the lower portion dark red,  the inner surface densely  glandular-punctate, oblanceolate, filaments 10 mm long, dark red, nun long. shaped,  Capsules 6—7 mm long. 1 nun long.  15-16 mm long,  5 nun wide;  anthers 2.5 mm long; style 8-9 Seeds dark brown,  irregularly  81  Phenology.  Flowering time unknown,  Distribution and habitat.  fruiting in Oct.  Known only from the wet cliffs of  the headwaters of the North fork of the Wailua River, and above.  in  At least two populations of more than 25  individuals are known.  Growing on vertical wet, mossy or  rocky cliffs with Machaerina, Hedyotis,  720  Pipturus,  Isachne,  Cyrtandra,  Bidens, Plantago,  Dubautia, Athyrium,  and  Metrosideros. Comments.  The viscid—hirtellous leaves,  calyx lobes of  pedicel,  pedicels and  i. iniki resemble L. daphnoides, from which it  differs in having broader, pellucid veins,  stems,  chartaceous  smaller fruit,  (when dry)  a shorter calyx,  leaves,  a shorter  and narrower corolla lobes that are distally white.  Named after Hurricane Iniki, which struck Kauai on Sept. 11,  1992.  The force of the hurricane broke off several  branches from the cliffs above the headwaters of the Wailua River in the same area where j. pendens grows. discovered by Lorence et al.  and seeds  These were  (from Lorence 7270)  sent to UBC where the plant was successfully grown and first seen in flower. or piercing,  In the Hawaiian language “iniki” means “sharp  as wind or pangs of love”  (Pukui and Elbert,  1992) Representative Specimens examined. fork of Wailua River, Flynn 5276  KAUAI:  (PTBG),  headwaters of N  Penman 13079  (PTBG).  82  10 mm  Figure 2.11. Lysimachia iniki. branch; C. leaf.  A.  flower; B.  flowering  83  8.  Lvsimachia kahiliensis St.  John,  1987.—TYPE: Kauai, Kahili Ridge, Aug. (holotype:  glabrous.  Forbes 271.K  at least 40 cm tall; stems reddish-brown,  Leaves alternate,  1—10 mm apart, petioles 1—2 mm  long; blades narrowly obovate, mm wide,  glabrous,  1909,  BISH!;isotype P!,US!)  Upright shrub,  12(—l5)  Phytologia 64(1):44,  coriaceous,  base attenuate,  35—50 mm long,  apex attenuate,  9—  upper surface  lower surface glabrous; angle of divergence of the  prominent secondary veins narrowly acute from the primary vein,  tertiary veins obscure.  Flowers solitary in leaf axils,  6—merous; pedicels 6—17 mm long; calyx lobes green, lanceolate,  9—11 nun long,  Capsules 6 mm long.  3 mm wide; corolla unknown.  Seeds,  dark brown,  irregularly shaped  1.5-3 nun long.  Phenol ogy.  Unknown.  Distribution.  Kauai.  from Kahili Ridge, Comment.  Possibly extinct.  exact location unknown.  This is a problematic species because it is known  from a single collection. base,  Collected only once  resembles L.  The leaf, with a nearly sessile  daphnoides.  Lysimachia kahiliensis differs  from the latter in that the secondary veins depart from the primary vein in a more acute angle, coriaceous, abruptly,  the leaves are much less  the apex is more acuminate,  the base tapers more  and the entire plant is nearly glabrous.  84  9.  Lysixnachia kalalauensis Skottsb.,  Meddel.  15:429—430,  Heller, Minn.  Bot.  Lysimachiopsis hillebrandil sensu  1944.  Stud. Bull.  Lysimachiopsis hillebrandii  Lysiinachiopsis helleri Hawaii Nat.  Parks,  9, May 1897,  (Hook.  (Knuth)  f.  third edition,  p.  & I.  1905.  Deg.,  Aug.  1938,  and I. Deg.,  392,  Plants  1983; Lysimachia  Pflanzenr.  Parks,  non  Heller.  IV.  237(Heft  Lysimachiopsis kalalauensis (Skottsb.)  Plants Hawaii Nat.  1983.-TYPE:  Plate LVIII,  ex. A.Gray)  0. Deg.  hillebrandii var. helleri R. Knuth, 22):310,  Goteborgs Bot. Tradg.,  third edition,  p.  0.  391,  Kauai, Kilohana Lookout above Kalalau Valley,  Cranwell, Selling & Skottsberg 3034  (BISH).  The description of the type location is in error.  Deg.  20  (Note:  Kilohana  Lookout is above Hanalei valley, Kalalau Lookout is above Kalalau Valley.  The habitat at Kalalau Lookout is more  typical for the species, there,  and there is an extant population  therefore it is much more likely that the type  collection was from this location rather than from Kilohana Lookout).  Lysimachia hanapepeensis St. John, Phytologia 64(1):44, TYPE:  Kauai,  2614A  (holotype: K; isotypes: A! ,F!  ridge west of Hanapepe,  Lysimachia lamiatilis St. TYPE:  John,  ,  1987.-  23 July 1895, Heller  GH! ,M0! ,NY!  ,  P! ,UC! ,US).  Phytologia 64(l):46,  Kauai, Wahiawa Marsh, Lydgate s.n.  1987.  (holotype: BISH!).  85  Shrub with sterns up to 4 m long, branching mostly from the base, with short lateral shoots; stems brown to dark red, densely rusty tomentose when younger, becoming glabrous. Leaves alternate, 8.5(—lO.5)  mm long; blades elliptic,  (—100) mm long, rounded, green,  (1-)3-23(-40) mm apart, petioles coriaceous,  (5—)7.5-  (50—)60—80  (15—)25-32(-45) mm wide, base attenuate to  apex acute to abruptly acuminate, upper surface dark  glabrous,  lower surface much lighter,  lightly pilose,  with scattered reddish-purple streaks; primary vein and secondary veins red,  secondary veins prominent, higher order  veins often obscured by thick cuticle. leaf axils,  Flowers solitary in  (5—)6—7(-8)—merous, urceolate, the petals often  tightly closed around the exserted style, abscission; pedicels  (12—)16-33 mm long,  even until corolla lightly pubescent,  often pendulous, rarely erect; calyx lobes, green,  spotted  with streaks of red, the margins slightly hyaline, nerves sometimes visible, (—5)  linear (7—)13—16(—20) mm long,  (2—)3—3.5  mm wide; corolla lobes obovate, the margins erose, red at  base becoming green in upper half,  but the veins red and inner  surface of corolla remaining red further distally than on the outer surface (13—)15—17 mm long,  (6.5—)7—9(—9.5)  filaments 10-12 mm long, red, anthers  mm wide;  (2.5-)3-5 mm long,  sometimes elongating beyond the corolla lobes and clasping the style; style 10-12 mm long, red at base, persistent in fruit. Capsules broadly ovoid to subglobose (7.5-)8-l0(-ll) mm long. Seeds dark brown,  irregularly shaped,  2—3.1 mm long.  86  Phenology.  Flowering Feb.- Nov.  Distribution.  Kauai.  Scattered populations in Lowland  Diverse Mesic Forest of western and central Kauai, Growing with Metrosideros, Acacia, Tetraplasandra, Styphelia,  Alyxia,  Psychotria,  Zanthoxylum,  Melicope,  Pouteria, Wikstroemia,  Vaccinium, Antidesma, Wilkesia, Remya,  Alphitonia,  Dianella,  970—1260 m.  Pleomele, Hedyotis,  Scaevola,  Dodonaea, Dubautia  and Nestigis. Comments.  A number of characters distinguish this species  from all others.  This is the only extant species in which the  corolla lobes are distally green with red veins and erose margins.  The corolla lobes frequently remain tightly closed  around the exerted style; possible exception of reflexed.  .  in all other species, with the ovoidea, the corolla lobes are always  Leaves of seedlings and sometimes young leaves of  older plants often have a silver color.  The short lateral  shoots is a feature also common in L. waianaeensis. Representative Specimens examined.  KAUAI:  ICalalau Lookout  Stone 3766, (K,BISH); Makaha Valley road, Flynn 1157, (BISH,PTBG) 6308  Flynn 3282,  (PTBG), Wagner 6067  (MO,PTBG), Marr 1235,1236,1237  Forbes 1041.K  256—257, (BISH),  (UBC); Kalalau Pali,  (BISH,P), Flynn 3892,1870  (UBC); Honopu trail, Hobdy 119 Marr 540-Marr 558  (BISH), Lorence  (PTBG), Marr 249  (BISH), Wood 1188  (PTBG), Marr  (UBC); Kaunuohua Ridge, Flynn 3275  Stern & Carlquist 1348  Awaawapuhi Trail, Flynn 1459  (RSA,US), Marr 271  (UBC);  (BISH,PTBG), Wood 1803  Marr 273,275, Marr 577-Marr 589  (PTBG),  (UBC); Puu Ku-dividing ridge  87  between Kalalau and Pohakuao, Wood 2395 2394  (PTBG,UBC), Wood 2390,  (PTBG); Ridge west of Hanapepe River Heller 2614  (F,MASS,NY,P,US); Waialae Ridge, Marr 620—Marr 635 Kohua Ridge Marr 266,267,270 Wood 1104  (PTBG).  (UBC);  (UBC); Kalalau, below Puuokila  88  10.  Lysimachia maxima  64(1):47,  1987.  var. maxima R. 237  (R. Knuth)  Knuth in F. Pax and R. 1905.  Deg.  & I.  391,  1983.—TYPE: Molokai,  Deg.,  Phytologia  (1987):  950  in,  Knuth,  f.  Parks,  ex A. Gray  Pflanzenr.  Lysimachiopsis maxima  Plants Hawaii Nat.  north of Ohialele, John  John,  Lysimachia hillebrandii Hook.  (Heft 22) : 310,  by St.  St.  (Knuth)  Molokai,  0.  third edition, p.  south rim of Pelekunu valley,  Hillebrand s.n  just  (lectotype designated  fragment of B at BISH!).  Lysimachia ternifolia St. John, Phytologia 64(1):49, TYPE:  IV.  Pelekunu Tr., Forbes 242.Mo  1987.-  (holotype:  BISH;  isotype: NY!).  Nomenclatural note. from “Molokai!  Hillebrand  Pali of Pelekunu,  (1888)  first described plants  and smaller forms,  quite  glabrate, with subsessile leaves from Maunahui; E. Maui! Haleakala; var.’  ,  at heights of 3000 to 4000 ft.” as L. hillebrandii  which was not validly published.  it var.-  Knuth  (1905)  named  maxima and cited the Pelekunu and Haleakala  collections. L. maxima,  I have not seen any specimens from Maui labelled  however it is possible that the plant cited was j..  remyi subsp.  calicdnis, which also has densely tomentose stems  and whorled leaves.  Sturdy upright shrubs 1-2  in  tall; stems light brown or green,  densely brown tomentose at tip, older stem.  Leaves 3(—4)  retaining pubescence on the  per node,  sometimes alternate,  89  (O.l—)O.5—43(—70) mm apart, at base; blades ovate,  petioles  obovate or elliptic,  slightly rugose, margins revolute, (18-)23-30(—50) mm wide, abruptly acuminate,  (l—)2—3(—5) mm long, coriaceous,  (38-)55-60(-95)  base attenuate,  red  mm long,  apex acute to  upper surface dark green,  sparsely pilose,  lower surface lighter green, moderately pilose with red streaks;  secondary and tertiary veins prominent, pellucid.  Flowers solitary in leaf axils, pedicels  (5—)6(—7)—merous, campanulate;  (16-)25-30 mm long, densely tomentose,  lobes green,  lanceolate,  9-lO.5(-ll)  mm wide; corolla lobes obovate, (15—)15.5—l6(—17.5) mm long, (6.5—)7.5—8.5(—lO) mm long, 9—11 mm long.  (3—)3.5—4(-4.5)  lighter at tips  9—10 mm wide;  filaments  anthers l.5—2(—2.5) mm long; style  Capsules 8—9 mm long.  irregularly shaped,  Phenology.  red,  mm long,  erect; calyx  Seeds dark brown,  1-2 mm long.  Flowering June-July.  Distribution and Habitat.  Molokai.  Known from only the  leeward side of the island on the western rim of Pelekunu valley in Lowland Wet Forest from a single population of 45-50 individuals, Vaccinium, Labordia,  Growing with Metrosideros, Dicranopteris,  Psychotria,  Lycopodium, Machaerina, Hedyotis,  Cheirodendron,  Broussaisia, Comments.  950 in.  Styphelia,  Dubautia,  Sadleria, Elaphoglossum,  Scaevola and Cyrtandra.  This species has been distinguished by having  leaves that are ternately arranged While this is often true,  (Wagner et al.,  1990).  leaves are sometimes alternately  90  arranged and occasionally there are up to four leaves per node.  .  maxima appears to be most closely related to  .  remyi in which the leaves are also occasionally ternately arranged.  It differs from the latter in having much broader  leaves, thicker secondary veins, lobes,  longer calyx and corolla  secondary and tertiary veins that are more pellucid and  in having the upper surface of the leaf darker and the lower surface lighter.  The angle of divergence of the secondary  veins is also more obtuse in L. maxima than it is in Representative Specimens examined.  Pelekunu valley, 1135,  Marr 1137.  .  remyi.  MOLOKAI: western rim of  just north of Ohialele,  950 m, Marr 1123-Marr  91  11.  Lysimachia ovoidea St.  TYPE:  John,  Phytologia 64(l):48,  Kauai, west side of Wainiha Valley along narrow ridgetop  separating Wainiha and Manoa drainages, Kalanaililia, BISH!;  600-700 in,  southwest from  20 May 1976, Fay 581  (holotype:  isotype BISH!,PTBG!)  Sprawling shrub mostly branching from the base; in  long,  reddish-purple to green,  glabrous.  Leaves alternate,  (6-)7.5-12(—15) coriaceous, wide,  waxy,  (45—)60—80(—100)  barely open,  petioles  (18—)25—36(—41)  mm  slightly  secondary veins prominent,  Flowers solitary in leaf axils,  almost urceolate; pedicels  pendulous,  thin at base,  5—6—  (20—)24—30  broadening toward  dark red for full length; calyx lobes dark red,  narrowly lanceolate, lobes obovate, long,  mm long,  lower surface glabrous,  tertiary veins obscure.  calyx,  mm apart,  becoming  apex acute to cuspidate, upper surface  light green,  mm long,  pilose at the tip,  (1-)5-36(-48)  lighter; primary vein dark red,  merous,  stems up to 3  mm long, dark red; blades elliptic,  base rounded,  glabrous,  (—32)  1987.—  deep maroon,  4-5.5 mm wide;  mm long;  style  filaments  (2-)3-4 mm long,  mm long.  irregularly shaped,  6—8.5 mm anthers 1—1.5  Capsules 5-6 mm long. 1—2 mm long.  Flowering April-Aug.  Distribution and Habitat. populations,  1.5—2 mm wide; corolla  the margins lighter,  (3-)4-5(-5.5)  Seeds dark brown,  Phenology.  4—5.5 mm long,  Kauai.  Known from only two  the type location and the ridge between Limahuli  92  and Hanakapiai Valleys,  above Limahuli waterfall,  Growing in Lowland Wet Forest with Metrosideros, Eugenia,  Cyanea,  hex, Alyxia,  Pittosporum,  Psychotria,  Dicranopteris, Diplopterygium,  Broussaisia, Xylosma,  Scaevola, Dubautia,  Freycinetia, Bidens, Vaccinium, Melicope, Comments.  615—680 m.  Cibotium.  This is the only species in which all individuals  of a population have dark red pigment for the full length of the calyx and pedicel.  kalalauensis are  This species and j.  the only species that have pendant pedicels.  Leaf shape of j.  ovoidea is similar to that of L. waianaeensis and L. hillebrandii,  but the areoles are obscure  areoles of L. waianaeensis). smaller calyx lobes than  Lysimachia ovoidea also has hillebrandii.  .  (vs. prominent  The growing tips of  most Hawaiian Lysimachia present a gradual increase in leaf size from those that are just being initiated to those that are mature.  In j..  ovoidea,  the second youngest leaf is  nearly full-sized and significantly larger than the youngest leaf.  This is seen also in L. waianaeensis and j..  kalalauensis, but the transition is not nearly as abrupt. L.  In  ovoidea the angle between the petiole and the stem is  acute,  nearly parallel to the stem;  angle is more obtuse,  often nearly 900 to the stem.  Representative Specimens examined. Ridge,  Christensen 283,316  123 8-Marr 1254 (BISH),  (UBC);  Flynn 2164  533,534,535  (UBC).  in other species this  KAUAI: Wainiha—Manoa  (BISH), Wichman 253  (PTBG,UBC) Marr  Limahuli-Hanakapiai Ridge, Penman 18  (MO,PTBG,RSA), Marr 525-Marr 531, Marr  93  12.  Lysimachia pendens Marr,  sp. nov.—TYPE: Kauai, headwaters  of the north fork of the Wailua river, Lorence 5349,  (holotype: PTBG!).  720 in,  Figure 2.12.  23 July 1987, (NOTE: This is  not intended to effect the valid publication of a nomenclatural entity.)  Lysirnachiae waiaholeensi affinis a qua praecipue differt  foliis latioribus,  lanceolatis,  non linearibus; pagina  inferiore foliorum brunnea manente et puberula ubi matura; pedicello longiore.  Small, many branched, delicately pendulous or prostrate shrubs; stems 20—60 mm long, densely tan tomentose when young, eventually glabrate.  Leaves alternate,  (0.1—).5-3.5(-5) mm  apart, petioles l(—2) mm long; blades narrowly lanceolate, soft coriaceous, attenuate,  (20-)25-30(-45)  apex attenuate,  mm long,  2-4 mm wide, base  upper surface green glabrous,  lower  surface green, brown pilose; secondary and higher order veins obscure.  Flowers solitary in leaf axils,  (6-)9—l2(—l4) mm long, densely tomentose, long,  5—7—merous; pedicels  green, occasionally red below calyx,  erect; calyx lobes narrowly ovate,  4—6 mm  2-2.5(-4) mm wide; corolla lobes red to the tips,  obovate,  7.5-8.5 mm long,  long, red,  5-6 mm wide;  filaments 3.5-4 mm  anthers 1 mm long; style 3.5-4.5 mm long.  5-6.5(-7) mm long. 1.8 mm long.  Seeds dark brown,  Capsules  irregularly shaped,  1.2-  94  Phenology.  Flowering June-July.  Distribution and habitat.  Growing on vertical wet,  Kauai.  mossy or rocky cliffs with Machaerina, Selaginella, Dubautia,  Plantago, Hedyotis,  Athyrium,  Isachne,  Pipturus,  and Metrosideros.  Bidens,  Cyrtandra,  Known from only several  small populations at the headwaters of the north fork of the Wailua River, Comments.  720 m.  This species is a low,  with pendant stems. L.  almost mat—forming shrub  The characters that distinguish it from  filifolia are its broader leaves and tomentose leaves,  stems and pedicels. Representative Specimens examined.  KAUAI: Headwaters of the  north fork of the Wailua river, Wood 95,344 7253  503  (BISH), (UBC).  Penman 13078  (PTBG), Lorence  (BISH), Mann 261,262, Marr 470-Marr  95  10 mm  B  10 inn  Figure 2.12. branch.  Lysimachia pendens.  A.  flower; B.  flowering  96  13.  Lysimachia remvi Hillebr.,  Fl. Hawaiian Isi.  Lysimachiopsis remyi Heller, Minn. Bot. Lysimachiopsis remyi Parks, Remy,  no.  1:876,  1888.  1897.  (Hillebr.) Heller, Plants Hawaii Nat.  third edition, p. 1851—1855,  Stud.  284,  458  392,  1983.-TYPE: Maui, Voyage de M.J.  (lectotype here designated:  P1 ;isolectotype: K!).  Nomenclatural note.  Hillebrand (1888)  his description of Lysimachia remyi:  cited 6 collections in  “Maoui!  Waihee, Waiehu; Molokai! Halawa, Waikolu, Remy.” 1862)  Lysimachia hillebrandii var.7  Haleakala,  collected also by  angustifolia  (Gray,  collected by Remy on Maui is cited in Hillebrand (1888)  as a synonym.  A sheet from GH appears to be one of the  collections cited by Hillebrand.  It bears two collections,  one labelled “Lysimachia hillebrandii var.  linearis H.  f., W.  Maui: Gulch of Waihee” and the other “Lysimachia hillebrandii Hook.  f. var.  ancrustifolia, Molokai: Gulch of Halawa”.  On  this sheet the collection from Waihee lacks fruit and flowers and is otherwise a poor specimen; the collection from Halawa bears fruit and a single flower but has unusually narrow leaves for plants from Molokai,  nevertheless the angle of  divergence of the secondary veins in the leaves is typical of that of j.  remyi subsp.  subherbacea.  The specimen cited as  “Haleakala” by Hillebrand may have been a specimen from the Wilkes 1838—1842 Expedition, hillebrandii Hook.  f.  (US 76574),  verging to var.  no collection location is given.  labelled “Lysimachia  angustifolia”, however,  This specimen belongs to  L.  97  remvi subsp.  caliginis, distributed on both West and East Maui  (Haleakala). I have designated Remy 458 as the lectotype because it is a good flowering specimen and would appear to be among the material cited by Hillebrand.  The collection location is not  stated on the specimen chosen as the lectotype, however, St. John (unpublished manuscript) viewed a duplicate from GH and lists the collection location as Waiehu in his description of L.  stene St. John, which is based upon the same collection.  Sprawling to erect shrubs with stems up to 5 m long; stems red or green,  glabrous, glabrate or densely reddish—brown  tomentose.  Leaves alternate or whorled,  apart, petioles  (O.l—)O.5—30(55)  mm  (l—)2—lO(—12) mm long; blades linear,  oblanceolate, ovate,  elliptic, rarely orbicular, coriaceous,  sometimes undulate and rugose, margins sometimes revolute, (l5—)20—60(—95) mm long,  (l—)4—17(—33) mm wide, base and apex  acute to attenuate or rounded, upper surface light green to dark green,  glabrous, glabrate to tomentose,  slightly lighter, glabrous, pilose,  lower surface  or densely light brown  tomentose sometimes with scattered red streaks, veins sometimes pellucid. merous,  Flowers solitary in leaf axils,  campanulate, rarely urceolate; pedicels  mm long, glabrous to pilose, lobes entirely green,  (l—)l.5-3(-4.5)  (3—)lO—30(—70)  sometimes tomentose, erect; calyx  or red at the base,  or ovate, glabrous to densely tomentose long,  5—8—  linear,  lanceolate  (2.5—)4-8(-lO) mm  mm wide; corolla lobes obovate, red  98  (5—)8—14(—17) mm long,  (5—)6—9(—15) mm wide; filaments  (3—)4.5—7.5(—1l) mm long; style 3—6.5(—7) mm long. (4—)4.5—8(—l0) Comments.  Capsules  mm long.  This is an extremely variable species that  encompasses all plants from Oahu, Molokai and Maui that are upright shrubs, have leaves 4—20 mm wide,  red corolla lobes  less than 15 mm long and calyx lobes less than 10 mm long. Leaf morphology and spacing is often extremely variable among as well as within populations.  There are however, groups of  populations that share vegetative characters that distinguish them from other groups. populations,  Floral characters also differ within  but the range of variation overlaps among the  different groups of populations.  While it is possible to  determine which group of populations a specimen belongs to based on leaf morphology, this is not always possible if using floral characters alone.  Four subspecies are recognized,  based primarily upon vegetative differences.  Key to subspecies la.  Angle of divergence of upper secondary veins in the  leaves, more obtuse than lower; alternate. lb.  West Maui  leaves dark green,  13c. i.  remvi subsp. remyi.  Angle of divergence of secondary veins in the leaves  nearly uniform;  leaves dark or light green,  alternate or  whorled 2a.  Leaf veins pellucid; leaves often whorled; stems  (2)  99  usually densely tonientose.  13a. 2b.  renivi subsp.  Leaf veins not pellucid;  whorled; 3a.  j.  East and West Maui  leaves alternate,  (20—)45—60(—95)  mm apart,  if  otherwise alternate;  mm long; corolla lobes  (6—)ll.5—l4  mm long. Molokai 13d.  3b.  (3)  (.l—)l—22(—55)  closely spaced then nearly ternate,  (-17)  rarely  stems nearly glabrous or tomentose  Leaves distantly spaced,  leaves  caliginis.  Leaves closely spaced,  alternate;  leaves  (6—)8--lO(-l3)  L.  renwi subsp.  (.5-)l-6(-20)  (l5-)20--30(-60) mm long;  mm long.  mm  subherbacea. apart,  corolla lobes  East and West Maui 13b. L.  remvi subsp. kipahuluensis.  100  13a. Lysimachia remyi subsp.  caliginis  nov.; Lysimachia caliginis St. 1987.-TYPE:  E. Maui,  E.  John,  of Ukulele,  1919, Forbes 864.M (holotype:  (St. John)  Marr comb.  Phytologia 64(1):43, along edge of stream, July  BISH!;isotype:  NY[2J!).  (NOTE:  This is not intended to effect the valid publication of a nomenclatural entity.)  Lysimachia pentophylla St. TYPE:  John,  Phytologia 64(1) :48,  East Maui, Koolau Gap at treeline,  John & Mitchell 21266  (holotype:  BISH;isotype:  Lysimachia kukuiensis St.  John,  TYPE: W. Maui,  open bog,  Ewart 140  Puu Kukui,  (holotype:  Upright shrubs, long,  2 Sept.  1945,  St.  BISH!).  Phytologia 64:45-46, 1540 in,  1987.-  18 Dec.  1987.1928,  BISH!).  sometimes sprawling, with stems up to 2.5  often growing in dense clumps; stems light brown,  densely reddish—brown tomentose,  even when older.  in  often  Leaves  sometimes alternate, but often whorled, with up to 5 leaves per whorl,  0.5-19(-35)  mm apart, petioles  (1-)2-3(-5)  often dark red; blades lanceolate to ovate, undulate, long,  rugose,  the margins revolute,  (4-)8—l4(-20)  acute to attenuate, surface lighter, pellucid,  mm wide,  coriaceous,  (20—)30—45(-55)  base acute to attenuate,  upper surface dark green,  slightly pilose,  mm  apex  glabrous,  lower  scattered red streaks; veins  secondary and tertiary veins prominent.  solitary in leaf axils,  mm long,  Flowers  (5—)6—7—merous, campanulate; pedicels  101  (8-)13-20(—40)  tomentose,  mm long, glabrous to pilose,  purple veins, 3(-4.5) (—14)  erect; calyx lobes often  often red toward calyx,  dark red toward base,  sometimes  otherwise green with prominent reddish—  lanceolate to ovate,  (4-)5.5—7(—8) mm long,  2-  mm wide; corolla lobes dark red, obovate (7—)lO-1l  mm long,  6(—7) mm long, mm long.  (5—)7—8.5(—l2.5) mm wide; filaments red, anthers 1.5—2 mm long; style red,  (4—)5—6(-7)  Capsules 5—6.5(—7) mm long, widely ovate.  dark brown,  irregularly shaped,  Phenology.  Flowering May—December.  Distribution.  East Maui.  (3—)5—  Seeds  1.3-2.4 mm long.  Montane Wet Forests, growing at or  below treeline in Koolau Gap and Kaupo Gap, with Metrosideros, Blechnum, Vaccinium,  Styphelia,  Stenocivne,  Dubautia, Coprosma,  Cheirodendron, Machaerina, Rubus,  1660—1970 m.  West Maui.  Montane wet shrublands and wet forest on summits, growing with Metrosideros, Dicranopteris, Dodonaea, Broussaisia,  Sadleria, Lycopodium,  Dianella and Sphenomeris, Comments.  Styphelia, Coprosma,  Scaevola, Vaccinium,  1200—1760 m.  Lysimachia remyi subsp. kipahuluensis, j.  caliginis and  .  r.  subsp.  subherbacea differ from  .  .  subsp.  r.  subsp. remyi in having the angle of divergence of the secondary leaf veins nearly equal for all secondary veins, whereas those of L. r.  subsp. remyi become more obtuse toward Leaves of j. r.  the tip of the leaf.  subsp.  calicrinis and  subsp. kipahuluensis are generally less than 45 mm long, whereas those of  .  .  subsp.  subherbacea are generally  102  greater than 45 mm long. differs from j.  .  are often whorled, undulate,  caliginis  subsp. kipahuluensis in having leaves that larger, more ovate, darker,  and with revolute margins.  prominent and pellucid,  rugose,  Higher order veins are  and the stems are often densely  Lvsimachia remyi subsp. kipahuluensis and j.  tomentose. subsp.  Lysimachia remyi subsp.  .  caliginis form hybrid swarms on East Maui in Kaupo Gap  and near the old Waikau Cabin site in Koolau Gap.  In these  areas the range of leaf shape and size is continuous between the ovate leaves of subsp.  caliginis and the linear leaves of  subsp. kipahuluensis although these two species are not sympatric now,  they may well have been in the past.  Decades  of destruction of the native vegetation in Haleakala has undoubtedly affected species distributions. between j.  caliginis and subsp. remvi could also  subsp.  .  Hybridization  account for some of the variation in leaf shape and spacing in some specimens from West Maui. Representative Specimens examined.  EAST MAUI: Haleakala,  Koolau Gap, Marr 327-Marr 344, Marr 1019-Marr 1044 Degener 2551  (UBC),  (BISH,GH,MO,NY,P), Degener 2553  (BISH,GH,MO,NY,US), Degener 2552 (BISH,GH,MASS,MO,NY),  Rock 8632  (GH,NY), Degener 17668 (GH,BISH), Herbst 1620  (BISH),  Cariquist 1933  (RSA), Forbes 1014.M (P), Hobdy 745  (BISH),  Warshauer 2803  (BISH), St. John and Mitchell 21266  (BISH),  Perlman 10762,10767  (PTBG), Perlman 10769  (MO,PTBG); Location  unclear but in Haleakala Crater, Degener 2554  (BISH,GH,K,MASS,  103  MO, NY, US), s.n.  Degener 23296  (NY), Hitchcock 14960  (BISH,GH); Pipeline trail,  (BISH,GH,MASS,MO,NY,W);  (US), Woolford  Olinda, Degener 17686  East of Ukulele,  along edge of  stream, Forbes 864.M (BISH); Without locality: U.S. Expedition, Hook.f.,  under Captain Wilkes  (US),  as L. hillebrandil  verging to var. angustifolia.  Hitchcock 14820 Munro s.n. (GH,NY,US),  (US), Degener 25074  (BISH), Neal s.n.  Exploring  WEST MAUI:  (NY), Rock 8140  (BISH); Mt.  Puu Kukui, (GH,NY),  Eke, Degener 2550  Forbes 389.M (BISH); Hanakaoo, Forbes 62.M (BISH).  104  13b. Lysiinachja remyi subsp. kipahuluensis comb.  nov.  25:50, & I.  Lysimachia kipahuluensis St. John,  1971.  Deg.,  Lysimachiopsis kipahuluensis  Plants Hawaii Nat.  1983.-TYPE:  E. Maui,  Kuhiwa divide, Aug.  1945,  (St. John)  Pacific Science  (St. John)  0.  third edition, p.  Parks,  Deg.  391,  Haleakala, Lake Waianapanapa, Kipahulu  in low thicket,  crest of divide,  1720 in,  H. St. John and A.L. Mitchell 20,980,  BISH!;isotype: US!).  Marr  13  (holotype:  (NOTE: This is not intended to effect  the valid publication of a nomenclatural entity.)  Lysimachia angusta St. John, East Maui,  Phytologia 64(l):43,  Hana Forest Reserve near N.  rim of Kipahulu Valley  on steep inner slope of old cinder cone, Harrison 539  (holotype: BISH;  Lysimachia arta St. John, Maui,  Hana Forest Reserve,  Waianapanapa,  (holotype:  (holotype:  1973,  1987.-TYPE:  E.  near Waieleele, NE of Lake 2090 m,  29 June  BISH!).  Phytologia 64(l):44,  NE of Lake Waianapanapa,  Harrison 488  16 Nov.  isotype: BISH!).  Phytologia 64(l):43,  Lysimachia furcata St. John, East Maui,  1730in,  near rim of Kipahulu valley,  1973, Harrison 272  1987.-TYPE:  2120  in,  1987.-TYPE:  23 Nov.  1973,  DISH!).  Upright shrubs usually less than 1 m tall, mostly branching from the base,  but up to 2 in tall,  often growing in dense clumps  105  several meters square; stems light brown to red, glabrous to minutely pilose. petioles  coriaceous  (15-)20—45(—60) mm long,  (l—)2-6  mm wide, base attenuate and apex acute to acuminate,  upper surface light green, slightly pilose,  glabrous,  (5—)6—7(—8)—merous,  campanulate; pedicels toward calyx,  Flowers solitary in leaf  sometimes urceolate,  (5-)ll-l7(-31) mm long,  lightly pilose,  glabrous or pilose, at base,  lower surface paler,  secondary veins pellucid, tertiary veins  usually obscured by thick cuticle. axils  0.5-6(-20) mm apart,  (l-)2-3(-4) mm long; blades linear to oblanceolate or  narrowly ovate, (—14)  Leaves alternate,  but usually often maroon  erect; calyx lobes green,  linear to narrowly ovate, occasionally red  (3—)5—6(—9.5) mm long,  (l—)2—2..5(—4.5) mm wide;  corolla lobes red, much lighter toward margins, (6—)8—10(—l3) mm long,  (5—)6—7(—9.5) mm wide;  (3—)4.5—5.5(—8) mm long,  red,  style  (4-)5--6.5(-7) mm long,  6(-7)  mm long.  anthers red.  Seeds dark brown,  obovate,  filaments  (1—)1.5(—2)  mm long;  Capsules widely ovate, irregularly shaped,  5-  0.9-2.1  mm long.  Phenology.  Flowering March through November.  Distribution and habitat.  East Maui.  Montane Wet Forest in  Haleakala Crater above Paliku Cabin growing with Metrosideros, Cheirodendron,  Coprosma, Melicope, Myrsine, Blechnum,  Elaphoglossum, Peperomia, Astelia,  Rubus, Dubautia, Vaccinium.  Montane Mesic Forest of the east and west sides of Kaupo Gap. Subalpine Mesic Shrublands above treeline and into the forest  106  around the rim of Kipahulu Valley growing with Metrosideros, Styphelia,  Coprosina,  Blechnum, Vaccinium,  Deschampsia, Lycopodium,  Nephrolepis,  Dubautia,  from 1510-2330  in.  Growing with Metrosideros, Broussaisia, Cheirodendron, Vaccinium, Psychotria,  Selaginella, Machaerina,  Dicranopteris, Acacia,  Clermontia and Carex along streamsides or similar  high moisture environments in Kipahulu Valley, Waihoi Valley and Hana Forest Reserve from 690-1750  in.  West Maui.  wet cliffs of streamsides growing with Pipturus, Artemisia, 370—900  Selaginella, Hedyotis,  Bidens,  Coprosma and Metrosideros  in.  Comments.  Lysimachia remyi subsp. kipahuluensis is distinct  in having light green, ovate leaves, wide.  Along  closely spaced,  linear to narrowly  generally less than 6 mm wide,  but up to 14 mm  Plants with very narrow leaves are scattered along  streamsides and in bogs in Kipahulu Valley, Waihoi Valley and the Hana Forest Reserve.  Broader leafed plants, more typical  of the subspecies also occur in these areas. Plants from West Maui have somewhat longer corollas mm)  than do those from East Maui  (6-7 mm).  Representative Specimens examined. Paliku cabin,  17670, 21400  Marr 281,  EAST MAUI: Gulch SE of  Marr 678-Marr 705  (GH,MASS,MO,NY), Degener 17753 (BISH,NY), Degener 17754  (UBC), Degener  (GH,MO,NY,P), Degener  (NY), Higashino 849  (BISH,US),  Henrickson 3509  (BISH,RSA,US), St. John 21117  South of Kuiki,  east side of Kaupo Gap, Degener 17671  (GH,MO,NY,US), Rock 8605  (6—9  (GH,RSA,US);  (BISH,GH,NY,P,PTBG,US), Olson 6  107  (BISH,US);  Small gulch between Healani Gulch and Kaupo Gap,  Gagne and Montgomery 605  (BISH), Hobdy 500  of Haleakala Mtn., Degener 17755  (BISH); Cliffs SE  (BISH,GH,MASS,MO,NY,US);  Kalapawili Ridge E of Pohaku Palaha to Lake Waianapanapa, Marr 282—Marr 291,  Marr 636-Marr 666  (UBC), Harrison 531, 272  (BISH), Forbes 1195.M (K,US), Henrickson 3915 Henrickson 3913  (BISH),  Valley, Higashino 10190 Henrickson 3884  Penman 10517  (BISH,US),  (PTBG); Upper Kipahulu  (BISH); S of Paliku cabin, Kaupo Gap,  (BISH, RSA, US); Waihoi Valley, Harrison 26  (BISH); Waihonu Stream, Waihoi Valley, Nagata 1049 Upper Hana Forest Reserve “mid—camp”, 567  (BISH),  3217  stream bank, Harrison  inner slopes of old cinder cone Harrison 409  (BISH); Kipahulu Valley, 3212,  Palikea Stream, Wood 3201,  Stream, Anderson s.n.  442  WEST MAUI:  (UBC),  (UBC), Kokowai  (UBC); Upper Kipahulu valley, Marr 430  Nakalaloa stream, Mann 432, Marr 434-Marr  (UBC); Black Gorge, Mann 1258-Marr 1263  (PTBG).  3202,  (PTBG), Marr 1279—Marr 1284, Marr 1290,1291  below upper valley plateau, Mann 1287,1288,1289  (UBC).  (BISH);  (UBC), Wood 0333  108  13c.  Lysimachia remvi subsp. remvi. ,  Hook.  Phytologia 64 (1): 49,  Deg.,  Plants Hawaii Nat. Parks,  W. Maui,  Proc. Am. Acad. Arts Sci.  Lysimachiopsis angustifolia  1862.  Lysimachia ciliolata St. John.  1920,  Lysimachia hillebrandii  1987.  f. var. angustifolia Gray,  5:329,  Lysimachia stene St.  390,  Phytologia 64:43,  1987.-TYPE:  1983.  14 May  Forbes 2369.M (holotype: BISH!; isotype: W!).  TYPE: Maui,  Phytologia 64(1):44,  ridge left of Lahainaluna Valley,  Forbes 325A.M.  (holotype:  Feb.  ridge left of Lahainaluna Valley,  325B.M.  (holotype:  BISH! ;isotype:  Lysimachia lydgatei Hillebr. Lysimachiopsis lydgatei  Hillebrand s.n.  isotype:  GH,K).  1913,  Fl.  Feb.  1987.-TYPE:  1913, Forbes  BISH!).  Hawaiian Isl.  (Hillebr.)  1897.-TYPE: W. Maui,  Lahaina,  1987.—  BISH!).  Lysimachia lata St. John, Phytologia 64(1):46,  1:876,  & I.  third edition, P.  secondary ridge at right hand head of Olowalu,  Lysimachia elliptica St. John,  Maui,  0. Deg.  (Gray)  Heller,  p.  in Minn.  284, Bot.  (holotype:  Hanaulaiki, BISH!).  Stud.  on slopes and in gulches back of  (holotype:  fragment of B at BISH!;  Lysimachia occidentalis St. John, Phytologia 64(1):47, TYPE: Maui,  1888.  1000 m,  14 March 1972,  1987.-  Hobdy s.n.  109  Lysirnachia pedicellata St. John. TYPE: 431  Phytologia 64:48,  Lanai, ridge at head of Maunalei,  (holotype:  1916, Munro  BISH!).  Lysirnachia pilosula St. John,  Phytologia 64(l):48,  Maui, Hanaula, June 1910, Forbes 114.M.  Lysimachia scansoria St. John. Lanai, Munro s.n.  (holotype:  Spreading shrub up to 1.5 base,  14 Oct.  1987.-  in  1987.-TYPE:  (holotype:  Phytologia 64:49,  BISH!).  1987.-TYPE:  BISH!).  tall,  branching mostly from the  often with very short axillary shoots; stems green,  light brown or red, Leaves alternate, (l—)l.5—10(—12) oblanceolate,  glabrate to densely rusty tomentose.  (0.1-)0.5-30(-50) mm apart, petioles  mm long; blades linear,  ovate or elliptic,  (15—)20—55(—90)  mm long,  attenuate or rounded,  apex acuminate,  mm wide, acute,  coriaceous,  base acute,  or attenuate,  glabrous to densely brown tomentose,  lower surface much lighter, tomentose,  rarely orbicular,  (1—)2—17(—33)  upper surface dark green,  lanceolate,  glabrate to densely brown  especially along the primary vein; secondary veins  prominent or obscure,  tertiary veins usually obscure.  Flowers  solitary in leaf axils,  (5—)6—8—merous,  (3—)10--25(--65)  densely tomentose, usually erect,  nun long,  sometimes pendulous;  campanulate; pedicels  calyx lobes lanceolate,  sometimes dark  110  red at base, long,  glabrous to densely tomentose  (l-)l.5—3(-3.5)  (2.5-)4—7(—9) mm  mm wide; corolla lobes obovate,  (5—)8—lO(—l3)  mm long,  (4—)5—7(—9)  (2.5—)3—5(—7)  mm long;  style red (3—)4—6(—6.5)  Capsules shaped,  (5—)6—8(—l0)  mm long.  filaments red mm long.  Seeds dark brown,  irregularly  1.3-2.9 mm long.  Phenology.  Flowering Feb.  -  Distribution and habitat.  Oct. West Maui.  the leeward side of the island. Dodonaea,  Styphelia,  Cheirodendron,  670—1110 m.  Montane Mesic Forest on  Growing with Metrosideros,  Sadleria, Vaccinium,  Coprosma,  Dubautia, Hedyotis, Myrsine,  Astelia, Wikstroemia,  Comment.  mm wide;  red,  Lanai.  Selacrinella, Mountains,  Pipturus,  Alyxia,  Scaevola and Lobelia at  650 m.  This is easily the most variable taxon of Hawaiian  Lysimachia.  Leaf length, width and shape vary from linear and  glabrous to broadly elliptic or even orbicular and often densely toinentose. glabrous leaves,  At one extreme are plants with nearly  15 mm long and 2 mm wide,  been classified as j.. al.  (1990).  these would have  remyi in the classification of Wagner et  At the other extreme are plants such as a  specimen from Lihau,  Hobdy 519  densely tomentose leaves, length, width,  (BISH)  which has elliptic,  90 mm long and 30 mm wide.  The  and degree of pubescence of the calyx lobes is  also quite variable in many populations. leaves and tomeritose stems,  leaves,  previously classified as  lydgatei.  .  Plants with elliptic  and calyx lobes were This species is not  111  recognized here because of the continuum in form between it and nearly glabrous plants with linear leaves.  Representative Specimens examined. WEST MAUI: Hobdy 1212  (BISH); Lihau Hobdy 519  Marr 801-Marr 870  8425  (BISH), Marr 408-Marr 430,  (UBC), Hobdy 515,516,823  (BISH),  Penman  (MO,PTBG,UBC); Hanaula, Manawainui Gulch, Degener 17687  (GH,NY,US,BISH), Gustafson 2078,2069,2781 (BISH,P,US), Nagata 961 248  Kaulaula Canyon,  (BISH)  Degenen 22063  871-Mann 927  (RSA), Forbes 114.M  (BISH,US), Nagata 1923 (NY,MASS)  (BISH), Hobdy  Marr 350—Marr 356,  Marn  (UBC); ridge north of Pohakea Gulch, Degener 2556  (US); Pohakea Gulch, (BISH), Nagata 1918  Gustafson 2075  (RSA), Hobdy 2671,2690  (BISH), Wood 1170  Hanaula, Wood 1176,1179  (PTBG);  (PTBG);  south of  left-hand side of Olowalu,  Forbes 2300.M  (BISH,W); Olowalu,  Forbes 2368 .M  (NY, P), Forbes 2247 .M (BISH,K), FORBES 2248 .M  (BISH); Kahoolewa ridge, Hillebrand s.n.  Penman 10581  (PTBG); Waihee Valley,  (GH); Helu summit, Marr 1264—Mann 1268  Penman 10729,10728  (GH,NY[2j,US); Black Gorge,  (GH,NY); between Kinihapai and Ae streams,  Gustafson 3300  (RSA); upper Ukumehame Canyon, Hobdy 1259  (BISH); Wailuku Pali, Forbes 2440.M (BISH); Kaanapali, Rock 8164 (BISH).  (UBC),  (PTBG); Hale Pohaku, Marr 933-Marr 1018  (UBC); Puu Lio, Degener 12909 Degener 23735  right hand side of valley,  LANAI, Mtns.  (BISH);  location uncertain, Forbes s.n.  east end, Forbes 221.L (BISH,MO,NY,  P,US); Kaiholena, Forbes 387.L (BISH,GH,K,NY);  lowlands back of  (BISH); Manaha, Rock 8096  ridge above Maunalei, Munro 627  (BISH,W).  112  13th Lysimachia renwi subsp.  subherbacea (St. John)  nov.  (Hillebr.)  Lysirnachia subherbacea  64 (1): 49,  1987.  Hillebr.,  Fl. Hawaiian Is.  subherbacea Parks,  L. hillebrandii Hook.  (Hillebr.)  third edition,  0. p.  Halawa, Hillebrand s.n., of B at BISH!).  283, Deg.  393,  1888. and I.  1983.  St.  John,  Marr comb.  Phytologia  f. var. subherbacea Lysimachiopsis Deg.  Plants Hawaii Nat.  —TYPE: Molokai,  (lectotype here designated:  gulch of  fragment  (NOTE: This is not intended to effect the  valid publication of a nomenclatural entity.)  Lysimachia attenuata St. John,  Phytologia 64(1):43,  TYPE: Molokai,  Kahaunui Gulch,  12 May 1928, Degener 17676  (holotype:  isotype: MO!).  GH;  Lysimachia fauriei St. Molokai, Kamolo, GH!;  isotype:  John,  1,000 m,  Phytologia 64(1):44,  June 1910, Faurie 705  BISH!,P!).  1987.—  1987.-TYPE:  (holotype:  (St. John changed the collection  number on the P specimen to 705A).  Lysimachia kalaupapaensis St.  John,  1987.—TYPE: Molokai, Kalaupapa Pali, 14030  (holotype:  Phytologia 64(1):45, 520 m,  23 May 1918, Rock  BISH!).  Lysimachia mucronata St.  John,  Phytologia 64(1):47,  1987.-  TYPE: Molokai, Wailau Valley-Waiehu, June 1910, Forbes 528.Mo (holotype:  BISH!;  isotype:  GH!,K!,P!,W!).  113  Lysimachia munroi St. John, Phytologia 64 (1) :47, Molokai, 127  edge of Waihanau Valley,  23 Jun 1927, Munro  (holotype: BISH!).  Lysimachia rockii St. John, Molokai, Mapulehu, 6146  770 in,  1987.-TYPE:  (holotype:  ridge to Kamakou 770  BISH! ;isotype:  Lysimachia rufa St. John, Molokai,  A! ;isotype:  TYPE: Molokai, Waikolu, (holotype:  Nomenclatural note.  1987.-TYPE:  22 March 1910, Rock  BISH! ,P! ,US! ,W!).  1987.-TYPE:  28 April 1928, Degener 17679  BISH! ,NY!).  Lysimachia waikoluensis St.  John 12348  in,  Phytologia 64(1):49,  Puu Kaeo, Waikolu,  (holotype:  Phytologia 64(1):48,  John,  Phytologia 64(1):50,  Hanailoilo,  BISH;  1110  isotype:  Four specimens,  labelled “Moaui; voyage de M.J. Remy,  in,  23 Dec 1932, St.  BISH!).  two at K and two at P are 1851-1855 no 458’.  is the same collection number as the type specimen of from Maui. belong to j.  1987.-  .  This remvi  One specimen from each of these herbaria clearly  r.  added after 458,  subsp.  subherbacea and the letter “b” has been  perhaps by St.  specimen in 1976 as  .  John, who annotated the  occidentalis St.  John.  The other  specimen from K is labelled 458a and the second one from P remains 458.  Both of these clearly came from West Maui.  114  Branching shrubs, acquiring an almost vine like habit; stems brown to red, up to 5 m long, young stems rusty tomentose, older glabrate.  Leaves alternate, sometimes almost ternate,  (O.l-)l—22(—55) mm apart, petioles linear,  (1-)2—3(-7) mm long; blades  lanceolate, oblanceolate, or obovate,  sometimes slightly undulate,  coriaceous,  (20—)45-60(—95) mm long,  11(—20) nun wide, base acute to attenuate,  (3-)8-  apex acute to  attenuate, upper surface olive green, glabrous above, the veins slightly rugose, pilose,  lower surface lighter,  initially  becoming glabrous, with scattered streaks of red;  secondary veins prominent, tertiary veins obscure. solitary in leaf axils, (11—)20-30(—70)  (4—)5—7—merous, campanulate; pedicels  mm long, glabrous to densely tomentose, often  red toward calyx, pendulous; calyx lobes green, base,  Flowers  lanceolate to narrowly ovate,  often red at  (3—)5—8(—1O) mm long,  (2-)3—4 mm wide; corolla lobes obovate, rose pink to dark red at base,  lighter toward the tips  (5—)7.5—9.5(—15) long, long.  (6-)ll.5-14(-17)  mm wide; filaments red,  anthers 2-3.5 mm long; style red, Capsules  (5—)6-8(-9) mm long.  irregularly shaped,  Phenology.  mm long,  (4.5—)6—7.5(—ll) mm (5-)6.5-8(-ll) mm  Seeds dark brown,  1-2.1 mm long.  flowering March-Nov.  Distribution and Habitat.  Molokai.  Eastern side of the  island at higher elevations, on both leeward and windward sides,  490—1130 in.  Found in lowland inesic shrublands and  Lowland Mesic Forest dominated by Metrosideros, Stvphelia,  115  Dodonaea, Vaccinium, Wikstroemia, Dianella,  Psychotria,  Chamaesyce,  Alyxia,  Selaginella, Dubautia,  Pipturus,  Elaphoglossum,  Peperomia,  Nestigis,  Diospyros, Viola and Coprosma.  Oahu.  Collected in 1993 for the first time from the Waianae  Mtns.,  785—835 in.  Lipochaeta, Hedyotis,  Growing on cliffs with Dubautia, Schiedea,  Tetramolopium,  Carex, Viola,  Bidens, Panicum,  Lepidium, Lobelia and Pleoinele.  Comments.  Leaf and calyx size and shape are somewhat variable  both within as well as among populations of this species. Plants in gulches generally have broader leaves and longer stems than those on drier hillsides.  A few collections from  the ridges and gulches of the south side of Molokai resemble  I.. maxima in having nearly ternate leaves, but differ in that the leaves are not as broad,  nor are the calyx and corolla as  long as that of L. maxima. Representative Specimens examined. Gustafson 3002  (BISH),  (RSA), Forbes 166.Mo  Anderson 514  1169; Kamalo,  Faurie 706  Degener 33497  33500  (NY),  1948  (UBC); Wailau Valley (BISH), Degener 17677  (NY); edge of Waihanau Valley, Degener (BISH); Waikolu, Marr 391-Marr 404,  (UBC),  Carlquist 2216  (BISH), Degener 23733 (US),  806  (BISH); Makakupaia, Degener 33508,  Marr 1194-Marr 1207  Munro s.n.  Marr 1103-Marr 1122 23437  (BISH), Davis 772,  (GH,P); Kaulahuki, Evans s.n.  (GH),  Puu Kolekole,  (BISH), Marr 371-Marr 377, Marr 1138-Marr  Rock 7024  Marr 378—Marr 382,  MOLOKAI:  St. John 12348  (BISH,RSA), St. John  (NY), Degener 23734  (K),  Gillett  (NY); near Laianui, Degener 17678  (BISH,GH,NY); Hanalilolilo, Lorence 6324  (PTBG); Onini gulch,  116  Mill 502  (BISH), Davis 846  Kalaupapa Rim, Forbes 24.Mo,  Harrison s .n.  Penman 6602  (PTBG,UBC),  Lorence 5630  1102  John 19880  (GH,NY);  Davis 880  (UBC);  17676  St.  (NY);  (K,P),  Halawa Gulch,  Hillebrand s.n.  (BISH,NY),  Marr 1066-Mann  (GH,NY), (BISH),  Degener Rock 6146  ridge to Olokui along seacliff, Wood  (US 809323).  Makua Keaau Forest Reserve, 2496,2490,2633,2634  (BISH),  Degener 33498  no location given,  Hillebnand s.n.  (BISH),  Marn 1171-Marr 1193  Degener 34499  Cuddihy 1244  Kahuaawi gulch,  (MO,PTBG);  Warshauer 2317  ridge between Kaunakakai and Kupaia  (BISH),  (PTBG); Wailau Valley, 1258  (PTBG),  (PTBG), Marr 367—Mann 370,  Degener 17674  (BISH),  Krajina 620612006  Ravine N of Puu Makaliilii,  gulches,  Southworth s .n.  Rock 14030,  (BISH,MO, PTBG),  (UBC);  (BISH),  (UBC);  (BISH,MO,US),  (UBC); Kawela,  Lorence 5615  (BISH), Marr 1208—Marr 1229  Remy 457 QAHU:  Ohikilolo area,  (PTBG), Wood 2633,2490  (P),  Remy 458b,  Waianae Mtns., Wood (UBC,PTBG).  117  14.  Lysimachia scopulensis Marr sp. nov.  —TYPE: Steep cliffs  of Kalalau Valley rim, north of Kahuamaa Flat, (PTBG!).  Wood 634  Figure 2.13.  3 Mar.  1991,  (NOTE: This is not intended  to effect the valid publication of a nomenclatural entity.)  Lysimachiae waianaeensi similis magnitudine et forma calycis,  sed foliis paene sessilibus, areolis et venis secundariis obscuris.  Folia atro—viridia et saepe pulverulenta,  deorsum curvata,  ad apicem  caulibus semper pulverulentis.  Branching shrubs up to 75 cm tall; stems red or green, pulverulent when young. succulent, 4.5(-5)  glandular,  Leaves alternate, coriaceous,  (O.5-)2-7(-8)  mm apart, petioles  almost (2-)3—  mm long; blades linear to narrowly oblanceolate,  sometimes narrowly obovate, tips slightly to extremely recurved, acute,  (33—)55—65(-86) mm long,  (5—)8—ll(—23)  apex acute, upper surface glabrous,  olive green, pulverulent when young, dark green,  dark green, drying  lower surface glabrous,  drying lighter than above; primary vein and  petiole sometimes dark red above and below, prominent,  mm wide base  tertiary veins obscure.  secondary veins  Flowers solitary in leaf  axils,  (5-)6(--7)—merous,  green,  erect; calyx lobes green or red at the base, the tip  campanulate; pedicels 20-45 mm long,  acute, margins scarious, widely ovate, pulverulent (2.5—)4—5 mm long, burgundy,  (2-)3-4(-4.5) mm wide; corolla lobes obovate, 10-11 mm long,  6-7 mm wide; filaments 5 mm long,  118  anthers 2 mm long; style 5 mm long. long.  Seeds dark brown,  Phenology.  Capsules 6.5-7.5(-9)  irregularly shaped,  1.5-2.5 mm long.  Flowering March.  Distribution.  Kauai.  Growing on steep cliffs of Diverse  Lowland Mesic Forest in the upper part of Kalalau Valley, 880 m.  Associated with Hedvotis,  Nototrichium,  Stenogyne,  Remya, Wilkesia,  Dubautia,  Coprosma, Vaccinium, Rumex, Comments.  mm  780-  Chamaesyce, Hibiscadelphus,  Melicope, Lepidiurn,  Lobelia, Myrsine,  Lipochaeta, Metrosideros,  and Exocarpus.  The most distinguishing characteristics of this  species are the recurved leaf tips, upper leaf surfaces,  the dark green and shiny  and the pulverulent young leaves and  stems. Representative Specimens examined. Puuokila, Wood 1008 2036  (MO,PTBG).  KAUAI:  (UBC,PTBG), Wood 798,  Kalalau rim, below 1233  (PTBG), Wood  119  10mm  ....  •1 ; ‘: *  ;•  *:*.::  10mm  B  A  Figure 2.13. branch.  Lysimachja scopulensis.  A.  flower; B.  flowering  120  15.  Lysimachia venosa  (Wawra)  St. John,  1987.  Lysimachia hillebrandii Hook.  Wawra,  Flora 5:523,  Deg.  & I.  393,  1983.—TYPE:  2165  (holotype: W!;  Deg.,  1874.  f.  Phytologia 64:50, ex A.  Lysiinachiopsis venosa  Plants Hawaii Nat. Parks, KAUAI,  Gray var. venosa  1600 ra, Wawra  fragment at BISH!).  Shrubs 0.5—1 m tall; stems brown, pilose when young, glabrous.  Leaves alternate,  1—6 mm long; blades obovate, long,  (12—)20-35(-48)  upper surface dark green, above,  (3-)lO-15(-25) coriaceous  mm wide,  dark veins,  green,  apex acuminate,  tertiary veins  6—7—merous; pedicels  erect; calyx lobes green with prominent  narrowly lanceolate,  corolla lobes obovate, filaments 7 mm long,  mm  lower surface lighter than  Flowers solitary in leaf axils,  15—27 mm long,  mm apart, petioles  base attenuate,  glabrous,  becoming  (50-)75-80(-100)  glabrous; secondary veins prominent,  obscure.  0.  third edition, p.  summit of Mt. Waialeale,  isotype: W!,  (Wawra)  dark red,  13—16 mm long,  4—6 mm wide;  15-19 mm long,  10-11 mm wide;  anthers 3 mm long;  style 9-10 mm long.  Capsules not seen.  Phenology.  Unknown.  Distribution.  Kauai.  Known from two collections from the  type location, where it was last collected in 1911.  No recent  collections from there despite several visits by botanists in recent years.  A 1991 collection (Wood 784)  from the  headwaters of the North fork of the Wailua River  (NE corner of  amphitheater) was of a broken branch that had been dislodged  121  from the cliffs above, possibly from the summit area of Mt. Waialeale. Comments.  The size and shape of the leaves of this species  closely resemble j. cilutinosa. dark red corolla,  is not viscid,  j..  venosa differs in having a  and has longer calyx lobes  versus those of L. glutinosa which are ovate. Representative Specimens examined.  Waialeale, Rock 8881  KAUAI:  (BISH,GH); Wood 784  Summit of Mt.  (PTBG).  122  16.  Lysimachia waianaeensis St. John.  1987.—TYPE:  Oahu,  Phytologia 64(1):50,  Puu Kanehoa, Waianae Mtns.  1934, St. John 14012  (holotype:  BISH;  830 in,  7 Jan.  isotype: BISH!).  Sprawling shrubs with leader shoots up to 4 m long and much shorter secondary shoots; stems green to light brown, with distinct linear lenticels, usually pilose at tips, Leaves alternate, (-15)  mm long,  (.5-)2-18(-26)  mm apart,  coriaceous,  nun wide,  base rounded,  (30—)50—65(—100) mm long,  i.e.  distinct.  Flowers solitary in leaf axils,  secondary,  campanulate; pedicels  mm wide,  long,  (10-)l5—25(—46)  green,  subglobose,  style  Phenology.  obovate,  filaments  mm long.  (l—)2.5—  rarely  (9-)ll—13(—l4) mm  (5—)6—7.5(—8)  (5-)6—7(-8) mm long.  (5.5-)7-9(—10)  irregularly shaped,  mm long, green,  (3-)4-5(-8) mm long,  corolla lobes red,  anthers 1.5—2 mm;  5-6(-7)-merous,  usually widely ovate,  (6—)7—9(—lO) mm wide;  areoles  and all higher order veins  erect; calyx lobes  lanceolate;  (2-)5-8  (8—)16—24(—39)  apex abruptly acuminate;  prominent,  3.5(—5)  petioles  often red; blades elliptic to slightly ovate or  obovate,  glabrous,  glabrate.  mm long,  Capsules  Seeds dark brown,  1.6-2.3 mm long.  Flowering Nov.-May.  Distribution and habitat. the Waianae Mtns,  Oahu.  540-1140 m,  Acacia,  Charpentiera,  Alyxia,  Pleomele,  Cyanea,  Lowland mesic forest only in  growing with Metrosideros, Selaciinella,  Wikstroemia, Viola,  Sida,  Bidens,  Hedyotis,  Claoxylon,  123  Chamaesvce,  Diospyros, Psychotria,  Eragrostis,  Hibiscus,  Antidesma, Comment.  Santalum,  Blechnum,  Elaphoglossum, Melicope,  Dodonaea,  Schiedea and Dubautia.  The most distinct characteristic of this species is  the prominent areoles in the leaves. Mtns.  Canthium,  In the southern Waianae  this species grows sympatrically with  gulches NE of Palikea  (800-870  Nanakuli and Lualualei hybrids between L.  and the ridge between  in)  (720-820  hillebrandii in  The evident lack of  in).  hillebrandii and j..  waianaeensis in this  area suggests that the two may be reproductively isolated from each other.  This species typically lacks pigment on the calyx  and pedicel and the calyx is ovate.  However, the calyx of an  exceptional plant grown in the greenhouse from seed collected in Makaha Valley,  on Kamaileunu Ridge  (Waianae Kai), was  entirely red and lanceolate. Representative Specimens examined. Puu Hapapa,  Krajina 620401115  (BISH), Degener 17672  (BISH,GH,MO,NY,US), Degener 12368 Makaha Valley, 5053  33505  Kamaileunu Ridge,  (PTBG, BISH), Penman 6818  Degener 4138  (GH,MO,NY,US),  Penman 5103  Obata 279  N.  (MO,PTBG),  (K,MO,US),  (BISH), Degener  central ridge Degenen 21278  (BISH); Puu Kaua, Mann 231 Mann 778—Mann 788  Penman 5110  (BISH),  (BISH);  of Puu Kawiwi, Penman  (PTBG), Forbes s .n.  Mann 239-Marr 245,  (NY,W), Obata s.n. 1155  (NY), Degener 21004  (BISH,GH,MO,NY,US), Nagata 1132  (P); Makaleha Valley,  Mann 233,234,  OAHU: Waianae Mtns.,  (UBC),  (BISH), Degener 33507  Takeuchi 2059,3719  (BISH), Montgomery sn.  (PTBG),  (BISH), Nagata  (BISH); Makua Valley, Degenen  124  17684  (GH,NY,US), Degener 17682  17680  (GH); N slope Mt. Kaala, Degener 19439  Carlquist 2354 Fosberg 13011  Reserve,  (NY); Puu Kanehoa, Degener (GHNY),  (RSA); E ridge PUU Kalena Hartung s.n. (BISH,GH), Kerr s.n.  Ohikilolo, Wood 2635  (BISH); Makua Forest  (PTBG), Koiahi, Wood 2494  (PTBG); Pahole NAR, Kukuiula Gulch, Fosberg 13072 Kapuna Gulch, Marr 448,449,450 Marr 777  and Lualualei, Marr 1295,1296 Oahu, 2211  U.S. (W).  (BISH),  (UBC), Pahole gulch, Marr 769—  (UBC); Mokuleia, Forbes 1775.0  Olelo Gulch, Marr 443-Marr 447  (BISH),  (BISH); Napepeiau  (UBC); ridge between Nanakuli (UBC); Location unclear: Mtns.,  Exploring Expedition 1838-1842  (US); Kaala Wawra  125  chapter 3 Allozyme diversity in endemic Hawaiian Lysimachia  3.1  Introduction The extreme geographical isolation of the Hawaiian  Islands makes dispersal of propagules to the islands a rare event.  Many plants have evolved in isolation from their  closest relatives, resulting in a flora that is 89% endemic at the species level  (Wagner et al.,  1990).  Although the  Hawaiian flora and fauna are frequently cited as dramatic examples of adaptive radiation,  it is often overlooked that  68% of the 216 native (32 endemic)  plant genera have not  undergone adaptive radiation or are represented by only two species  (Lammers,  species—rich.  1990).  Nevertheless, many genera are quite  For example, the 52 species of the endemic  genus Cyanea have presumably evolved from a single ancestor (Wagner et al.,  1990).  For genera in which a monophyletic  origin can be demonstrated, the Hawaiian Islands are an ideal location to investigate the initial stages of speciation. contrast to studies of continental taxa, however,  In  few  biosystematic studies of island genera have utilized allozyme or DNA analysis of species to elucidate phylogenetic relationships and genetic divergence associated with primary speciation following long-distance dispersal 1987b).  (Crawford et al.,  Only five genera of Hawaiian angiosperms have been  analyzed using allozyme variation to estimate the degree of genetic divergence that is associated with adaptive radiation:  126  Tetramolopium  (Lowrey and Crawford,  and Ganders,  1985),  and Wilkesia  (Witter and Carr,  (Aradhya et al.,  Dubautia  1991).  1985),  Bidens  (both n=13 taxa and n=14 taxa) 1988),  and Metrosideros  Species in each of these genera are  morphologically and ecologically diverse,  yet with the  exception of the n=14 species of Dubautia, are high.  (Helenurm  genetic identities  This contrasts with continental taxa where species  pairs that have high genetic identities are usually very similar morphologically Outside of Hawaii,  (Crawford et al.,  l987b).  allozyme variation within and among  populations of insular taxa has been studied in genera from the Juan Fernandez Islands 1992,  1993),  the Galapagos Islands  Wendel and Percival, 1990).  (Crawford et al.,  1990)  l987a,  1990a,  (Wendel and Percy,  and the Bonin Islands  1990;  (Ito and Ono,  Most of these studies addressed questions regarding  (continental)  ancestor—(insular)  derivative relationships of  genera possessing a single insular species.  While these have  been useful in estimating genetic variation within populations, (1992,  except for Ito and Ono  l990a),  (1990)  few studies have estimated the level of genetic  divergence among insular species of larger speciose)  and Crawford et al.  (i.e. more  monophyletic lineages.  The endemic Hawaiian species of Lysimachia were chosen for evaluation of genetic variation within and among insular congeneric species for two reasons.  First,  morphological differences among the species,  despite gross several  morphological characters suggest that the group is  127  monophyletic.  Secondly, the availability of morphometric  data, used to produce a taxonomic revision of the Hawaiian species, permits a comparison of morphological divergence with genetic divergence, as estimated from allozyine variation. studies have made such a direct comparison (Hamrick,  Few  1989).  Extensive ecological and morphological divergence have accompanied adaptive radiation of Lysimachia in the Hawaiian Islands.  Species grow in a diversity of habitats including  montane bogs, waterfall spray zones,  subalpine mesic  shrublands, montane dry and wet forests and lowland mesic shrublands from 250-2300 m elevation.  All are woody shrubs,  some with stems up to two cm in diameter. their herbaceous continental relatives,  This contrasts with  a change of habit that  is common in insular evolution (Carlquist, habit varies from upright to scandent, with stems up to 5 m long, vegetation.  1974).  Growth  or sometimes vinelike,  supported by the surrounding  The fruit is a dehiscent capsule,  dispersed locally by gravity.  thus seeds are  Founding distant new  populations is more problematical.  A reasonable vector of  dispersal is the strong winds that accompany hurricanes. Seeds stuck in the mud on the feet of birds is also a possible mode of dispersal. Of the 16 species and 4 subspecies, L. haupuensis St. John,  (.  forbesii Rock,  and j. kahiliensis St.  John) have not  3  been collected for 60 years or more and are probably extinct. The characters that are most useful to distinguish among species include leaf size and shape  (Figure 3.1), as well as  128  calyx and corolla size,  shape and pigmentation.  Most species  have a reddish corolla, however that of L. glutinosa Rock is white and  .  kalalauensis Skottsb.  is green.  Observations of  plants in the greenhouse indicate that the species are protogynous.  Seif-pollinations within flowers fail to set  fruit, however plants are self—compatible and geitonogamous pollination is possible. The extant high islands are the youngest of a 6,000 mile long chain of islands, the Emperor and Hawaiian Chain.  These  began forming over a hot spot in the Pacific plate at least 70 million years ago (Walker,  1990).  The older islands, now  converted mostly to coral atolls, were once high islands, undoubtedly with much greater habitat diversity than they presently contain,  and could have been the source of  propagules for dispersal to the extant high islands. reason,  For this  it cannot be assumed that Kauai was the site of the  original founding event or that species in any extant lineage have been evolving for less than 5.7 million years. another monophyletic group, Hawaiian Drosophila,  In  the drosophiloid lineage of the  it appears that the ancestor of the  Hawaiian species arrived in Hawaii 10 million years ago, using sequence data from the alcohol dehydrogenase locus Hunt,  1991)  (Thomas and  129  4.,  c  3.  2.  16.  14.  12.  $  18.  7.  17.  I  19.  13 Figure 3.1. Representative leaves of endemic Hawaiian Lysimachia by island: Kauai (1-11), Oahu (12-14), Molokai (15— 16), Maui (17—19). 1. L. glutinosa, 2. L. iniki, 3. L. kalalauensis, 4. L. ovoidea, 5. L. kahiliensis, 6. j. scopulensis, 7. L. venosa, 8. L. daphnoides, 9. L. filifolia, 10. L. pendens, 11. . haupuensis, 12. i. hillebrandii, 13. i. forbesii, 14. L. waianaeensis, 15. i.. maxima, 16. L. remyi subsp. subherbacea, 17. . remvi subsp. remyi, 18. j. remvi subsp. kipahuluensis, 19. . remyi subsp. caliginis.  130  Species of Lysimachia are distributed on all of the major islands with the exception of Hawaii, Kahoolawe and Niihau. The number of species per island correlates roughly with the age of the island  (dates from MacDonald et al.,  1990).  species grow on Kauai, the oldest extant high island million years),  Maui  one species and one subspecies on  (1.5—1.8 million years),  (1.3 million years),  on East Maui  and one species and two subspecies  11 species and 4 subspecies,  in the present study, these are: Hillebr,  .  Rock,  hillebrandii Hook F.  Skottsb.,  filifolia C.N.  L. maxima  L. pendens Marr,  L.  .  scopulensis Marr, venosa  (Wawra)  ex A. St.  renwi subsp.  remyi subsp.  L.  daphnoides  Gray, John,  subherbacea  L.  John and  .  L. glutinosa  ovoidea St. (St.  Marr, j.  (Hillebr.)  and j. waianaeensis St.  St.  Gray)  L. kalalauensis  caliginis  (St. John)  were included (A.  Forbes and Lydgate,  (R. Knuth)  remyi subsp. kipahuluensis remyi,  and three subspecies on West  (0.7—0.8 million years).  Fifteen taxa,  L.  (4.5-5.7  four species and one subspecies grow on Oahu  (2.1—3.6 million years), Molokai  Twelve  iniki Marr,  John)  John,  Marr, L.  remyi subsp. Marr,  John.  .  Lysimachia  could not be  included because their populations are too inaccessible.  The  taxa that were sampled represent most of the morphological variation of the genus with the exception of collections from Oahu.  Regrettably,  may be extinct: i. hillebrandii  two species from the Koolau Mtns., Oahu, forbesii  (last collected in 1934),  (last collected from the Koolau Mtns.  One population of L.  hillebrandii was sampled,  and  L.  in 1980).  but this came  131  from the Waianae Mtns. corolla,  The dimensions of the leaves,  and pollen of  calyx,  forbesii were the largest of any  species of Lysimachia in the world. Most taxa are endemic to a single island and are often restricted to a single ecological zone. from a single location.  Several are known  Populations are small and scattered,  typically consisting of from 10—100 individuals. Distributions are typically allopatric, with rare exceptions. One location of sympatry is on the steep slopes and cliffs of upper Kalalau Valley on Kauai.  Here the ranges of  distribution of L. glutinosa, L. kalalauensis and L. scopulensis overlap somewhat.  Hybrids between  and L. kalalauensis, and between  glutinosa and  .  scopulensis have been found in this area. populations of L. glutinosa and isolated from each other.  .  .  glutinosa .  Elsewhere  kalalauensis are quite  In the southern Waianae Mtns.  on  Oahu, plants of j. hillebrandii grow sympatrically with j.. waianaeensis in some populations.  Further field observations  are needed to determine whether these two species hybridize. In Haleakala crater on East Maui,  individuals in two  populations display the full range of morphological intermediacy between j.. remyi subsp. subsp. kipahuluensis, swarms.  caliginis and L. remyi  and are interpreted as being hybrid  Although the ranges of these two subspecies  apparently do not overlap at the present time, have in the past.  they may well  The structure of the vegetation in this  132  area was severely damaged by decades of damage from feral goats and pigs. An important question concerning studies of differences among species in general and insular adaptive radiation in particular is whether or not gross morphological differences are the result of changes at fewer loci than would be expected given the degree of morphological divergence evident (Helenurm and Ganders,  1985; Crawford et al.,  1987b).  Gottlieb (1984)  summarized literature on the genetic basis of morphological differences among species, structure,  shape,  and found that characters of  orientation and presence versus absence were  typically controlled by one or two genes.  Characters of  dimension, weight and number, were usually controlled by numerous genes.  Many of the characters that are often used to  distinguish one species from another fall within the first category.  The most direct means of assessing genetic control  of differences among species is,  of course, progeny analysis  of F 1 and F 2 hybrids as well as backcrosses.  Island taxa are  ideal for these types of studies because internal reproductive barriers are often lacking.  An alternative approach is to use  allozyme analysis as an indirect means of measuring genetic variation within and among populations or species by comparing the number and frequency of alleles that code for several different proteins.  133  3.1.1  Objectives of allozyme analysis  The objectives of this study were to:  1)  estimate the  degree of genetic variation within and among species using allozymes; 2)  to compare these results with published results  from other insular taxa,  as well as with continental taxa; 3)  to compare morphological divergence to allozyme divergence among a group of congeneric species that have undergone adaptive radiation; and 4)  to evaluate the taxonomic revision  proposed in Chapter 1.  3.2 3.2.1  Materials and Methods Sample collection. Young leaves were collected in the field from 1028  individuals in 48 populations  (Table 3.1).  Plants were  sampled randomly and whenever possible at least 30 plants per population were sampled.  Vouchers of populations collected  have been deposited at UBC.  Leaves were refrigerated until  they could be sent via express mail to UBC and stored at _800 C until immediately prior to enzyme extraction. one year,  Even after  enzyme activity of samples stored in this manner was  comparable to that of fresh leaves.  KAAPU KMAKA KWAIA  kalalauensis  kalalauensis  kalalauensis  L. kalalauensis  DSECO  daphnoides  daphnoides  j.  L.  FKTHR  L.  Kauai  Kauai  Kauai  Kauai  Kauai  Kauai  Kauai  Kauai  Kauai  Kauai  Kauai  Kauai  Table 3.1 continued on next page.  pendens  FKONE  L. pendens  DBIGB  NSPTR  scopulensis  L.  OWAIN  ovoidea  OLIMA  L.  .  .  .  ovoidea  KHONO  glutinosa  L.  .  GHONO  glutinosa  L.  Kauai  Kauai  GKALR  glutinosa  L. GKALL  Island  Pop. Code  Species  615  615  1230  1230  800  615  650  1110  1015  1140  1110  1230  1230  1260  Altitude (meters)  SW of Kulanaililia.  SE of trail. Head of N fork of Wailua River, south wall. Head of N fork of Wailua River, west wall.  Kokee; Alakai Swamp,  Kokee; north-south ridge below and E of Kalalau Lookout. Kokee; Alakai Swamp trail.  Wainiha Pali;  on ridge SW of Limahuli Falls.  Kokee; Waialae Ridge.  Kokee; North of Makaha Ridge road.  Kokee; Awaawapuhi trail.  Kokee; Edge of Honopu Valley.  Kokee; Honopu trail.  Kokee; Road between Kalalau and Puu 0 Kila Lookouts. Kokee; Below Kalalau Lookout.  Locality  Collection localities for populations of endemic Hawaiian Lysimachia sampled for Table 3.1 allozyme analysis.  I-I  1050 1075  W. Maui W. Maui W. Maui W. Maui W. Maui  HPUKE HWAIK HLUAF HLUAS WNBLG WMHEL WMMN2 WNNN3  WNNPP  waianaeensis  L. waianaeensis  L. hillebrandii  j. waianeensis  L. remyi subsp. kipahuluens is i.. remyi subsp. remyi L renwi subsp. remyi remvi subsp. remyi . remvi subsp. remyi  Oahu  Oahu  Oahu  Oahu  Oahu  Table 3.1 continued on next page.  .  HPUKS  860  1230  340  830  830  800  800  710  250  L. waianaeensis  Oahu  FOTHR  250  615  filifolia  Oahu  Kauai  Altitude (meters)  L.  FOTWO  FKFIV  Island  filifolia  pendens  Pop. Code  L.  .  Species  E of summit.  Manawainui Plant Preserve.  Hanaula.  Hanaula.  Helu,  Waianae Mtns.; Kamaileunu Ridge in Makaha Valley; N of small peak N of Puu Kawiwi. Waianae Mtns.; ridge between Lualualei and Nanakuli Valleys. Waianae Mtns.; ridge between Lualualei and Nanakuli Valleys. lao Valley; Black Gorge.  Waianae Mtns.; ridge E of Puu Kaua.  Waianae Mtns.; ridge SE of Puu Kaua.  Koolau Mtns.; Waiahole Ditch Trail.  Head of N fork of Wailua River, east wall. Koolau Mtns.; Waiahole Ditch Trail.  Locality  Table 3.1 continued. Collection localities for populations of endemic Hawaiian Lysimachia sampled for allozyme analysis.  Ridge SE of Paliku Cabin. Ridge SE of Paliku Cabin.  West Kaupo Gap,  Kalapawili Ridge; rim of Kipahulu Valley.  1260 985 1050 1000 2000 2090 2215 2280 2155 1785 2300  W. Maui W. Maui W. Maui W. Maui E. Maui E. Maui E. Maui E. Maui E. Maui E. Maui E. Maui  WNLIS WNHPL WMHPS WMHPY EMPAL EMHPA EMKUI EMKUK EMLWA EMKAW EMKPR  subsp.  subsp.  subsp.  L. remyi subsp. remyi remyi subsp. . kipahuluensis remyi subsp. . kipahuluensis j. remyi subsp. kipahuluensis j. remvi subsp. kipahuluensis L. remvi subsp. kipahuluensis remyi subsp. . kipahuluensis . remyi subsp. kipahuluensis  Table 3.1. continued on next page.  ridge SW of summit.  ridge SW of summit.  rim of Kipahulu Valley.  below spring.  Above Lake Waianapanapa.  Kuiki,  Ridge N of Kuiki.  Halepohaku,  Halepohaku summit.  Halepohaku,  Lihau summit.  ridge west of summit.  1130  W. Maui  WMLIL  subsp.  remvi remyi L. remyi remyi i. renwi remyi remyi . remyi  .  Lihau,  Altitude (meters)  Island  Species  Pop. Code  Locality  Collection localities for populations of endemic Hawaiian Lysimachia continued. Table 3.1. sampled for allozyme analysis.  treeline. below treeline.  Koolau Gap, Koolau Gap,  Kipahulu Valley, Palikea stream.  Makakupaia; ridge S of Onini gulch. Makakupaia; bottom of Onini Gulch. Waikolu; SE of Puu Kaeo.  1970 2920 1000 920 1040 970 850 1050 920 950  E. Maui E. Maui E. Maui Molokai Molokai Molokai Molokai Molokai Molokai Molokai  EMKOT EMKOB EMKIP MKAWE MKOLE MMAKA MONIN MWAIK MKAUN MMAXI  S of road.  Small ridge between Kauanakakai and Kupaia Gulches. Pelekunu; N of Ohialele.  W of Puu Kolekole,  West fork Kawela Gulch.  Koolau Gap, Waikau Cabin site.  2030  E. Maui  EMWAI  below Paliku Cabin.  Kaupo Gap,  1845  E. Maui  EMKAE  j. remyi subsp. kipahuluensis X j.. remvi subsp. caliginis L. remyi subsp. kipahuluensis X . remyi subsp. caliginis . remyi subsp. caliginis . remyi subsp. caliginis . remyi subsp. kipahuluensis remyi subsp. . subherbacea L. remyi subsp. subherbacea L. remyi subsp. subherbacea L. remvi subsp. subherbacea remyi subsp. . subherbacea L. remyi subsp. subherbacea L. maxima  Locality  Altitude (meters)  Pop. Code  Species  Island  continued. Collection localities for populations of endemic Hawaiian Lysimachia Table 3.1 analysis. sampled for allozyme  138  Wherever possible, were sampled.  at least two populations of each taxon  In some cases the limited number of populations  sampled is a reflection of their limited distribution. Lysimachia maxima is known from a single population and ovoidea is known from only two populations.  .  Lvsimachia  filifolia is known from one locality on Oahu and one on Kauai. The two populations from Oahu that were sampled are in adjacent small gulches separated by 30-40 meters.  Likewise,  L. pendens is known only from a large amphitheater approximately 200 meters across.  Populations were sampled  from different walls of this amphitheater.  Lysimachia remyi  subsp. kipahuluensis was collected from both East Maui arid West Maui.  Collections from East Maui are referred to as L.  remyi subsp. kipahuluensis as L.  (EM)  and collections from West Maui  remyi subsp. kipahuluensis  (WM).  Populations of this  subspecies from East Maui were analyzed separately from those from West Maui. swarms between L.  Two populations,  EMKAE and EMWAI, hybrid  remyi subsp. caliginis and  .  remyi subsp.  kipahuluensis, were excluded from calculations of genetic variation at the island level, however they were included in all other analyses.  3.2.2  Electrophoretic procedures Leaf tissue was ground in the 0.1 M Tris-HC1, pH 7.5  extraction buffer of Soltis et al.  (1983).  Extracts were  absorbed onto Whatman 3MM chromatography paper wicks.  Samples  were electrophoresed on 12.5% starch gels using three buffer  139  systems from Soltis et al.  (1983).  resolve isocitrate dehydgrogenase phosphoglucoisomerase malic enzyme  (ME),  (PGI),  System 2 was used to (IDH);  system 8 to resolve  triosephosphate isomerase  leucine amino peptidase  to resolve alcohol dehydrogenase  (LAP); and system 9  (ADH), malate dehydrogenase  (MDH),  shikimate dehydrogenase  (MNR),  6-phosphogluconic dehydrogenase  phosphoglucomutase (PGM)  (SKD), menadione reductase  and diaphorase  (6-PGD), (DIA).  The pH of each  gel buffer was changed from those of Soltis et al. follows:  (TPI),  system 2, pH=8.6; system 8,  pH=8.l;  (1983)  system 9,  as  pH=6.7.  Gels were electrophoresced in a refrigerated cabinet. System 2 was run for 4 hr at a constant voltage of l5OV, System 8 was run for 10 hr at 11OV,  at 190V.  Following electrophoresis,  System 9 was run for 6 hr  enzymes were visualized  according to the protocols of Soltis et al.  (1983), with the  exception of DIA and MNR which were visualized using protocols of Wendel and Weeden (1989). agarose overlays. “1”,  3.2.3  TPI and DIA were stained using  The most anodal isozyme was designated as  likewise, the most anodal allozyme was designated “a”.  Data Analysis Results were analyzed in such a way as to compare the  distribution of allozyme variation at several levels: within populations; 2) 3)  1)  among populations of the same species;  among species; 4) within islands;  5)  among islands.  140  3.2.3.1.  Analysis of allozyme diversity  Four measures of genetic variation within populations were calculated:  the number of alleles per locus  percent of loci polymorphic  (P)  polymorphic locus  (Ar),  (according to Nei,  the  (a locus was regarded as  polymorphic if there were two or more alleles, frequency of at least 0.01),  (A),  each with a  the mean number of alleles per  and the expected heterozygosity (H)  1987,  177,  P.  equation 8.1).  Measurements  of H are calculated from:  H=l-’X’ where m is the number of alleles, frequency of the ith allele. effect on the value of H.  and xi is the population  Very rare alleles have little  Maximum diversity occurs when all  alleles have equal frequencies.  Some hypothetical examples  illustrate the effect of the number and frequency of alleles on the value of H for the simplest case where only one locus is analyzed.  If only two alleles,  “a” and “b”,  equal frequencies of a=0.5 and b=0.5, 0.25= 0.5,  and H  then a 2 +  =  1-0.5  =  0.01 + 0.81  =  0.5. =  then a 2 + b 2  However,  0.82,  are present at  and H  0.25 +  =  if a=01 and b=0.9, =  1  0.82  -  =  0.18.  Now consider the effect on H as the number of alleles increases. +  0.111  =  If a=b=c=0.333, then a 2 + 0.333,  and H= 1— 0.333  =  0.667.  + C 2  =  0.111 + 0.111  If a=b=c=d=e=0.2,  the + + 2 = 0.04+0.04+0.04+O.04+O b d c e na .04 =  0.2 and H  =  1  —  0.2  =  0.8.  Gene diversity statistics Ht,  Dst,  (unbiased for sample size), H,  and G t were calculated to compare the distribution 5  141  of variation within and among populations of the same species. The measure of genetic diversity is H , 5 as H above).  Diversity among populations is 5 D t ,  the total genetic diversity, Gst,  (calculated the same and Ht is  therefore Ht=H 5 + Dst.  Finally,  the proportion of the total genetic diversity that is  distributed among populations of a species, GstDst/Ht.  is calculated as  This latter measure is also known as the  coefficient of gene differentiation. Nei’s genetic identities  (I)  for each pair of populations.  (Nei,  1972), were calculated  This statistic compares the  degree to which two populations share the same alleles at the same frequency.  When 1=1, the two populations share all  alleles at identical frequencies.  When 1=0,  they have no  alleles in common. Gene diversity statistics Nei’s genetic identities GENESTAT-PC 3.3  (Lewis,  (I)  (Nei and Chesser,  (Nei,  1993).  1972)  1983)  and  were calculated using  A phenogram based on UPGMA  (unweighted pair-group method with arithmetic mean)  clustering  of Nei’s genetic identities was produced using BIOSYS-1 (release 1.7)  3.2.3.2.  (Swof ford and Selander,  1989).  Principal Components Analysis  Principal components analysis  (PCA)  was used to  characterize and ordinate morphological variation and allozyme variation among populations.  The measurements used for PCA of  allozyme variation were the allele frequencies at each  142  allozyme locus from each population. two alleles were detected.  For ADH and DIA,  To avoid redundancy,  only  the  frequencies of the alternative alleles, Adh—la and Dia—ib were not included.  The eight vegetative and eight reproductive  characters that were used in Chapter Two  (page 25),  in the PCA of morphological characters.  Population means of  were used  the factor scores were used in all subsequent analyses. Pearson product—moment correlation coefficients were calculated to compare the ordination of populations based on allozyme analysis to that based on morphological characters. SYSTAT  (Wilkinson,  1990)  was used to compute the PCA’s and the  Pearson correlation coefficients.  143  3.3.  Results  3.3.1.  Electrophoretic Patterns  Eleven isozyme loci, coding for the following eight enzymes could be scored in all populations: three for TPI, two for MDH,  and one each for SKD, DIA, PGM, ADH,  IDH,  and PGI.  Inconsistent staining or poor resolution of Mnr—i, Pgm—2, Pgi— 2, Mdh-i,  Mdh-2, McTh-5, Lap-i, Me,  6-Pgd-1,  precluded their inclusion in the analysis.  and 6-Pgd-2 Interpretation of  the genetic basis of the banding patterns was based on the number of isozymes reported for diploid plants Wendel,  1989)  and analysis of F 1 hybrids.  (Weeden and  Nearly all  individuals displayed three,  five or seven bands at the two  most anodal isozymes of MDH,  and the cathodal isozyme of TPI.  Because these are dimeric enzymes, the phenotype of diploid homozygous plants should be one band; three bands would be expected for a heterozygous plant.  A phenotype of five and  seven bands requires the presence of three and four alleles respectively, duplicated.  indicating that the gene for that locus is Plants were scored in accordance with this  explanation. Allele frequencies for each population are reported in Appendix 1,  Table A3.1.  All of the 11 loci that could be  scored were polymorphic, as were the 10 that could not be consistently scored.  All species shared the same highest  frequency allele at four loci, Adh-i, Mdh-4, Skd-i, Dia-i. Most species shared the same highest frequency allele at the remaining loci as well.  However, at a few loci of some  144  species, especially those on Kauai, there was a different highest frequency allele.  The geographical and taxonomic  distribution of loci whose highest frequency allele differed from the most common allele (compared to all other species)  is  presented in Table 3.2. The following “private alleles”, those that were detected in a single species, were not necessarily diagnostic of the species they occur in because they had very low frequencies: Pgm-la,  in  .  lb in  Pgm-lb and Pgi-ld in  .  daphnoides, Idh-le and  kalalauensis, Mdh-4c and Pgm-lf in .  waianaeensis.  .  ovoidea, and Adh  None were present in all individuals  of a population or even in all populations of a species. Some plants,  especially  zone of stain for MDH, Mdh-5.  .  glutinosa, had an additional Evidently, at this locus there  is a null allele in some species because no stain was detected in many plants. involving  .  In the progeny of some hybrid crosses  glutinosa two bands were detected at Mdh-5.  These were interpreted as the homodimer of the protein contributed by L. glutinosa and the heterodimer between the protein of the allele from L. glutinosa and that of the null allele contributed by the other species.  This locus could not  be scored because of inconsistent staining as well as the possibility that more than one null allele exists.  145  Table 3.2. Geographic and taxonomic distribution of the highest frequency alleles for loci at which the allele with the highest frequency is different from the most common allele (based on comparison to all other species), for that locus, in taxa of Hawaiian Lysimachia. Allele  Taxon  Island  Pgi-la Idh-la Idh-ld Idh-lb Pgm-2e Tpi-la Idh-lb Pgi—la  L. glutinosa L. glutinosa L. glutinosa L. ovoidea ovoidea . j. scopulensis L. pendens L. daphnoides  Kauai Kauai Kauai Kauai Kauai Kauai Kauai Kauai  waianaeensis L. waianaeensis waianaeensis .  Oahu Oahu Oahu  Tpi—2a Idh-lb Pgm—2e Tpi-3a Idh-la Idh-lb Mdh-3b Mdh-3b  .  remyi subsp. remyi subsp. j. subsp. remvi j. .  . .  subherbacea subherbacea subherbacea  remyi subsp. kipahuluensis remyi subsp. caliginis  Molokai Molokai Molokai Maui Maui  146  Malic enzyme displayed two closely spaced zones of stain. This was difficult to interpret because ME is a tetrameric enzyme with one isozyme present in diploid plants Wendel,  1989).  single band.  (Weeden and  The phenotype of homozygous plants should be a Heterozygous individuals should display five  bands, the two homotetraxaers and three heterotetramers. Duplication of the locus for ME should result in a complicated pattern of at least three interlocus heterotetramers in addition to the two homotetramers in individuals that are homozygous for different alleles at the different loci.  The  phenotype of individuals heterozygous at one or both of the duplicated loci would be even more complicated.  The lack of a  suitable explanation for the banding pattern, prevented the inclusion of this enzyme in the analysis, however at least two alleles were detected,  one apparently unique to  .  glutinosa.  In some cases the banding pattern of Dia—1 on gels of system 8 differed from that seen using system 9 gels, but was the same as that seen for MNR (when stained on system 9). example,  For  some plants that appeared heterozygous for DIA using  system 9,  appeared homozygous when stained using system 8.  similar observation has been made in Daucus carota and Wricke,  1991).  Wendel and Weeden  (1989)  (Westphal  report that the  stains for DIA and MNR are not necessarily specific for one enzyme;  i.e.  in some species the same enzyme is visualized  using the different stains, while in others, the enzyme that is visualized by the stain for DIA differs from the enzyme  A  147  visualized by MNR.  In order to be consistent, Dia—1 was  always scored using gels from buffer system 9.  3.3.1.1.  Allozyme variation within and among populations of  species  Estimates of genetic variation within populations are presented in Table 3.3. (A),  The mean number of alleles per locus  ranged from 1.1 for a popi.lation of J. remvi subsp.  kipahuluensis to 2.1 for a population of remyi. of  .  .  remvi subsp.  (The low value of 1.0 from a population of five plants pendens undoubtedly reflects the small sample size.)  The mean number of alleles per polymorphic locus  1 ranged (A)  from 2.0 for populations of many species, to 3.0 for a population of L. glutinosa. (P), L.  The percent of polymorphic loci  ranged from 9% for one population each of j. glutinosa,  remyi subsp. kipahuluensis,  and  .  remvi subsp.  to 72% for one population of j. waianaeensis. heterozygosity (H), remyi subsp.  The expected  ranged from 0.01 for populations of  .  caliginis and L. remyi subsp. kipahuluensis, to  0.24 for one population each of and  caliginis,  waianaensis.  .  remyi subsp.  subherbacea  148  Table 3.3. Genetic variability in 48 populations of endemic Hawaiian Lysimachia. Mean number of individuals scored for each enzyme (N); mean and standard error for number of alleles per locus (A); mean number of alleles per polymorphic locus (Ap): percentage of polymorphic loci (P); mean and standard error for expected heterozygosity (H); number of unique alleles (U). ISLAND Species Pop.  N  A±s.d.  Ap  P(%)  H±s.d.  U  KAUAI L. daphnoides DBIGB 24.6 DSECO 8.0  1.9±. 3 l.5±.2  2.7 2.3  54 36  0.135±.052 0.116±.062  2 0  mean  1.6  2.5  45  0.126  L. glutinosa GKALR 18.8 GKALL 17.5 GHONO 28.4  l.2±.2 l.7±.3 l.6±.3  3.0 2.7 2.7  9 36 36  0.062±. 062 0. l18±.059 0.l08±.053  mean  1.5  2.8  27  0.096  kalalauensis KHONO 25.1 KAAPU 25.0 KMAKA 9.3 KWAIA 22.5  l.6±.2 l.6±.2 1.2±.1 1.5±.2  2.4 2.2 2.0 2.0  45 54 18 54  0.122±. 0.140±. 0. 075±. 0.130±.  mean  1.5  2.1  ovoidea OLIMA 8.4 OWAIN 37.6  1.5±. 2 1.5±. 2  mean  16.3  21.57  .  0 0 0  0 0 0  054 059 052 051  2  43  0.117  2.0 2.0  45 54  0.169±.062 0.149±.058  0 2  1.5  2.0  50  0.159  L. scou1ensis NSPTR 7.2  l.4±.2  2.0  36  0.135±.058  0  j. pendens FKONE 12.2 FKTHR 6.5 FKFIV 5.1  1.4±. 2 1.2±. 1 1. 0±.0  2.0 2.0  36 18 0  0. 077±. 035 0. 061±.042 0. 000±. 000  0 0 0  mean  1.2  1.3  27  0.046  20.5  .  Table 3.3.  23.0  7.9  continued on next page.  149  Table 3.3. continued. Genetic variability in 48 populations of endemic Hawaiian Lysimachia. ISLAND Species Pop.  A±s . d.  Ap  P(%)  H±s.d.  OAIU-WAIANAE 3:. waianaeensis HPUKE 28.1 HPUKS 8.5 HWAIK 23.2 HLUAS 12.5  1.9±.3 1.5±.2 1.5±.2 1.5±.2  2.3 2.0 2.0 2.7  72 45 54 27  0.239±. 0. 191± . 0.173±. 0.171±.  mean  1.6  2.2  L. hillebrandii HLUAF 21.5  1.5±.2  OAHU-KOOLAU L. filifolia FOTWO 21.8 FOTHR 21.4 mean  N  18.1  21.6  MOLOKAI L. maxima MMAXI 27.8  L!.  remyi subsp. MWAIK 26.6 MKOLE 22.2 MONIN 22.9 II4AKA 26.5 MKAWE 19. 9 mean  23.6  WEST MAUI !. remyi subsp. WMBLG 11.1  Table 3.3.  U  065 072 062 089  1 0 0 0  50  0.194  2.0  54  0.132±.056  0  1.3±. 1 1. 3±.2  2.0 2.5  27 18  0.051±.032 0.088±.060  0 0  1.3  2.3  22  0.070  1.5±. 2  2.2  45  0.148±.065  0  subherbacea 1.7±.2 1.5±.2 1.5±.2 l.7±.3 1.6±.2  2.3 2.2 2.2 2.6 2.2  54 45 45 45 54  0.223±. 068 0. 167±.073 0. 2 13±. 075 0.219±. 079 0.238±. 077  0 0 0 0 0  1.6  2.3  49  0.212  kipahuluensis 1.4±.2 2.0  36  0.087±.043  continued on next page.  0  150  Table 3.3. continued. Genetic variability in 48 populations of endemic Hawaiian Lysimachia. ISLAND Species Pop.  N  remvi subsp. WNHEL 41.4 WMHPL 29.1 WNHPY 9.9 WMNN2 28.1 WMMPP 27.2 WNHPS 24.6 WNLIS 54.5 WNLIL 16.1 WMLIM 11.2 mean  26.9  EAST MAUI remyi subsp. EMPAL 26.3 EMHPA 23.0 EMLWA 27.6 EMKPR 26.7 EMKAW 13.6 EMKIP 22.0  L.  mean  23.2  k. remvi subsp. EMKOT 29.2 ENKOB 9.0 mean  19.1  A±s . d. remvi 1.8±.2 l.8±.3 1.5±.2 2.0±.3 1.6±.2 l.5±.2 2.1±.3 1.7±.3 1.3±.l 1.7 kipahuluensis l.4±.2 l.1±.1 1.4±.2 l.4±.2 1.4±.2 1.3±.1 1.3 caliginis l.2±.l l.l±.l 1.1  Ap  P(%)  H±s.d.  U  2.3 2.3 2.0 2.7 2.2 2.2 2.5 2.3 2.0  63 63 45 63 54 45 72 54 27  0.130±. 056 0.115±.045 0. 106±.052 0. 162±. 050 0.139±. 050 0. 089±. 037 0.144±. 042 0. 142±. 055 0.055±. 033  0 0 0 0 0 0 0 0 0  2.3  54  0.120  2.0 2.0 2.0 2.0 2.0 2.0  36 9 36 36 36 27  0.075±.042 0.010±.010 0.l13±.053 0.078±.039 0.048±.023 0.064±.046  2.0  30  0.065  2.0 2.0  18 9  0.031±.026 0.009±.009  2.0  13  0.020  j. remvi subsp. kipahuluensis X L. (populations of hybrid swarms) EMKAE 28.5 1.5±.2 2.0 EMWAI 28.2 1.4±.2 2.0  remyi subsp. 54 36  0.127±.055 0.055±.031  mean  28.4  1.5  2.0  54  0.091  All pops.  21.4  1.5  2.2  40  0.117  25  .064  Is land endemics (Dejoode and Wendel, 1992)  1.32  0 0 0 0 0 0  0 0  caliciinis 0 0  151  Mean species values for subsp.  (A)  ranged from 1.1 for j. remyi  caliginis to 1.7 for L. remvi subsp. remyi;  from 2.0 for glutinosa; to 54% for  varied from 13% for  .  L. remyi subsp. remvi, and  varied from 0.02 for remyi subsp.  varied  remvi subsp. kipahuluensis to 2.8 for j.  .  (P)  (Ar)  remvi subsp.  caliginis  hillebrandii; and  .  (H)  L. remvi subsp. caliciinis to 0.21 for L.  subherbacea  (Table 3.3).  Graphical presentation (Figure 3.2)  of genetic variation  within populations emphasizes the fact that there is not necessarily a stepwise decrease in the level of heterozygosity from the oldest through to the youngest island,  although  populations from East Maui are the least heterozygous. fact,  some populations from West Maui,  variable,  if not more so,  In  Oahu and Molokai are as  than populations from Kauai.  Genetic diversity statistics are presented in Table 3.4. Total diversity was lowest in j.. (Ht=0.02),  and highest in  .  remyi subsp.  remyi subsp.  caliginis  subherbacea  (Ht=0.23); diversity within populations was lowest in j.. subsp.  caliginis  subherbacea  (H = 5 O.O2)  (H = 5 0.2l).  and  .  (Dst=O.O3l),  remvi subsp.  lowest among populations of L. (D t 5 O.OO4)  and  .  remvi subsp.  Variation among populations was  highest in j.. waianaeensis (DstO.Ol4)  and highest in j.  remyi  remyi subsp.  kalalauensis  .  subherbacea  (Dst=O.Ol8)  and  remvi subsp. kipahuluensis caliqinis  (Dst=0.OOl).  3 0 C)  2  I  00 .  0 80  I: 70  .40 30 0 -20 10  0,05 0.00  00  00  I  0  080  I  I  I  K  OW  OK  M  I  I  I  0  00  0  I  I  EM  A  E  I  I  I  I  I  I  00  0  000000  00000  •  0 00  0  0  0  0  0  0  0  0  00  I  I  K  OW  OK  M  I  0 00  oo  0  000000  I  -  0  000  00  :  0  0  000000  -  I  WM  .  •  I  oo  .  -  I  8 o8ooooo o  I  I  0.25  0.10  I  0  •  0  0.20  I  8  O08880OOO  1  I  .  .  >,  I  I  I  WM  EM  A  E  I  I  I  I  I  I  0 0  0  00 00  I  0  8  0  00  0  0  0 0 0 0 0  I  I  I  I  I  K  OW  OK  M  WM  EM  A  I  I  E  I  Figure 3.2. Genetic variation by island for populations of endemic Hawaiian Lysimachia are compared to average values for species of various geographical distributions. Islands are arranged in decreasing age from left to right: K, Kauai; OW, Oahu-Waianae Mtns.; OK, Oahu-Koolau Mtns.; M, Molokai; WM, West Maui; EM, East Maui. From Hamrick and Godt(1990): A, average of 449 species; E, average of 81 endemic species. From DeJoode and Wendel(1992): I, average of 62 insular endemic species.  152  153  Table 3.4. Nei’s genetic diversity statistics calculated for 11 species and four subspecies of endemic Hawaiian Lysimachia. Values are means calculated from 11 loci. Ht=total diversity; =diversity within populations; Dt=variation among 5 H populations; Gt=proportion of variation among populations. ISLAND Species  No. of Pops. Ht  KAUAI L. daphnoides L. glutinosa L. kalalauensis L. ovoidea L. scopulensis L. pendens  2 3 4 2 1 3  0.129 0.103 0.132 0.166 0.135 0.049  0.126 0.096 0.117 0.159 0.135 0.047  OAHU-WAIANAE MTNS. L. waianaeensis L. hillebrandii  4 1  0.255 0.132  0.194 0.132  OAHU-KOOLAU MTNS. L. filifolia  2  0.076  0.069  1 5  0.146 0.230  0.149 0.212  1  0.087  0.087  9  0.131  6  MOLOKAI L. maxima remyi subsp. subherbacea WEST MAUI I!. remvi subsp. kipahuluensis . remyi subsp. remyi EAST MAUI j. remyi subsp. kipahuluens is . remvi subsp. calicfinis remvi subsp. . kipahuluensis X j. remvi subsp. caliginis mean for each taxon all populations Island Endemics (DeJoode and Wendel,  5 H  t 5 D  Gst  0.003 0.007 0.014 0.008  0.022 0.068 0.109 0.046  —  —  0.002  0.041  0.031  0.140  —  —  0.006  0.083  —  —  0.018  0.077  0.120  0.011  0.082  0.069  0.065  0.004  0.059  2  0.021  0.020  0.001  0.050  2  0.102  0.091  0.011  0.109  0.119 0.203  0.112 0.118  0.010 0.086  0.073 0.421  1992)  0.064  154  Compared to the other species, greater geographical distances separate populations of j..  remyi subsp.  .  kalalauensis, L. waianaeensis and  subherbacea.  This may have contributed to the  higher values of Dst for these three species.  The percent of  the total variation that is partitioned among populations, G t 5 , varied from 4.1% for  .  pendens, to 14% for L.  waianaeensis. There is a stepwise decrease in the levels of genetic divergence among populations within a single island from the oldest through to the youngest islands as seen in the values for Dst and Gst in Table 3.5.  In other words, populations on  Kauai have diverged the most from each other,  followed by  Oahu, Molokai and Maui.  3.3.1.2.  Allozyme variation within and among species  Mean intra-taxon coefficients of Nei’s genetic identity  varied from 0.927 among populations of 0.998 for i.  renwi subsp.  caliginis  .  waianaeensis, to  (Table 3.6).  The range of  variation was greatest among populations of L. waianaeensis (0.877—0.958) (0.995-0.997).  and least among populations of i. pendens With the exception of populations of  .  waianaeensis, all intraspecific values for I were greater than 0.930.  155  Table 3.5. Genetic diversity statistics of endemic Hawaiian Lvsimachia for each island. Values are means calculated from 11 loci. =diversity within populations; 5 Ht=total diversity; H t=variation among populations; Gst=proportion of variation 5 D among populations. ISLAND  Hs  Ht  Dst  Gst  KAUAI OAHU—WAIANAE OAHU—KOOLAU MOLOKAI WEST MAUI EAST MAUI  O.107±.03 0.18l±.06 O.069±.04 0.201±.06 O.117±.03 0.054±.03  O.207±.06 O.223±.07 0.076±.04 0.229±.07 0.128±.04 O.061±.03  0.100 0.041 0.006 0.028 0.011 0.007  0.482 0.186 0.083 0.121 0.085 0.114  156  Table 3.6. Identities  Intraspecific variation in Nei’s Genetic (I) for species of endeniic Hawaiian Lysimachia.  Species  No. of Populations  L.  cilutinosa  3  L.  daphnoides  2  L.  kalalauensis  4  L.  ovoidea  2  L.  scopulensis  1  L.  pendens  3  .  filifolia  2  L.  waianaeensis  4  L. hillebraridii  1  Range of I  Mean I  0.983—0.998  0.989 0.993  0.972—0.988  0.978 0.982  0.995—0.997  0.996 0.989  0.877—0.958  0.927  9  0.930—1.000  0.978  5 j. remyi subsp. kipahuluensis (East Maui)  0.986—0.999  0.995  remyi subsp. 1 kipahuluensis (West Maui)  .  .  remvi subsp.  remyi  j. reinyi subsp. kipahuluensis X L. subsp. ca1icinis  2  0.976  remyi subsp. caliginis  2  0.998  j. renwi subsp. subherbacea  5  L. maxima  1  .  remyi  0.943—0.992  0.972  —  —  157  The lowest mean inter—taxon genetic identity was for the pairwise comparison of L.  .  remyi from East Maui  glutinosa to the two subspecies of  (1=0.744 and 0.745)  highest pairwise comparisons were for remyi subsp. kipahuluensis and L.  kalalauensis  remyi subsp.  remyi  (WM)  (1=0.974),  The lowest value was for j. .  kalalauensis and  .  (1=0.986), and for j.  for j..  .  filifolia  filifolia and j.  renwi,  I ranged from 0.796-1.000.  renwi subsp.  subherbacea from  remyi subsp. remyi from West Maui,  highest for subspecies j. subsp.  The  (1=0.980).  Among subspecies of i.  Molokai and  (Table 3.7).  remvi subsp.  and the  caliginis and L.  remyi  kipahuluensis from East Maui.  Within islands, the greatest range of genetic identities was among populations on Kauai  (1=0.743-0.998).  The range of  genetic identities within an island is greatest on Kauai and decreases on progressively younger islands, with the least divergence among populations on East Maui (Table 3.8).  (1=0.965-1.000)  The greatest divergence among islands is between  populations from East Maui and those on Kauai  (1=0.707-0.956).  Using the mean genetic identities calculated for each pair of taxa, Kauai,  the mean I among taxa on each island are as follows: 1=0.858; Oahu,  1=0.941.  1=0.924; Molokai,  1=0.933; Maui,  ovoidea  scopulensis  scandens  filifolia  waianaeensis  hillebrandii  L.  L.  L.  L.  L.  L.  L.  4.  5.  6.  7.  8.  9.  10.  .  Table 3.7.  11.  0.853 (0. 846—0. 858) 0.918 (0. 906—0. 93 1) 0.840 (0. 805—0. 886) 0.934 (0. 934—0. 934) 0.863 (0. 863—0. 863) 0.858 (0. 820—0. 892) 0.938 (0.938—0.938) 0.929 (0.873—0.983) 0.902 (0.845—0.974) 0.976 (0. 924—10. 000)  0.818 (0. 764—0.876)  0.888 (0. 861—0. 903)  5  0.862 (0. 828—0. 895) 0.906 (0.888—0.926) 0.888 (0.843—0. 929) 0.910 (0. 847—0. 979) 0.905 (0.884—0.926) 0.850 (0.818—0.882) 0.945 (0. 926—0. 971) 0.931 (0.910—0.953)  4  0.910 (0. 855—0. 951) 0.949 (0. 924—0. 986) 0.903 (0.877—0.954) 0.974 (0. 961—0.987) 0.901 (0.842—0.966) 0.963 (0.954—0.982) 0.928 (0.922—0.933) 0.921 (0. 857—0. 973) 0.986 (0. 980—0. 999)  3  0.823 (0. 798—0. 864) 0.774 (0. 743—0. 798) 0.766 (0. 752—0. 775) 0. 847 (0.830—0.870) 0.827 (0.811—0.839) 0. 797 (0.763—0.855) 0.841 (0. 830—0.851) 0.779 (0. 770—0. 794) 0.810 (0. 772—0. 840) 0.853 (0. 840—0. 862)  2  of endemic Hawaiian Lysimachia. for species Mean (Minimum-Maximum)  0.926 (0. 901—0.948) 0.888 (0. 879—0.901) 0.829 (0.789—0.865) 0.828 (0. 825—0. 832) 0.812 (0.797—0.827) 0.893 (0. 881—0. 901) 0.812 (0. 767—0.859) 0. 872 (0. 867—0.876) 0.852 (0. 847—0. 858) 0.838 (0.802—0.871) 0.899 (0. 898—0. 899)  (I)  continued on next page.  remyi subsp. subherbacea 12. . remyi subsp. kipahuluensis (West Maui) 13. . renwi subsp. remyi  maxima  kalaluaensis  L.  3.  daphnoides glutinosa  L.  1  Genetic identities  1. 2.  Species  Table 3.7.  (7) 0,  1  Genetic identities  (I)  Table 3.7.  continued on next page.  14. j. remyi subsp. 0.819 (0.796—0.848) kipahuluensis (East Maui) 15. j. remyi subsp. 0.809 (0.804—0.811) caliginis 16. j. remyi subsp. 0.827 (0.800—0.861) caliginis X remvi subsp. . kipahuluensis  Species  Table 3.7.  0.910 (0.885—0.950) 0.901 (0.887—0.927) 0.918 (0.888—0.956)  0.745 (0.726—0.759) 0.755 (0.729—0.775)  3  0.744 (0.707—0.771)  2  0.847 (0.838—0.866) 0.839 (0.835—0.842) 0.854 (0.833—0.875)  0.801 (0.769—0.831) 0.832 (0.760—0.902)  5  0.835 (0.768—0.888)  4  for species of endemic Hawaiian Lysimachia. Mean (Minimum-Maximum)  (-n  I-I  0.899 (0.883—0. 913) 0.915 (0.894—0. 942)  0.813 (0. 802—0.826) 0.825 (0. 807—0. 844)  Table 3.7. continued on next page.  16.  15.  14.  13.  0.980 (0.923—10.000) 0.908 (0. 869—0. 938)  remvi subsp. subherbacea remyi subsp. . kipahuluensis (West Maui) i:• remyi subsp. remyi L. remvi subsp. kipahuluensis (East Maui) j. remyi subsp. caliginis j. remyi subsp. caliginis X I.!. remyi subsp. kipahuluens is  11.  maxima  hillebrandii  0.896 (0. 83 0—0. 947) 0.818 (0. 786—0. 840)  L.  10.  12.  L.  9.  .  0.900 (0.854—0.950) 0.965 (0. 958—0. 973) 0.973 (0.959—0.986) 0.931 (0.892—0.970) 0.985 (0.978—0.993)  L.  8.  waianaeensis  0.895 (0. 886—0.904) 0.908 (0. 877—0. 941) 0.944 (0. 940—0. 947) 0.853 (0. 848—0. 861) 0.932 (0. 888—0. 956) 0.947 (0. 945—0. 949)  L.  7.  filifolia  7  0.829 (0. 785—0. 898) 0.845 (0.780—0.905)  0.901 (0. 834—0. 967) 0.844 (0.787—0.918)  0.883 (0. 878—0. 889) 0.897 (0. 881—0. 912)  0.958 (0.913—0.974) 0.889 (0.871—0.909)  0.918 (0.918—0.918) 0.929 (0. 900—0. 962) 0.986 (0.986—0.986)  9  0.898 (0. 893—0. 904) 0.910 (0.904—0.915)  0.949 (0. 909—0. 983) 0.909 (0.876—0.937)  0.933 (0. 905—0. 964) 0.942 (0. 942—0. 942)  10  of endemic Hawaiian Lysimachia.  0.909 (0.877—0.949) 0.883 (0. 842—0. 946) 0.919 (0.836—0.961) 0.921 (0.887—0.962)  8  (I) for species Mean (Minimum-Maximum)  Genetic identities  6  continued.  Species  Table 3.7.  0  16.  15.  14.  13.  12.  .  0.925 (0.845—0.985) 0.873 (0.796—0.947) 0.851 (0.797—0.910) 0.871 (0.792—0.945)  0.981 (0.935—0.998) 0.912 (0.897—0.934) 0.902 (0.898—0.906) 0.919 (0.900—0.938)  12  0.992 (0.970—1.000) 0.988 (0.965—0.996)  14  0.989 (0.979—0.997)  15  of endemic Hawaiian Lysimachia.  0.932 (0.884—0.978) 0.925 (0.892—0.979) 0.940 (0.891—0.978)  13  (I) for species Mean (Minimum-Maximum)  Genetic identities  0.951 (0.921—0.982)  11  continued.  remyi subsp. kipahuluensis (West Maui) . remyi subsp. remyi . remvi subsp. kipahuluensis . remyi subsp. caliginis remvi subsp. caliginis X remvi subsp. . kipahuluensis  Species  Table 3.7.  0  0.959 (0.905—0.992) 0.931 (0.845—0.985) 0.874 (0.796—0.947)  0.938 (0.892—0.986) 0.981 (0.923—1.0) 0.906 (0.869—0.938)  0.915 (0.834—0.986) 0.851 (0.780—0.918)  0.906 (0.764—1.0)  0.830 (0.707—0.950)  Wmaui  Emaui  Molokai  0.916 (0.836—0.962)  0.903 (0.811—0.987)  Koolau (Qahu)  0.920 (0.877—0.958)  Molokai  0.884 (0.770—0.973)  0.876 (0.763—0.982)  Waianae (Oahu)  Koolau (Oahu)  0.989 (0.989—0.989)  0.880 (0.743—0.998)  Kauai  Wai’anae (Oahu)  0.913 (0.854—0.973)  Kauai  ISLAND  Mean (Minimum-Maximum)  0.928 (0.884—0.979)  0.979 (0.930—1.000)  West Maui  0.992 (0.970—1.)  East Maui  Nei’s genetic identities (I) within and among islands for endemic Hawaiian species Table 3.8. Islands are arranged from oldest to youngest (left to right, top to bottom). of Lysimachia.  163  The phenogram produced using UPGMA delineates two distinct groups: and L.  daphnoides;  (Figure 3.3)  in group one are j. glutinosa  in group two are the remaining species.  Lysimachia glutinosa and L.  daphnoides cluster together and  have lower genetic identities in pairwise comparisons with all other species because all individuals of L. of L. daphnoides are fixed for Pgi-la. daphnoides also carry Pgi-ld, species.  glutinosa and most  Some individuals of  .  an allele unique to this  All other species are fixed for Pgi—lb  (Pgi—lc was a  very rare allele in two populations).  Principal Components Analysis  3.3.2  Allele frequencies and mean measurements of floral and vegetative characters were available for 35 populations.  The  same character codes that were used in Chapter Two were also used here.  Fewer populations were included here than in  Chapter Two, identical,  3.3.2.1  therefore the results of the PCA are not  however the overall pattern remains the same.  Principal Components Analysis of floral characters  Component loadings for the PCA of floral characters are presented in Table 3.9. for PEDLENM, FILLENM.  CORLENM,  The first component had high loadings  CORWIDM,  CALLENM,  CALWIDM STYLENM,  and  The second component had high loadings for CALPUBE.  The first two axes represent 79% of the variation.  164 Figure 3.3. UPGMA phenogram derived from Nei’s genetic identities of 48 populations of 11 species and four subspecies of endemic Hawaiian Lvsimachia. See Table 3.1 for population codes. Cophenetic correlation coefficient is .798. .80  83  .87  Nei’s Genetic Identity .90  .93  .97  1.00 1.02 DBIGB daphno ides DSECO daphnpides GKALR a1utinoe GKALI. cilutinosa  GHONO cilutinosa KHONO )calalauensis WMBLG 9yj subsp.  kioahuluensis  WMLIS 9 j subsp.  .ubsp.  j  WMKPP 9 j subsp.  iay.i  WMMN2  Yorwo filifolia subsp.  WMLIK  YOTHR filifelia WMMPL  twny.i subap. ri  WNLIL  j subsp. 5  WNEPS pj subsp. KAAPO )calalauensis  KMAX1 kalalpuensis ENSPY 59 j subsp. KWkTh kalalauensis HLOkP hillebrandjj NXkXI  AJ.EA  WNNEL  subsp.  )1SPR scopulensis HPUKS waianaeensis ENPAL  pyJ subsp. kioshuluensis  EMKPR  rjyj subsp.kioahuluensis  ENLWA  subsp. )cioahuluensis  EMKAE  pjj subsp. kioahciluensis  EMEIP  subsp. kiciahujuensis  ENSPJL  rjyJ subsp. kioshuluensis  EXXOB 5pj  subsp. calipinis  E)1XAW  Li subsp. caliginis  ENKOl’  oixi subsp. caliginis  X  subsp. kiahu1uensis  .  EMWAI 1 j subsp. caliojnip X r. subsp. kiciahuluensis OLIN?. Ovoidea  OVAlS Ovoidea SWAIE pjjj subap. subherbacea SKAWE pj subsp. subberbacea MONtH  subsp. subberbacea  KNAKA  subsp.  subherbacea  MKOLE  subsp.  subherbaCea  FKONE Dendens FKIHR pendens ?KFIV cienclens SPOKE waianaeensis 5t0A8 wajanaeensjs  HWAIK wajappeensis  165  Table 3.9. Principal components analysis of floral characters of endemic Hawaiian Lysimachia averaged over 35 populations. Variable  Component Loadings 1 2  CALPUBE PEDLENM CORLENM CORWIDM CALLENM CALWIDM STYLENM FILLENM  0.092 0.751 0.963 0.902 0.881 0.852 0.853 0.870  —0.971 0.032 0.026 0.022 0.017 0.269 -0.184 -0.077  5.299 66.235 66.235  1.058 13.226 79.461  Eigenvalues % Variance % Cum. Var.  Table 3.10. Principal components analysis of vegetative characters of endemic Hawaiian Lysimachia averaged over 35 populations. Variable ULPUBE LLPUBE LEAFLM LEAFWM PETLM MINNTRND MAXNTRND STDNTRND Eigenvalues % Variance * Cum. Var.  Component Loadings 1 2  3  0.128 0.060 0.852 0.890 0.805 0.666 0.695 0.478  0.664 0.764 —0.275 —0.207 —0.235 —0.289 0.509 0.661  0.609 0.480 0.078 0.100 0.288 0.160 —0.473 —0.567  3.342 41.777 41.777  1.977 24.707 66.484  1.271 15.893 82.377  166  3.3.2.2  Principal Components Analysis of vegetative  characters Component loadings for the PCA of vegetative characters are presented in Table 3.10. loadings for LEAFLM,  LEAFWM,  The first component had high PETLM.  high loadings for ULPUBE, LLPUBE,  The second component had  and STDNTRND.  The first two  axes represent 66% of the variation.  3.3.2.3  Principal Components Analysis using allele  frequencies The population frequencies of 33 alleles were included in this analysis.  The following alleles had relatively high  loadings for the first component: Idh-lc, 3c,  Tpi-2a and Tpi-2c  (Table 3.11).  Idh-lb, Mdh-3b, Mdh  Alleles with high  loadings for the second component were: Pgi-la, and Pgm-le. variation.  Pgi-lb, Pgm-lc  The first two components represent only 27% of The first 11 components have eigenvalues greater  than one and represent 82% of the variation.  3.3.2.4  Comparison of morphological and allozyme results  The ordination of populations differs depending on which characters were used in the PCA (Figure 3.4).  The Pearson  correlation coefficient calculated from the comparison of vegetative characters and allozyine data is 0.489 and the comparison of floral characters to allozyme data is 0.372 (Table 3.12).  167  Table 3.11. Principal components analysis of endemic Hawaiian Lysimachia based on allele frequencies from allozyme analysis of 35 populations. Variable  Component Loadings 1 2  3  4  5  TPI—1A TPI—1B TPI—1C TPI—2A TPI—2B TPI—2C TPI—2D TPI—2E TPI—3A ADH—1A IDH—1A IDH—1B IDH—1C IDH—1D IDH—1E MDH—2A MDH—2B MDH—2C MDH—2D MDH—3A MDH—3B MDH—3C SKD—1A SKD—1B SKD—1C DIA—1A PGM—1C PGM—1D PGM—1E PCI—lA PGI—1B PGI—1C PGI—1D  —0.171 0.117 —0.099 0.612 0.047 —0.593 0.048 0.190 0.453 —0.416 0.312 0.706 —0.766 0.232 —0.027 —0.027 —0.676 0.601 0.486 0.244 —0.379 0.274 —0.121 0.299 —0.326 —0.073 0.171 —0.459 0.401 0.284 —0.276 —0.273 0.140  —0.011 —0.118 0.119 0.153 —0.390 —0.217 —0.158 0.257 0.245 —0.292 —0.255 0.240 0.112 —0.658 0.084 0.084 0.208 —0.289 0.331 0.038 —0.184 0.221 —0.146 0.097 0.024 0.040 —0.641 —0.349 0.525 —0.898 0.901 0.012 —0.544  —0.012 0.051 —0.049 —0.533 —0.041 0.459 —0.176 —0.011 0.253 0.458 0.123 0.040 0.004 —0.157 0.592 0.592 —0.147 0.167 —0.538 0.689 —0.732 0.296 —0.300 0.339 —0.170 —0.023 0.035 —0.032 0.000 —0.045 0.037 —0.024 0.113  0.043 —0.890 0.885 0.005 0.035 0.135 0.177 —0.310 —0.367 —0.034 —0.224 —0.057 0.129 —0.096 0.067 0.067 —0.478 0.460 0.093 —0.050 0.072 —0.049 —0.031 0.166 —0.228 0.578 0.111 —0.056 0.034 —0.072 0.069 —0.042 0.031  —0.360 0.056 —0.017 0.052 0.351 0.131 0.052 —0.308 —0.163 —0.099 —0.308 —0.168 0.273 —0.183 0.131 0.131 0.121 —0.155 0.064 —0.037 —0.225 0.359 0.459 —0.464 0.174 0.159 0.382 —0.659 0.537 0.047 —0.064 —0.228 0.389  Eigenvalues % Variance % Cmii. Var.  4.637 14.952 14.952  4.146 12.562 27.514  3.239 9.814 37.328  2.851 8.641 45.969  2.458 7.450 53.419  168 Figure 3.4. Comparison of principal components analysis of vegetative, reproductive and allozyme data of endemic Hawaiian Lysimachia. Points are of population averages labelled as follows: D=L. daphnoides, U=j. remyi. subsp. kipahuluensis, C=. remvi subsp. caliginis, P=L. pendens, W=L. filifolia, G=L. glutinosa, H=L. waianaeensis, K=L. kalalauensis, S=j. remyi subsp. subherbacea, M=L. maxima, O=. ovoidea, R=. remyi subsp. remyi, N=L. scopulensis, X=. remvi subsp. kipahuluensis X . . subsp. caliginis (hybrid swarm).  REPRODUCTIVE 2  I  I  I  •  W. j•U.  s  X.  8  c. 0.  0  U. 8.  0.  —1  P.  C.  H.  8.  8.  P.  -2  .8  1<.  8.  -  .0  -2  -1  1  0  2  3  pci ALLOZYMES  VEGETATIVE  2  3  1  2  .0 .M•S  U.U.N.  8.  U  X  -  H  K  0  K 0  •  .13 I  -3  -2  I  -i  0  pci  1  2  -2  -2  -i  I  I  I  0  1  2  pci  3  169  Table 3.12. Summary table of Pearson correlation matrix comparing principal component scores from allozyme, floral, and vegetative data. Allozyme axes are abbreviated as PCISO1 and PCISO2, floral axes as PCFLOW1 and PCFLOW2, and vegetative axes as PCVEG1 and PCVEG2; “1” and “2” refer to the first principal component score and the second score respectively. Values marked with “*“ are not significant. PCVEG1  PCVEG2  PCFLOW1  PCFLOW2  PCISO1  0.489  —0.272  0.372  —0.128  PCISO2  —0.100  0.355  —0.695*  —0.029  170 34  Discussion  3.4.1  Allozynte variation within species of Hawaiian  Lys intachia  Two recent reviews of allozyme variation in plant populations provide a basis for comparison with Hawaiian Lysimachia  (Figure 3.2).  Hamrick and Godt  (1990),  calculated  average values of genetic diversity within populations, based on a compilation of 449 species in 165 genera. are continental taxa. of polymorphic loci per locus 0.149.  (A),  For the “average” species the percent  (P),  is 1.96,  is 50%,  For 81 endemic species  (1992),  the mean number of alleles  and the total diversity  (Ht),  and Ht=0.096.  DeJoode and  summarized allozyme variation in 62 insular  species from 16 genera.  The “average” insular endemic species  has even less variation than continental endemics, A=1.32 and Ht=0.064. are P=41%  is  (few were insular endemics),  the averages were P=40%, A=1.80, Wendel  Most of these  For Hawaiian Lysimachia,  (range=14-54%), A=1.5  (range=0.020—0.212)  (range=l.l-1.7)  (Table 3.3).  P=25%,  these averages and Ht=0.117  Allozyme variation in the  “average” Hawaiian Lysimachia is greater than the “average” insular species.  In fact,  comparing the values for Ht and P,  the level of variation in several species of Lysimachia is similar to that found in continental species should be pointed out however, upon the number of loci,  (Figure 3.2).  It  that P is especially dependant  and the 11 that were successfully  resolved in this study is a relatively small sample.  171  If genetic bottlenecks follow colonization from the oldest through to the youngest island, a stepwise reduction in genetic variation within populations would be expected. Hawaiian Bidens deviate from this expectation in that the mean values of genetic variation in populations from all islands are nearly equal and are actually slightly higher on the youngest island, Hawaii  (Ganders,  1989).  Hawaiian Lysimachia  do not entirely follow the expected trend either,  although  there is a marked decrease from Molokai to West Maui and finally to East Maui  (Figure 3.2).  One explanation for this  is that populations of Lysimachia are typically small, consisting of fewer than one hundred individuals.  In small  populations the loss of genetic variation through random genetic drift is greater in successive generations than it is in large populations.  This could explain why populations on  the older islands have less variation than would be expected. Two processes are at work here.  On the one hand,  required for the accumulation of novel alleles. hand,  time is On the other  the more that time has elapsed, the greater the  possibility of loss of variation through random genetic drift especially in small populations. A serious limitation in this comparison is the lack of samples from two species,  .  forbesii and L. hillebrandii,  that may now be extinct in the Koolau Mtns. on Oahu. populations of j..  The two  filifolia cannot be considered to represent  the level of genetic variation that may have once been present in species on that volcano.  In addition,  it cannot always be  172  assumed that the individuals that found new interisland populations necessarily come from the adjacent, next oldest island.  In the Hawaiian Madiinae,  it is hypothesized that  some species in the Maui complex (includes the islands of Maui, Molokai and Lanai)  have evolved from an ancestor from  the more distant island of Kauai rather than from the adjacent Oahu (Carr et al.,  1989).  From the oldest to the youngest islands genetic variation  within populations does not decrease in a stepwise manner, however,  the level of variation among populations  does decrease in this way.  (Table 3.5)  Populations from the younger  islands share the same alleles at nearly the same frequencies and few if any unique alleles are present.  On Kauai there has  been more time for divergence in allele frequencies among populations as well as time for mutations that can result in new alleles.  Six of the seven “unique” alleles that were  detected are restricted to species on Kauai, the seventh from the oldest part of Oahu, the Waianae Mtns. Almost without exception,  (Table 3.3).  it is only on Kauai that some  species do not share the same highest frequency allele at all loci  3.4.2  (Table 3.2).  Allozyme variation among species of Lysimachia The relatively high genetic identities among species of  Hawaiian Lysimachia is a consequence of most species sharing the same highest frequency allele at all or most loci.  This  supports the hypothesis that they have evolved from a single  173  ancestor.  The decrease in the range of intra—island genetic  identities from Kauai to East Maui  (Table 3.8)  suggests that  Kauaj was the first extant island to be colonized.  This trend  parallels the reduction in morphological diversity within islands from the oldest through to the youngest island (Chapter Two). Gottlieb (1977)  calculated that populations of the same  species had an average genetic identity of 0.95 ± 0.02 and the average for species belonging to the same genus was 0.67 ± 0.07.  Remarkably, Crawford (l990b)  reports that although the  range of variation has increased, the averages of Gottlieb (1977)  remain essentially the same even with the addition of  hundreds more, mostly continental, taxa.  The range of genetic  identities for populations of Lysimachia,  0.71-1.0,  intermediate to these values.  is  However, the mean of 0.89  indicates that the extent of divergence between most pairs of species is more typical of different populations of a continental species.  By comparison,  the mean genetic identity  among populations of a single species, Coreopsis integrifolia collected from Florida and Georgia, 0.786—0.997  (Cosner and Crawford,  is 0.925 and the range is  1994).  For the following reasons, newly founded populations of insular taxa are probably genetically depauperate compared to those.  Because of distance and barriers to dispersal, most  insular taxa are probably founded by a few or perhaps a single propagule.  In most cases,  it is propagules of self—compatible  hermaphrodite species that are able to found populations  174  following long distance dispersal to isolated oceanic islands, in accordance with “Baker’s Rule”  (Baker,  1967).  Such a  propagule potentially bears less genetic variation than an obligately outcrossed one.  However, Baker and Cox (1984)  cite  examples and discuss mechanisms whereby propagules of dioecious taxa may establish successful breeding populations, i.e. dioecy does not necessarily always evolve autochthonously,  on islands,  from hermaphrodite ancestors.  Compared to the few other insular genera (Table 3.13) which more than one species has been analyzed,  in  species of  Lysimachia have diverged from each other somewhat more than have those of Crepidiastrum, Bidens, Metrosideros, Tetramolopium, Wilkesia, or Dubautia  (n=13 species), but not  as much as those of Dubautia (ri=14 species), Wahienbergia or Robinsonia.  In fact, the degree of divergence among species  of the latter three genera is similar to that of continental species.  One factor that undoubtedly contributes to the  differences in the range of genetic identities among the different genera being compared is the fact that they have been in their respective archipelagos for different lengths of time.  175  Table 3.13. Comparison of Nei’s genetic identities among species of island genera. Taxon (Reference)  No. Taxa  Location  Range  Mean  Metrosideros (Aradhya, 1991)  5  Hawaii  0.79—1.00  0.92  Bidens (Sun and Ganders, 1986)  14  Hawaii  0.92—0.99  0.98  Dubautia (n=14 sp.) (Witter and Carr, 1988)  6  Hawaii  0.43—0.93  0.69  Dubautia (n=13 sp.) (Witter and Carr, 1988)  9  Hawaii  0.73—1.0  0.95  Tetramolopium (Lowrey and Crawford, 1988)  7  Hawaii  0.86—1.0  0.95  Lysimachia  15  Hawaii  0.71—1.0  0.89  Robinsonia (Crawford et al., 1992)  4  Juan 0.56—0.71 Fernandez  Wahlenbergia (Crawford et al., 1990)  3  Juan 0.68-0.95 Fernandez  Crepidiastrum (Ito and Ono, 1990)  3  Bonin Islands  Continental congeneric species Continental conspecific populations (Crawford, 1990)  0.75—0.99  0.63  0.84  0.67 0.95  176  This is especially important because divergence among insular species is more likely a function of the time necessary for new alleles to appear, rather than divergence through changes in polymorphic gene frequencies (Witter and Carr,  1988).  species of Dubautia are among the first  colonists of new lava, often in areas of low rainfall. have evolved from California tarweeds Carr,  1985)  (Hubbell,  They  (Baldwin et al.,  1991;  that are also adapted to dry environments  1968).  This suggests that one of the early arrivals  to Hawaii may have been the ancestor of Dubautia and the other members of the “silversword complex” and could explain the relatively lower genetic identities of these species as compared to the others. Although differences in the mean genetic identities of the genera listed above could be entirely related to differences in the length of time since they arrived in their respective archipelagos,  other aspects of their biology may  also have contributed to these differences.  Several species  of the “silversword complex” are self—incompatible al.,  1986),  et al.,  and species of Robinsonia are dioecious  1992).  (Carr et (Crawford  Because the establishment of a population from  a single self—incompatible ancestor of Robinsonia or the “silversword complex” is unlikely, Crawford et al.  (1992)  Carr et al.  (1986)  and  consider it possible that these taxa  were founded by more than one propagule.  As a consequence,  there may have been greater genetic variation in the founding population compared to other genera.  Although protogyny  177  promotes outcrossing in species of Lysimachia, plants often bear many flowers on a single stem, possible.  and geitonogamy is  The degree to which protogyny succeeds in promoting  outcrossing and influences the extent of genetic variation within populations of Lysimachia,  is unknown and would require  direct measurement of outcrossing rates. (1990)  Hamrick and Godt  report a mean heterozygosity of 0.074 for populations  of 113 species of self-pollinating species; the mean for animal—outcrossed species was 0.124 for populations of 164 species.  Levels of heterozygosity in species of Hawaiian  Lysimachia ranged from 0.02 to 0.21.  Thus some species have  levels of heterozygosity that are typical of outcrossing species, while other species are less variable than selfing species.  In populations from the younger islands,  it is of  course difficult to distinguish between reduced genetic variation as a consequence of self ing, or from reduced variation as a consequence of the founder effect. Obligate outcrossing in Robinsonia,  some species of the  “silversword complex” and possibly in Lysimachia, may have resulted in greater genetic variation among individuals. Because founders from an outcrossing population contain a smaller proportion of the overall variation present in the ancestral population,  it is possible that derivative  populations will diverge more from each other more rapidly than they would from populations of self ing species.  These  explanations do not seem to fit Hawaiian Bidens, however. Although species are self—compatible, the flowers are  178  protandrous and 9 of the 19 species are gynodioecious.  Both  of these aspects of their breeding system promote outcrossing. Outcrossing rates in 15 populations ranged from 0.43 to 0.88 (Sun and Ganders,  1988).  Despite this, Bidens species  nevertheless have high genetic identities. Genetic identities among species of Lysimachia from Kauai (1=0.743-0.998)  are lower than for all species of Bidens,  Tetramolopium, and Dubautia  (n=13 only).  Divergence among  species on Kauai was possibly accelerated by the evolution of three different corolla colors: white, green,  and reddish.  One hypothesis is that red was the ancestral color and that white and green resulted from mutations at different loci (discussed in more detail in Chapter 4).  Reproductive  isolation may have been relatively rapid if pollinators distinguished between the different corolla colors at the time each mutant appeared in populations of reddish flowered individuals.  The frequency of such a mutation could rapidly  increase and become fixed following establishment of a new population from a heterozygous individual or even genetic drift at the margins of existing populations.  Bawa  (1990)  has  hypothesized that because the geographical ranges of plants and their pollinators often do not coincide entirely,  floral  variants of a founder population could become fixed as a result of selection by a different set of pollinators. corollas of both L. glutinosa and  .  While  scopulensis are open, the  petals of j. kalalauensis are often tightly closed around the exserted style.  If and when the petals become reflexed the  179  corolla is still not as widely open as that of  criutinosa.  .  These differences also suggest a difference in pollinators. In the upper part of Kalalau valley on the island of Kauai, the white flowered  .  glutinosa, green flowered L.  kalalauensis and reddish flowered sympatrically.  Populations of L.  were sampled from this area.  .  scoiulensis grow nearly  scopulensis and j. glutinosa  However L. kalalauensis is  uncommon here and was not sampled from this location.  The  mean value for I between populations of L. glutinosa and j.. kalalauensis is 0.823; between L. cilutinosa and L. it is 0.766, is 0.949.  and between L. kalalauensis and  .  scopulensis  scopulensis it  Divergence is not only due to differences in  frequencies of shared alleles.  Ten alleles were detected at  relatively high frequencies in one or two species but not in all three.  Pgi—la was homozygous in all individuals of j..  glutinosa but was not detected in L. kalalauensis or scopulensis.  This is strong evidence that  .  glutinosa is  reproductively isolated from these two species. detected only in j. scopulensis glutinosa or L. kalalauensis. reproductive isolation of j.. Hybrids between  .  L. glutinosa and L.  (f=0.312)  .  Dia—la was  and not in L.  This is also evidence of scopulensis from the other two.  glutinosa and j. kalalauensis and between scopulensis do occur in this area.  However pollen stainability was very low in two putative natural F s of L. glutinosa and T.. 1  scopulensis.  collected from a result of a backcross to hybrid between j. glutinosa and  .  .  Seed  glutinosa of a  kalalauensis was grown in  180  the greenhouse.  This plant, putatively resulted from a  backcross with j.. criutinosa and also had low pollen stainability (see Chapter 4).  Other factors may have  contributed to reproductive isolation among these three species and possibly to allozyme divergence. differences in phenology, Unfortunately,  flower odor,  These include  and pollinator.  little is known about any of these.  All are  worthy of further research.  3.4.3.  comparison of allozyme and morphological variation  Species of Lysimachia are morphologically and ecologically diverse, yet genetic identities among them are quite high.  It is somewhat of a paradox that the estimate of  genetic variation, allozyme divergence,  suggests few genetic  differences, however the morphological differences among species suggests that there has been greater genetic divergence.  Continental taxa that have high genetic  identities are usually species pairs that are morphologically similar  (Crawford et al.,  1987b).  The most likely explanation for the occurrence of high genetic identities among morphologically distinct insular species is that divergence has been much more rapid for genes that regulate morphological development and ecological adaptation than for genes that code for enzymes of primary metabolism (Crawford,  et al.,  l987b).  A second possibility is  that there is a simple genetic basis for the expression of morphological characters.  In Hawaiian Bidens,  a number of  181  characters that are used to distinguish among species are controlled by one or two loci  (Ganders,  1989).  Characters  controlled by one locus are: erect vs. decumbent habit, determinate vs.  indeterminate flowering, achene awns smooth  vs. barbed and three vs.  five to seven leaflets.  Other  leaflet numbers are controlled by additional duplicated loci. Two loci control the following: coiled,  achenes straight or curved vs.  achene awns distinct vs. decurrent,  wingless;  achenes winged vs.  leaves pubescent vs. glabrous; disk corolla red vs.  yellow; achenes setose vs. glabrous. sterility)  is controlled by two loci  Gynodioecy (male (Sun,  1987).  The low correlation between morphological divergence and allozyme divergence among populations in Lysimachia 3.12) 1985).  is similar to that in Bidens Furthermore,  (Table  (Helenurm and Ganders,  in both studies,  scores from the first  PCA axis of allele frequencies represent a small proportion of the total variation, Lysimachia.  Overall,  l3.3 for Bidens and 14.9 for the species composition of morphological  clades differ from that of allozyme clades, thus the poor correlation between the two  (Table 3.12).  This can be seen  most clearly by comparing leaf shapes of species in Figure 3.1 to their position in the phenogram of Figure 3.3.  In the case  of Lysimachia daphnoides and L. glutinosa, allozyme divergence does parallel the possession of a unique morphological character.  Nearly all plants of these species also bear the  same alternate allele at Pgi—1 and are differ from the remaining species in having viscid leaves  (those of  .  ini]ci  182  are also viscid, but this species was discovered too late to be included in the present study).  Leaves of j.  glutinosa are  glabrous whereas those of L. daphnoides are viscid—hirtellous and they have diverged from each other in a number of characters including corolla color, habit, habitat and leaf size and shape.  Among the remaining species there are no  other examples of parallel allozyme and morphological divergence,  i.e., the presence in all individuals of a unique  allele as well as of a unique morphological character.  3.4.4.  Taxonomic implications  Taxonomic conclusions based on the groupings presented here need to be interpreted with caution because the overall genetic identities among populations and species are so high. Allozymes can be useful as taxonomic markers when they occur as mutually exclusive  (to one or several taxa among the group  of taxa under investigation) unique alleles.  However as  stated above, the differences between most species pairs were usually differences in allele frequency only.  Much of the  divergence among species is a function of differences in the frequency of alleles held in common rather than to the presence of unique alleles which occur at a significant frequency.  In general,  the results presented here support the  revisions proposed in Chapter Two.  Referring to Figure 3.3,  it can be seen that the following changes from the species delineations of Wagner et al. this electrophoretic study:  1)  (1990)  are in agreement with  recognition of j.  filifolia and  183  L. pendens,  as distinct species  included in  .  remyi;  3)  filifolia); 2)  recognition of  hillebrandii,  .  .  (j. pendens was formerly  recognition of subspecies of  ovoidea,  .  .  waianaeensis, L.  remvi subsp. caliginis and  .  subherbacea as distinct species or subspecies,  remvi subsp. (these had  previously been classified as L. hillebrandii sensu Wagner et al.  (1990)).  3.4.5.  Summary  The pattern of genetic variation in Hawaiian Lysimachia, as estimated from allozyme variation,  is similar to that seen  in the few other Hawaiian genera that have been investigated to date.  There is little divergence among most species as  compared with that seen among continental taxa, with the exception of L. glutinosa and L. daphnoides which do form a distinct dade at the lowest branch of the tree.  Further,  general there is little correlation between the pattern of allozyme divergence and morphological divergence.  in  184  Chapter 4 Fertility of artificial interspecific hybrids and breeding behavior of the endemic Hawaiian Lysimachia  4.1  Introduction The level of fertility of hybrids can be used as a tool  to provide insights into the processes of differentiation between parental taxa,  as well as interspecific relationships,  especially in conjunction with observations of chromosome pairing at meiosis.  There have been no previous attempts to  analyze the level of fertility of artificial interspecific hybrids of species of Lysimachia endemic to the Hawaiian Islands.  Although interfertility may indicate genetic  similarity of the parental genomes,  intersterility does not  necessarily imply greater genetic divergence unless its cause is known  (Davis and Heywood,  1973).  From progeny analysis,  hypotheses regarding genetic regulation of morphological characters can also be tested.  In general,  it appears that  reproductive isolation between congeneric species of Hawaiian plants is a result of ecological or geographical isolation rather than chromosomal differences such as aneuploidy, polyploidy,  or structural rearrangements.  This hypothesis is  based on the high pollen or seed fertility of interspecific hybrids among species of Bidens Tetramolopium (Lowrey, Wilkesia  (Carr,  Wikstroemia  (Ganders and Nagata,  1986), Dubautia, Argyroxiphium, and  l985a), Vaccinium (Vander Kloet,  (Mayer,  1984),  1991),  Schiedea  1993),  (Weller and Sakai,  1988),  185  intrasectional species of Lipochaeta 1980), Carr,  (Rabakonandrianina,  and intrasubsectional species of Portulaca (Kim and l990a).  The results of adaptive radiation in each of  these genera have been species whose morphologies and ecological requirements are often remarkably divergent. Chromosome numbers, available for 18.8% of Hawaiian flowering plant species  (Carr 1978,  1985b), are the same for  nearly all groups of species that are thought to have evolved from a common ancestor.  The only case of aneuploidy is in  Dubautia in which a group of n=l3 species are the result of aneuploid reduction from n=14 species  (Carr,  1985a).  In situ  polyploidy is unknown in Hawaiian plants except for a single plant of Portulaca that appears to be an allopolyploid between an indigenous species and in introduced one  (Kim and Carr,  1990a) Prior to this study nothing was known about the breeding behavior of Hawaiian Lysimachia,  i.e., the type of  compatibility system or the existence of mechanisms to promote outcrossing such as dioecy, dichogamy,  or heterostyly.  Species in Lysimachia section Seleucia, restricted to North America,  are self-incompatible  (Coffey and Jones,  1980).  Heterostyly, which is common in at least some genera of the Primulaceae (Richards,  1986)  is unknown in Lysimachia,  although a single Chinese species,  .  crispidens, has long and  short-styled flower morphs (Chen and Hu,  1979),  but stamen  lengths are evidently the same in the two morphs.  186  The purposes of the present study were to:  analyze the  1)  fertility of interspecific F 1 hybrids in order to elucidate the nature of reproductive barriers among species; and 2)  to  arrive at a better understanding of the breeding behavior of Hawaiian Lysimachia.  Somewhat ancillary to this study, but  also of interest, were observations of the morphology of F 1 hybrids.  4.2  Methods and Materials  4.2.1  Artificial crosses  Seeds and cuttings were collected in the field from 22 populations  (Table 4.1)  and grown in the greenhouse.  collection consisted of the following taxa: Gray)  Hillebr.,  A. Gray,  .  (R. Knuth) caliginis  subsp.  iniki Marr, L. kalalauensis Skottsb., .  .  .  maxima  remyi subsp. kipahuluensis (St.  L. remyi subsp. remyi  (St. John) Marr, L. remyi  (St. John) Marr,  waianaeensis St. John, and L.  .  (A.  ovoidea St. John, L. remyi subsp.  (St. John) Marr,  subherbacea  Lydgate.  daphnoides  glutinosa Rock, j.. hillebrandii Hook f. ex  .  St. John,  John) Narr,  .  This  L.  scopulensis Marr, j..  filifolia C.N.  Forbes and  All of these flowered except for j.. daphnoides and  scopulensis.  Lysimachia iniki and i..  ovoidea did not  flower in time for the results to be included here.  187  Table 4.1. Collection localities and numbers for seeds of species of endemic Hawaiian Lvsimachia used for artificial hybridizations. Species  Collection locality and voucher number.  L. glutinosa Kauai: Kokee, Kalua Puhi Trail, Marr 254. Kauai: Kokee, beside road between Kalalau and Puu 0 Kila Lookouts, Marr 255,276. Kauai: Kokee, below Kalalau Lookout, Marr 1304. L.  kalalauensis Kauai: Kokee, Nualolo Valley, Marr 271, Penman 11287. Kauai: Waimea Canyon, Waialae Valley, south of Waialae falls, Penman 11659. Kauai: Kokee, Honopu Valley, Mann 257. Kauai: Kokee, Awaawapuhi Trail Marr 273.  L.  filifolia Oahu: Koolau Mtns., Waiahole Gulch, Penman 11149.  L. waianaeensis Oahu: Waianae Mtns., Makaha Valley, Marr 241,242,244,1297.  Kamailenunu ridge,  L. maxima Molokai: Pelekunu Valley, north of Ohialele, Mann 1299, 1308. L.  .  remvi subsp. subherbacea Molokai: Kamakou, south slope of Puu Kolekole, Marr 372, 374, 377. Molo]cai: Kamakou, ridge south of Onini Gulch, Mann 380. Molokai: Kamakou, Onini Gulch, Mann 386,387,388. Molokai: Kamakou, rim of Waikolu Valley, ridge SE of Puu Kaeo, Mann 393,395,399,400. Molokai: Kamakou, small ridge between Kauanakakai and Kupaia Gulches, Mann 1306. remvi subsp. remvi West Maui: near summit of Lihau, Marn 410,415,416,424, 1302,1303,829. West Maui: ridge between Halepohaku and Ulaula, Mann 935. West Maui: Hanaula, Mann 346,347,350,354,355,356, 357,359,361,362,363,919.  Table 4.1.  continued on next page.  188  Table 4.1. continued. Collection localities and numbers for seeds of species of endemic Hawaiian Lysimachia used for artificial hybridizations. Species .  Collection locality and voucher number.  remyi subsp. kiiahuluensis West Maui: lao Valley, upper Nakalaloa stream, Marr 432,  436.  East Maui: Haleakala, notch in ridge above and SE of Paliku Cabin, Marr 281. East Maui: Kalapawili ridge, rim of Kipahulu Valley, Marr 282,283,284.  East Maui: Kalapawili ridge, Marr 285,290,291.  .  above Lake Waianapanapa,  remyi subsp. caliginis East Maui: Haleakala, Koolau gap at and below treeline, Marr 336,337,338,343,344.  Hybrid swarm between j. remyi subsp. kipahuluensis and L. remvi subsp. caliginis East Maui: Haleakala, Kaupo gap, Marr 296,308.  189  Prior to anther dehiscence, the female parent was emasculated and pollen was transferred from the male parent to the receptive stigma.  The manipulations included  interspecific crosses and self—pollinations. seif-pollinations were attempted: pollination within a flower; 2)  1)  Two types of  intrafloral,  intraplant,  i.e.  i.e. pollen  transferred from an older flower, to a younger emasculated one on the same plant in which the stigma was still receptive. Self—pollinations were attempted in order to determine if plants were self—incompatible. A total of 605 crosses were attempted among the 6 species and 4 subspecies that reached the flowering stage. possible,  several plants of each species were used for each  hybrid combination, however for some species, filifolia,  Whenever  only one plant flowered.  e.g.  .  The number of crosses  attempted per combination varied from 1 to 90 and averaged 16. Forty-five intertaxon hybrid combinations were possible but because species did not flower synchronously, combinations could be attempted. attempted for 24 pairs of taxa.  only 38  Reciprocal crosses were A few crosses were attempted  using an endemic Hawaiian species as the female parent and as the male parent, Asian species,  .  an indigenous species, j.. mauritiana, litchiangensis and L. barystachys.  and two The  endemic Hawaiian species of Lysimachia are believed to be most closely related to those from southeast Asia 1990).  (Wagner et al.,  Crosses with these non—endemic species were performed  in this admittedly small sample, to determine if  190  interfertility might be useful to elucidate the nearest ancestor of the endemic species.  Vouchers of hybrids and  species are deposited at UBC.  Pollen stainability  4.2.2  The fertility of parental and hybrid plants was estimated from pollen viability.  Viability was evaluated by observing  pollen grains that had been treated with a cytoplasm stain, cotton blue dissolved in lactophenol.  A minimum of 300 pollen  grains were examined after they had been treated for at least 24 hours.  Those that turned dark blue were scored as viable.  Those that were misshapen or that stained lightly were scored as non—viable.  Whenever possible, pollen stainability was  estimated for at least two flowers per plant.  As a control,  the fertility of pollen from herbarium specimens of plants from natural populations was also tested in order to evaluate the effect that greenhouse growing conditions might have upon pollen fertility. Although F 2 and backcrossed plants are currently in cultivation,  none have flowered.  Seeds were collected on  Kauai from a plant whose morphology is the same as artificially produced hybrids between L. glutinosa and L. kalalauensis.  The pollen parent of these seeds was probably  L. glutinosa because of the proximity of plants of this species to the hybrid.  Several plants were grown from these  seeds and one has flowered from which a pollen sample was taken.  Pollen viability of two wild plants, that based on  191  morphological intermediacy are putative hybrids between glutinosa and L.  .  scopulensis, was tested from herbarium  specimens.  4.2.3  Vacuolar flavonoids In a very preliminary analysis, vacuolar flavonoids were  isolated and visualized using thin layer chromatography methods based on Gornall and Bohm (1980).  192  4.3  Results Fruit—set in interspecific crosses  4.3.1  Percent fruit-set ranged from 0% to 100%  (Table 4.2),  however intermediate levels of fruit—set were more typical. Combinations that resulted in either 0% or 100% fruit-set had low sample sizes, generally fewer than four. involving  .  Crosses  cilutinosa failed more often than crosses  involving other species.  Fruit—set among all other  combinations was almost always greater than 50%.  Sometimes  there were differences between reciprocal crosses, e.g. the success rate of crosses in which j.  glutinosa was the male  parent was consistently higher than when it was the female parent. Hybrids grew vigorously in the greenhouse environment, however not all hybrid combinations had flowered at the time of this writing.  4.3.2 Fruit set in seif-pollinations  Observations in the greenhouse indicated that flowers were protogynous.  Stigmas were receptive 2—4 days prior to  anther dehiscence and there was little overlap in the time of stigma receptivity and anther dehiscence.  As flowers mature,  the filaments bend toward the style and actually clasp it. When the corolla falls from the plant, pollen is easily deposited on the stigma, yet in hundreds of observations of these unmanipulated,  selfed flowers,  fruit—set was very rare.  remyi subsp. kipahuluensis(WM)  i.. remyi subsp.  remyi subsp. kipahuluensis(EM)  j. remyi subsp. caliginis (EM)  remyi subsp. subherbacea L. maxima  9.  11.  13.  15.  16.  .  .  remyi  **  83(6)  38(13) 18(17)  7  ** cross not attempted Table 4.2 continued on next page.  17.  L. waianaeerisis  8.  .  L.  7.  filifolia  L. kalalauensis  2.  100(3) 85(11)  L.  1.  clutinosa  2  0(3)  **  **  **  **  100(2) 100(11)  **  **  100(1)  **  **  **  33(3) 33(3)  **  50(4) 100(1)  100(4)  **  45(11) 17(6)  50(18) 100(4)  64(11) 80(4)  **  0(2)  69(23) 100(2)  54(24) 50(2)  71(38) 45(18)  (number of cross attempted) Species 9 11 13  67(6)  **  8  Percent Fruit—Set  70(10) 70(10)  33(3) **  66(3) **  **  100(1)  66(4) 100(2)  60(10) **  71(14) 33(3)  15  Table 4.2. Percent fruit—set of artificial crosses among species of endemic Hawaiian Upper values of a pair are male parent (row) X female parent (column), lower Lysimachia. values are the reciprocal cross.  87(8) 82(11) 86(14) 100(9) 74(34) 55(56) 76(13) 66(24)  . remyi subsp.  kipahuluensis(WN)  L. remyi subsp. remyi  . remvi subsp.  kiahu1uensis(EM)  !. remvi subsp. caliginis(EM)  i. remvi subsp.  9.  11.  13.  15.  16.  **  17.  70(10)  L. waianaeensis  8.  cross not attempted  subherbacea L. maxima  79(19) 57(7)  filifolia  L.  7.  **  89(18) 100(1)  kalalauensis  L.  79(43) 35(17)  2.  glutinosa  L.  1.  16  (number of cross attempted) Species  100 (3)  **  **  **  0(1)  **  100(1) 100(1)  **  **  **  **  0(1) 100(1)  100(1)  **  100(1) 50(2)  17  Percent Fruit-Set  Table 4.2. continued. Percent fruit—set of artificial crosses among species of endemic Hawaiian Lvsimachia. Upper values of a pair are male parent (row) X female parent (column), lower values are the reciprocal cross.  195  In the 597 intrafloral self—pollinations that were attempted, fruit—set for all species ranged from 0% to 40% However,  (Table 4.3).  in the 40 intraplant attempts at self-pollination the  percentage of fruit-set was markedly higher.  4.3.3 Morphology of hybrids Hybrid plants were morphologically intermediate in characters such as the size and shape of calyx lobes,  corolla  lobes and leaves, the degree of pubescence and growth habit. Progeny of the same cross closely resembled each other in the characters mentioned above, with the exception of corolla pigmentation. e.g. L.  Crosses involving species with red corollas  filifolia or subspecies of j. remvi,  glutinosa  (white corolla)  and either j.  or L. kalalauensis  (green corolla),  generally produced progeny with red corollas  (Figures 4.1 and  4.2), with two exceptions j.  (Table 4.4).  One of the progeny of  glutinosa X j. remyi subsp. remvi had a white corolla,  while a sibling had a red corolla  (Figure 4.3).  exception was two progeny from the cross  .  The other  remyi subsp.  kipahuluensis X j. kalalauensis, both of which had greenish corollas.  196  Table 4.3. Percent fruit-set of seif—pollinations of species of endemic Hawaiian Lysimachia, intrafloral self s and intraplant selfs. Self-pollination that were not attempted are marked by “na”. Percent Fruit-Set (flu Intrafloral  Species cilutinosa  Intraplant  19(32)  0(2)  j.. kalalauensis  8(12)  na  L.  35(14)  50(2)  L. waianaeensis  0(5)  na  .  remvi subsp. kipahuluensis (WM)  40(5)  100(1)  I.. remyi subsp.  19(122)  75(12)  13 (184)  80(10)  j. remyi subsp. caliginis  25(24)  100(1)  k. remyi subsp.  25(199)  75(12)  .  filifolia  kipahuluensis (EM)  j.  remyi subsp.  subherba cea  remyi  0  • co (U  4i ..  r4  Oco co  .4 U) o\Q  o  rd  $4  .  —I  . 0 U) — -I-) 4J4iq-  -o  U.) 0) a)H  CU  1 p.  •1  ---J  H  th°  co 4 p U) ..QU) -  U) U)tC)  co •.-41 a)  a)r  •(U  r1 H  CU  4.)  .4.)  0  0  U]  a)  -H -i  a)4)  C)  4)  w  ow  (U -i  Q)4)  thO  S---r  •..l U) % N  r  S >4rc’.J  0 7.4 -r rj  .p4 .0  r10  lI0r  cU.—. a)  U)-4 U) U)  >  rI  r- p4.r-1 q-4 ..Q  0 CUrH  -I -‘-I  a) 0)4.)  -H  a)  ‘-U)  0 U)  .l  (J)4J  c  Cfl CU U)  -1 p.iI  a) : w  (U  a)CU  q-4 .Q  4.) •-i Qr1rI -I CU —I  CU  U)v-4 U) a).  4—  T-0  cU U) a)r1r  •  —S  .4-)LL)  00.  S rI LC) thO br4 t- % Cl  198  Table 4.4. Corolla pigmentation of F 1 hybrids between species of of endemic Hawaiian Lysimachia. Only those combinations in which the parents have different colored corollas are included here. Female parent No.  Male parent No.  Sibling No.  Corolla pigmentation  crlutinosa X L. kalalauensis 255.2 257.1 1 light yellow above; red base 255.2 257.1 2 light red above; red base 255.2 257.1 3 light green above; red base  .  L. kalalauensis X L. glutinosa 271.1 255.2 1 mostly white with red veins 271.1 255.2 2 mostly white with red veins 271.1 255.17 1,2 red L. glutinosa X L. filifolia 1304.1 11149.1 1  red  filifolia X L. . 11149.1 255.1  glutinosa 1  red  L. glutinosa X . 1304.1 432.1  remyi subsp. kipahuluensis 1 red  (West Maui)  remyi subsp. kipahuluensis X L. glutinosa 436.107 255.2 1 red  (West Maui)  • cilutinosa X . 255.2 281.7  (East Maui)  .  remvi subsp. kipahuluensis 1 red  remyi subsp. kipahuluensis (East Maui) 281.2 255.2 red 1,2 285.2 255.2 1,3,4,5, red 6,7,8  .  j. glutinosa X . 255.8 359.1 255.8 359.1  glutinosa  remyi subsp.remyi 1 red 2 white with red streaks  j. remyi subsp.remyi X 355.103 255.2 1 356.201 255.2 1,2 361.2 255.2 1,2 Table 4.4.  X L.  glutinosa red red red  continued on next page  199  Table 4.4. continued. Corolla pigmentation of F 1 hybrids between species of endemic Hawaiian Lysimachia. Female parent No.  Male parent No.  Sibling No.  Corolla pigmentation  L. glutinosa X L. remyi subsp. remyi 255.1 254.4  424.1 1301.1  1,2 1  j. remvi subsp. remyi X i. 410.101 255.2 1,2  L. glutinosa X 255.8 255.2 255.16  .  372.201 393.5 395.205  red red  glutinosa red  remyi subsp. subherbacea 1 red 1,2 red 1 red  L. remyi subsp. subherbacea X 372.1 387.202 393.3 399.4  255.2 255.2 255.2 255.2  kalalauensis X 271.1 387.202  .  1 1 1,2,3,4 1  .  1  .  red red red red  cilutinosa  remyi subsp. subherbacea red  L. remyi subsp. subherbacea X . kalalauensis 400.1 271.1 red 1 L. filifolia X L. 11149.1 271.1  kalalauensis 1 red  remyi subsp. subherbacea X . kalalauensis 410.101 271.1 1,2 red  .  L. remyi subsp. kipahuluensis (East Maui) X L. kalaluaensis 282.4 282.13  257.1 257.1  1,2 1  green outside, green outside,  red on inside red on inside  200  I.. glutinosa  Kauai Honopu  W Maui Manawainui  Figure 4.3. Leaves and flowers of two siblings of the cross L. glutinosa (#255.8) (female parent, left) X L. remvi subsp. remyi (#359.1) (formerly classified as . remyi, thus the label). Pollen stainabilities from left to right: not counted, 42%, 78%, 75%.  0 2  .3  4  5  6  7  8  9  10  ii  12  13  14  15  16  17  18  19  20  21  22  23  24  25  Leaves and flowers of three siblings of the cross Figure 4.4. L. glutiriosa (#255.2) (female parent, left) X . kalalauensis Pollen fertilities from left to right: 78%, 75%, (#257.1). 83%, 66%, 73%.  201  Hybrids between L. glutinosa and L. kalalauensis had corollas that were neither white nor green, but instead, were either red or nearly white, though not as white as that of glutinosa (Table 4.4).  In this small sample,  .  it was clear  that the pigmentation of the hybrid corollas differed among siblings of the same cross  (Figure 4.4)  progeny derived from different parents  as well as among (Table 4.4).  A second character that is not expressed in an intermediate manner is “viscidness”.  Entire plants of  .  glutinosa are very viscid, however progeny of crosses involving this species and non—viscid species, were not viscid. Leaf size and shape of a hybrid between  .  remvi subsp.  caliginis from Maui and L. glutinosa from Kauai was remarkably similar to that of L. maxima from Molokai.  Although it is  unlikely that L. maxima is a hybrid of these species, this observation suggests that relatively minor genetic differences separate maxima,  .  glutinosa, j. renwi subsp. calicrinis and I.  at least in terms of genes that control leaf shape.  4.3.4 Pollen stainability of species Pollen stainability of plants grown in the greenhouse varied from 36% to 99%. 57—86%  (Table 4.5).  Appendix 1,  The mean for each species ranged from  Data are presented for each plant in  Table A4.1.  At least one plant from each species  had pollen stainability greater than 75%.  Pollen stainability  202  of herbarium specimens also varied.  No species had  consistently high or consistently low pollen stainability.  4.3.5.  Pollen stainability of hybrids  Progeny of hybrids involving  .  glutinosa, L.  kalalauensis, L.  filifolia and the three subspecies of j..  remyi flowered.  Pollen stainability of hybrids varied from 0-  (Table 4.6).  98%  Data for individual hybrids and the parents  involved in the cross are presented in Appendix 1, Table A4.2. Mean pollen stainabilities of crosses involving either j.. glutinosa or  .  kalalauensis ranged from 41—85%, however, the  overall range was from 0—97%. generally higher  (72-93%)  Mean pollen stainability was  in hybrids among j.  filifolia and  the subspecies of L. remyi, yet here as well, the overall range was from 9-98%. included L.  Some plants of hybrid combinations that  glutinosa had pollen stainabilities of less than  25%. The progeny of the natural hybrid between j.. glutinosa and L.  kalalauensis, putatively backcrossed to J.. cilutinosa,  had 5% pollen stainability. .  crlutinosa and j.  stainability.  The two natural hybrids between  scopulensis, had 6% and 17% pollen  L. glutinosa L. kalalauensis  .  .  L. j.  j.  L. L.  .  Habitat 77 76 70 57 68 73 64 78 47 86 80 80 86 77 54  Mean  55 51 66 30 32 37 52 27 5 53 28 70 75 9 9 6 2 2 3 9 3 35 10 47 28 2 3 4 1  13 10 5 25 32 18 10 18 21 11 19 15 6 34 16 11 2 4 3 13 3 60 10 86 53 2 6 6 1  Mm  % Pollen Stainability N N SD Plants Flowers  95 87 74 78 88 97 71 98 78 98 98 91 94 99  Max  Pollen stainability in cotton blue of species of endemic Hawaiian Lysimachia.  greenhouse greenhouse wild greenhouse maxima wild ovoidea greenhouse remyi subsp. caliginis wild remyi subsp. kipahuluensis greenhouse wild greenhouse remyi subsp. remyi greenhouse subherbacea subsp. remvi wild greenhouse filifolia greenhouse waianaeensis wild  Species  Table 4.5.  0  r.  subsp.  subherbacea/. mauritiana*  18  28  61  95 90 92 94 56 85 31 45 11 2 18 16 8 2 16 6 13 3 34 11 82 89 70 72  3  93  9 15 16 35  78  1  98 33 21  48  — — —  97 90 93 77 97 88 94 94 90  0 23 23 13 18 20 17 71 71 31 38 19 18 22 16 35 10 8 7 2 3 12 12 12 3 2 2 2 10  18 3 10 21 24 26 4 4 4 1 23  Max Mm  56 67 62 43 52 42 68 85 79 41 64  Mean  * Pollen Stainability N N SD Plants Flowers  *  This must be regarded as a putative hybrid because morphologically it resembles their endemic Hawaiian parent; see text for further discussion.  L.  L. glutinosa/L. kalalauensis L. glutinosa/L. filifolia L. qlutinosa/L. r. subsp. calicrinis L. glutinosa/L. r. subsp. kipahuluensis L. glutinosa/L. . subsp. renwi L. glutinosa/L. . subsp. subherbacea . kalalauensis/L. r. subsp. kipahuluensis kala1auensis/. . subsp. remyi . . kala1auensis/. . subsp. subherbacea L. kalalauensis/L. filifolia . subsp. ca1iginis/. . subsp. . subherbacea kipahu1uensis/. subsp. r. subsp. r. subherbacea filifolia kipahu1uensis/. subsp. . . subsp. rexnyi/L. subsp. r. caliginis . r. subherbacea remyi/. . subsp. . subsp. . filifolia j. . subsp. remvi/j.  Cross  1 hybrids between species of endemic Pollen stainability in cotton blue of F Table 4.6. Hawaiian Lysimachia, and between endemic species and non—endemic/non—Hawaiian species of Lysimachia.  1.J 0  205  The pollen stainability of putative hybrids between endemic Hawaiian species and indigenous or non—Hawaiian species was high  (Table 4.6), given the differences in  morphology and chromosome number (Carr,  1978),  (.  L. barystachys has n=24  mauritiana has n=10 (Ko et al.,  chromosomes numbers are available for i..  1986), no  litchiangensis.  The  morphology of the hybrids resembles that of the endemic Hawaiian parent.  Preliminary two dimensional thin layer  chromatography of foliar flavonoids indicated that litchiangensis and L. profiles.  barystachys have similar flavonoid  This profile differs from both L. nauritiana and  remyi subsp. subherbacea. and  .  .  remyi subsp.  .  Flavonoid profiles of 3.. mauritiana  subherbacea are similar to each other.  Based on these results,  it would appear that putative hybrids  between the endemic Hawaiian species and L.  litchianciensis and  j. barystachys resulted either from contamination (pollen from another endemic Hawaiian species)  or self pollination.  Flavonoid data are equivocal concerning the veracity of the putative hybrid between  .  remyi subsp.  subherbacea and J..  mauritiana because both species have similar profiles, therefore it is not possible to detect in the hybrid the complementary pattern that would be exhibited by a true hybrid.  The wide range of pollen sizes and stainability  (Figure 4.5)  of the hybrid between  .  remvi subsp.  subherbacea  and L. mauritiana was much greater than was seen in any other cross,  and strengthens the case for this plant being a hybrid,  not a self.  206  a  Pollen from the cross j. renwi subsp. Figure 4.5. (female parent) X . mauritiana.  remyi  207  4.4.  Discussion Interpretation of results presented here are tentative  pending further study.  It is difficult to draw conclusions  from the reduced fertilities of some hybrids of Hawaiian Lysimachia because the species themselves, both in the greenhouse as well as in the natural environment, have such a wide range of fertility. Species of Hawaiian Lysimachia had a greater range of pollen stainability as well as minimum values lower than those of the few other species that have been measured.  Pollen  stainabilities from other genera are as follows: nearly 100% in Bidens  (Ganders and Nagata,  and Carr,  1990b); 77-99% in Wikstroemia  1984); 85-95% in Portulaca  than 90% in Tetramolopium (Lowrey,  (Mayer,  1986).  (Kim  1991); greater  There are several  possible reasons for reduced pollen fertility in species of Lysimachia.  One could be the fact that the species, which by  the high chromosome number must be polyploid, may not have fully “diploidized” genomes.  A result of this could be the  formation of multivalents and unequal segregation of chromosomes during meiosis.  Additional possibilities include  male sterility or inbreeding depression.  The latter would  seem to be a logical explanation because many populations are small and isolated, and although protogyny increases the likelihood of outcrossing, within plant pollination can occur. Biparental inbreeding depression is possible and in fact would seem to be likely in a species with a dry dehiscent fruit that inevitably deposits many seeds close to the maternal parent.  208  The values for genetic variation that were calculated from allozyme analysis  (Chapter 3),  indicated that many populations  have levels of heterozygosity that are typical of self ing species.  However, the levels of heterozygosity of other  populations are typical of outcrossing species. A somewhat surprising result, that is not expressed clearly in Tables 4.5 and 4.6, but is evident from careful study of Tables A4.l and A4.2 in Appendix 1, hybrids were more fertile than their parents.  is that many Restoration of  fertility in the hybrids could be a manifestation of heterosis.  This could be especially pronounced when  populations of one or both parental species are experiencing significant inbreeding depression. Mean pollen stainability is not consistent in hybrids among the four species and three subspecies tested here. Although some hybrids did have very low pollen stainability, in all combinations there was at least one plant that produced nearly 100% fertile pollen.  Even in hybrids between such  morphologically divergent species as j. glutinosa and L. filifolia  (Figure 4.1), post—zygotic barriers to crossability  are lacking.  Furthermore,  it is impossible to distinguish  between the possible effect of chromosomal divergence between the two parents as the cause for reduced pollen fertility in the hybrid, versus inheritance in the hybrid of the factors that contributed to the reduced fertility of the parents themselves. 43-93%,  The range of mean hybrid F 1 pollen stainability,  is similar to that published for hybrids among species  209  of several other Hawaiian genera as follows: close to 100% in Bidens  (Ganders and Nagata,  (Lowrey, 1980),  1986),  1984),  75—100% in Tetramolopium  66—100% in Lipochaeta  70—99% in Wikstroemia (Mayer,  Portulaca  (Kim and Carr,  1990a).  (Rabakonanadrianina, 1991),  and 44—97% for  Pollen stainability of  hybrids among Wilkesia, Dubautia and Argyroxiphium (genera of the monophyletic “silversword alliance”) (Carr,  l985a).  ranges from 9—99%  Wherever interspecific hybrids had reduced  pollen stainability, cytogenetic evidence implicated aneuploidy and/or reciprocal translocations as the cause of disruption of meiosis leading to reduced pollen stainability in these plants  (Carr,  l985a).  The relatively high fertility of the hybrid between an endemic Hawaiian species and  .  mauritiana and the similarity  of their flavonoid profiles in the prelimary analysis could indicate a closer relationship between these species than was anticipated. subsp.  remyi X  Plants of a backcross between the F 1 of j.. remvi .  mauritiana crossed with the  remyi parent, have been produced.  .  remyi subsp.  Lysimachia mauritiana could  theoretically have been the ancestor of the endemic species, however it is markedly different in a number of morphological characters as well as chromosome number.  Analysis of DNA  could provide further insights into the possible relationships between the endemic species and L. mauritiana as well as with Asian species. The reduced fruit-set of intrafloral pollinations appears to be the result of protogyny,  rather than genetically  210  controlled self-incompatibility.  However, more intraplant  crosses need to be attempted to verify this for several species.  If the ancestor of the Hawaiian species could be  identified,  it might be possible to determine whether or not  protogyny evolved in situ and was selected for as a means of promoting outcrossing. It is interesting to consider the genetics of corolla pigmentation.  All but one cross between  cilutinosa and the  .  red flowered species, produced red corollas suggesting that red is dominant to white.  (Table 4.4), The first plant to  bloom of the cross L. crlutinosa (white corolla) kalalauensis (green), had a red corolla.  X L.  This lead to the  hypothesis that red was the ancestral color at least of the extant Hawaiian species and that at two different points in the biosynthetic pathway for the production of red pigments, there were mutations resulting in the green flowers in one lineage and white ones in the other lineage.  Although  possible outgroups for the Hawaiian species have white or yellow corollas  (some of the yellow flowered species have red  at the base of the corolla),  from a biosynthetic standpoint,  it is more parsimonious to hypothesize that the ancestor of the Hawaiian species arrived with the ability to biosynthesize anthocyanins,  rather than de nova evolution of the necessary  enzymes once the lineage became established in Hawaii.  If one  assumes that white and green corollas result from fixation as homozygous recessives, but at different loci, then when the two are crossed, the hybrid is heterozygous at each locus and  211  red is restored.  Complicating these hypotheses somewhat is  the fact that later crosses between kalalauensis produced light green,  almost white corollas,  though not as white as L. glutinosa. epistasis may be involved.  glutinosa and j.  .  This suggests that  In addition, there are the red and  the white flowered siblings of the cross of remyi subsp. remvi to consider.  .  glutinosa X L.  Plants with white flowers  have not been collected from populations of red—flowered species,  suggesting that the “allele” for white is missing.  If white is expressed only as a homozygous is dominant to white,  recessive, and red  then progeny between L. cilutinosa and  red-flowered species should always be red. that the “allele” for white is present, populations of red-flowered species.  It is possible  but very rare in  It would appear that the  inheritance of corolla color in Lysimachia is complex, requiring further, and extensive studies to dissect its various components.  Progeny analysis of F s and backcrosses 1 2  could provide necessary information to interpret the genetic control of corolla pigmentation.  4.4.2.  Summary  Although not all species could be included in the crossing program,  it appears that speciation in Hawaiian  Lysimachia has not been accompanied by pre—zygotic or post— zygotic reproduction isolation.  Geographical separation alone  appears to be the only means of pre-zygotic reproductive isolation as even morphologically distinct species from  212  different islands and different ecosystems could be crossed to produce fertile seed.  Interspecific hybrid fertility suggests  that post-zygotic reproductive isolation is also absent. Observations of xneiosis in hybrids, the fertility of F 2 hybrids and the inclusion of more species into the crossing program are needed to test these hypotheses further.  213  Chapter 5 Summary 5.1  Dissertation findings The objectives of this dissertation were to produce a  taxonomic revision, to estimate the degree of genetic divergence that has accompanied adaptive radiation of the endemic Hawaiian Lysimachia, and to determine if species are isolated by post- or pre-zygotic reproductive barriers.  The  16 species and 4 subspecies of Lysimachia recognized in the taxonomic treatment places the genus within the 20 most speciose groups in Hawaii. unknown species,  The discovery of three previously  since 1987,  suggests that more species remain  to be found in areas previously unvisited by botanists. The extensive morphological variation of Lysimachia, with little,  if any,  allozyme divergence among species,  is similar  to that seen in other studies of Hawaiian congeners.  As is  the fact that geographical isolation alone appears to be an adequate isolating mechanism, because pre—zygotic and post— zygotic isolating mechanisms appear to be lacking. Interisland and intraisland founder events perhaps beginning on Kauai, or on an older island, to progressively younger islands, has resulted in taxa that are morphologically distinct, are usually restricted to a single island and occupy narrow ecological niches.  Following each dispersal event,  ecological and morphological divergence from the ancestral population is potentially relatively rapid because in small populations the effects of random genetic drift will be  214  maximized and there is a greater potential for novel genetic recombinations to become fixed. A combination of factors may have contributed to the fact that morphological divergence among species is greater than measures of genetic divergence (i.e. genetic identities and hybrid fertility), tend to indicate.  Among these are :  1)  the  low genetic diversity of the initial founding population in Hawaii; 2) 3)  random genetic drift in populations that are small;  the relatively slow rate of accumulation of mutations at  allozyme loci in comparison to loci that control morphological characters. The present study did not attempt to identify the likely progenitor of the endemic Hawaiian Lysimachia.  Preliminary  investigation of flavonoid composition indicated that j. mauritiana has a similar suite of compounds to the endemic species and there is little if any overlap in the flavonoid profiles of the endemic Hawaiian species and two Asian species.  5.2  Areas for future research With a thorough taxonoinic revision now in place,  of interesting questions can now be addressed.  a number  Among these  are the following: 1)  Use tools such as flavonoids or DNA sequences to  attempt to identify the nearest relative of the endemic Hawaiian Lysimachia.  215  2)  Use the greenhouse collection to investigate the types  of physiological adaptations that have accompanied speciation and the genetic basis for morphological divergence among species and other aspects of divergence. 3)  Chromosome counts of all species, observations of  meiosis in hybrids, and estimates of the fertility of backcrosses and F 2 plants would be useful to confirm that post—zygotic isolating mechanisms are lacking. 4)  Investigate the degree to which speciation has  resulted in divergence in floral fragrances and pollinator specificity.  216 References Aradhya, K.M. D. Mueller-Dombois and T.A. Ranker. 1991. Genetic evidence for recent and incipient speciaiton in the evolution of Hawaiian Metrosideros (Myrtaceae). Heredity 67:129—138. Baker, HG. 1967. Support for Baker’s Law—as a rule. Evolution 21:853—856. Baker, H.G. and P.A. Ccx. 1984. Further thoughts on dioecism and islands. Annals of the Missouri Botanical Garden, 71:244—253. Baldwin, B.G., D.W. Kyhos, J. Dvorak, and G.D. Carr. 1991. Chioroplast DNA evidence for a North American origin of the Hawaiian silversword alliance (Asteraceae). Proceedings of the National Academy of Sciences, 88:1840 —1843. Bawa,  1990. K.S. rainforests. 21:399—422.  Plant-pollinator interactions in tropical Annual Review of Ecology and Systematics,  Bennell, A.P. and C.M. Hu. 1983. The pollen morphology and taxonomy of Lysimachia. Notes from the Royal Botanic Garden, Edinburgh 40(3):425-458. Carlquist, Univ. Carr,  S. 1974. Press.  Island Biology.  New York:  Columbia  1978. G.D. Chromosome numbers of the Hawaiian flowering plants and the significance of cytology in selected taxa. American Journal of Botany 65(2):236 —242.  Carr, G.D. 1985a. Monograph of the Hawaiian Madiinae (Asteraceae): Argyroxiphium, Dubautia, Wilkesia. Allertonia 4(1) :1—123. Carr, G.D. 1985b. Additional chromosome numbers of Hawaiian plants. Pacific Science 39(3):302—306. Carr,  G.D., E.A. Powell and D.W. Kyhos. Self 1986. —incompatibility in the Hawaiian Madiinae (Compositae): an exception to Baker’s Rule. Evolution 40(2):430-434. Carr, G.D. 1985a. Monograph of the Hawaiian Madiinae (Asteraceae): Argyroxiphium, Dubautia, Wilkesia. Allertonia 4(1) :1—123.  217 Carr,  G.D., R.H. Robichaux, M.S. Witter and D.W. Kyhos. 1989. Adaptive radiation of the Hawaiian Silversword Alliance (Coxnpositae—Madiinae): A comparison with Hawaiian picture-winged Drosophila. Pp. 79-98, in Genetics, speciation and the founder principle, eds.L.V. Giddings, K.Y. Kaneshiro, and W.W. Anderson. New York: Oxford Univ. Press.  Carson, H.L. 1987. The process whereby speceis originate. Bioscience 37(10) :715—720. Chen,  F.H. and C.M. Hu. 1979. Taxonomic and phytogeographic studies on Chinese species of Lysimachia. Acta Phytotax. Sinica 17:21—53.  Coffey, V.J. and S.B. Jones. 1980. Biosysteinatics of Lysimachia section Seleucia (Primulaceae). Brittonia, 32(3) :309—322. Cosner, M.E. and D.J. Crawford. Comparison of isozyme 1994. diversity in three rare species of Coreopsis (Asteraceae). Systematic Botany 19(3): 350-358. Crawford, D.J.,T.F. Stuessy, and Silva 0, M. 1987a. Allozyine divergence and the evolution of Dendroseris (Compositae: Lactuceae), on the Juan Fernandez Islands. Systematic Botany 12(3) :435—443. Crawford, D.J., R. Witkus and T.F. Stuessy. 1987b. Plant evolution and speciation of oceanic islands. Pp. 183 -199 in Differentiation patterns in higher plants, ed. K.M. Urbanska. London: Academic Press. Crawford, D.J.,T.F. Stuessy, Lammers, T.G. Silva 0, M., Pacheco, P. 1990a. Allozyme variation and evolutionary relationships among three species of Wahienbergia (Campanulaceae), in the Juan Fernandez Islands. Botanical Gazette l51(l):119—l24. Crawford, D.J. l990b. Plant Molecular Systematics: Macromolecular approaches. New York: John Wiley and Sons. Crawford, D.J.,T.F. Stuessy, Haines, D.W., Cosner, M.B., Silva 0, M., and Lopez, P. 1992. Allozyme diversity within and divergence among four species of Robinsonia (Asteraceae: Senecioneae), a genus endemic to the Juan Fernandez Islands, Chile. American Journal of Botany 79(8) :962—966.  218 Crawford, D.J.,T.F. Stuessy, Rodriguez, R. and Rondinelli,M. 1993. Genetic diversity in Rhaphithamnus venustus (Verbenaceae), a species endemic to the Juan Fernandez Islands. Bulletin of the Torrey Botanical Club 120(1) :23—28. Cronquist, A. 1981. An integrated system of classification of flowering plants. Columbia University Press, New York. Davis, P.H. and V.H. Heywood. 1973. Principles of Angiosperm Taxonomy. Robert E. Krieger Publishing Company, New York. Dejoode, D.R. and J.F. Wendel. 1992. Genetic diversity and origin of the Hawaiian islands cotton, Gossypium tomentosum. American Journal of Botany 79(1l):1311 —1319. Degener, 0. and I. Degener. 1983. Plants of Hawaii National Parks illustrative of plants and customs of the South Seas. Page proof of ed. 3. Publ. privately: Volcano: Hawaii, pp. 389—393. Federov, A., ed. 1974. Chromosome numbers in flowering plants. Koenigstein, West Germany: Koeltz Science Publishers. Forbes, C.N. 1909. Some new Hawaiian Plants. Occasional Papers of the Bernice P. Bishop Museum, 4(3):222-223. Forbes, C.N. 1914. New Hawaiian Plants-IV. Papers of the Bernice P. Bishop Museum,  Occasional 6(1):39.  Forbes, C.N. 1916. New Hawaiian Plants—V. Occasional Papers of the Bernice P. Bishop Museum, 6(3):74-75. Fosberg, F.R. 1983. Derivation of the flora of the Hawaiian Islands. Pp. 396—408, in A natural history of the Hawaiian Islands, ed. Alison Kay. Honolulu: Univ. of Hawaii Press. Gagne, W.C. and L. C. Cuddihy, 1990. Vegetation. Pp. 45-114, in Manual of the Flowering Plants of Hawai’i, eds. Wagner, W.L., D.R. Herbst, and S.H. Sohmer. Honolulu: Univ. of Hawaii Press, Bishop Museum Press. Ganders, F.R. and K.M. Nagata. 1984. The role of hybridization in the evolution of Bidens. Pp. 179-194, in Plant Biosystematics, ed. W.F. Grant, Toronto: Academic Press.  219 Ganders, F.R. 1989. Adaptive radiation in Hawaiian Bidens. Pp. 99—112, in Genetics, speciation and the founder principle, eds. L.V. Giddings, K.Y. Kaneshiro, and W.W. Anderson. New York: Oxford Univ. Press. Goldblatt, P., ed. Index to plant chromosome numbers 1981. 1975-1978. Monographs in systematic botany from the Missouri Botanical Garden. Vol. 5. Missouri Botanical Garden, St. Louis. Goldblatt, P., ed. 1984. Index to plant chromosome numbers 1979-1981. Monographs in systematic botany from the Missouri Botanical Garden. Vol. 8. Missouri Botanical Garden, St. Louis. Goldblatt, P., ed. 1985. Index to plant chromosome numbers 1982-1983. Monographs in systematic botany from the Missouri Botanical Garden. Missouri Vol. 13. Botanical Garden, St. Louis. Goldblatt, P., ed. 1988. Index to plant chromosome numbers 1984—1985. Monographs in systematic botany from the Missouri Botanical Garden. Vol. 23. Missouri Botanical Garden, St. Louis. Goldblatt, P., and D.E. Johnson, eds. 1990. Index to plant chromosome numbers 1986—1987. Monographs in systematic botany from the Missouri Botanical Garden. Vol. 30. Missouri Botanical Garden, St. Louis. Gornall, R.J. and B.A. Bohm. 1980. Flavonoids and the taxonomy of Boykinia and related genera. Canad. J. Bot. 58:1768—1779. Gottlieb, L.D. 1977. Electrophoretic evidence and plant systematics. Annals of the Missouri Botanical Garden 64(2) :161—180. Gottlieb, L.D. in plants.  1984. Genetics and morphological evolution The Am. Nat. 123(5):681-709.  Gray, A. 1862. Characters of new or obscure species of plants of monopetalous orders in the collection of the United States South Pacific Exploring Expedition under Captain Charles Wilkes, U.S.N. with occasional remarks, Proceedings American Academy of Arts. 5:328-329. Hamrick, J.L. structure and Plant Portland:  1989. Isozymes and the analysis of genetic in plant populations. Pp. 87-105, in Isozymes Biology, eds. D.E. Soltis and P.S. Soltis. Dioscorides Press.  220 Hamrick, J.L. and M.J.W. Godt. 1989. Allozyme diversity in plant species. Pp. 43-63, in Plant population genetics, breeding and genetic resources, eds. A.H.D. Brown, M.T. Clegg, A.L. Kahier, and B.S. Weir. Sunderland, MA: Sinauer. Handel-Mazzetti, H. 1928. A revision of the Chinese species of Lysimachia, with a new system of the whole genus. Notes, Royal Botanic Garden, Edinburgh, 16:51 -122. Helenurm, K. and F.R. Ganders. 1985. Adaptive radiation and genetic differentiation in Hawaiian Bidens. Evolution 39(4) :753—765. Heller, A.A. 1897. Observations on the ferns and flowering plants of the Hawaiian Islands. Minnesota Bot. Stud. 1:874—876. Hickman, James C. ed. 1993. The Jepson Manual, Higher Plants of California. Berkeley: Univ. of California Press. Hillebrand, W. 1888. Flora of the Hawaiian Islands: a description of their phanerograms and vascular cryptograms. Cariwinter, Heildeberg, Germany; Williams and Norgate, London; B. Westerman and Co., New York. Holmgren, P.K., N.H. Holmgren, L.C. Barnett. Herbariorium, New York Botanical Garden,  1990. Bronx.  Index  Hubbell, T.H. 1968. The biology of islands. Pp. 385—395, in A natural history of the Hawaiian Islands, ed. Alison Kay ed. Honolulu: Univ. of Hawaii Press. Hutchinson, J. 1969. Evolution and phylogeny of flowering plants. Academic Press, London. Huynh, K.L. 1970. Le pollen et la systematique chez le genre Lysimachia (Primulaceae). I: Morpholgie generale du pollen et palynotaxonomie, Candollea 25:267—296. Huynh, K.L. 1971. Le pollen et la systematique chez le genre Lysimachia (Primulaceae). II: Considerations generales, Candollea 26/2:279—295. Ito,  M. and Ono, M. 1990. Allozyme diversity and the evolution of Crepidiastrum (Compositae) on the Bonin (Ogasawara) Islands. The Botanical Magazine, Tokyo 103:449—459.  Kim I.S. and G.D. Carr. 1990a. Cytogenetics and hybridization of Portulaca in Hawaii. Systematic Botany 15(3) :370—377.  221 Kim I.S. and G.D. Carr. 1990b. Reproductive biology and uniform culture of Portulaca in Hawaii. Pacific Science 44(2) :123—129. Knuth, R. 1905. Monograph of Lysimachia, Primulaceae. 309—312, in Das Pflanzenreich. IV. 237, (Heft 22). Ko,  Pp.  S.C., I.T. Im, and Y.S. Kim. A cytotaxonomic study 1986. on the genus Lysimachia in Korea. Korean Journal of Plant Taxonomy 16(3):1—11.  Lammers, T.G. 1990. Campanulaceae. Pp. 420-488, in Manual of the Flowering Plants of Hawai’i, eds.Wagner, W.L., D.R. Herbst, and S.H. Sohmer. Honolulu: Univ. of Hawaii Press, Bishop Museum Press. Leveille, H. 1911. Plantae novae sandwicenses. Spec. Nov. Regni Veg. 10:149—157.  II.  Lewis, P.O. 1993. GeneStat-PC 3.3. Department of Statistics, North Carolina State Univ., Raleigh, Carolina.  Repert.  North  Lowrey, T.K. 1986. A biosystematic revision of Hawaiian Tetramolopium (Compositae: Astereae). Allertonia 4:203 265. Lowrey, T.K. and DJ. Crawford. Allozyme divergence 1985. and evolution in Tetramolopium (Compositae:Asteraeae) on the Hawaiian Islands. Systematic Botany 10(1) :64—72. MacDonald, G.A., A.T. Abbott, F.L. Peterson. 1990. Volcanoes in the Sea. The geology of Hawaii. Univ. of Hawaii Press, Honolulu. Mayer, S.S. 1991. Artificial hybridization in Hawaiian Wikstroemia (Thymelaeaceae). Am. J. Bot. 78(1): 122-130. Moore, R.J., ed. 1974. Index to plant chromosome numbers for 1972. International bureau for plant taxonomy and nomenclature. Netherlands: Oosthoek, Schillema and Holkema. Moore, R.J., ed. 1977. Index to plant chromosome numbers for 1973-1974. International bureau for plant taxonomy and nomenclature. Netherlands: Oosthoek, Schillema and Holkema. Nei, M. 1972. Genetic distances between populations. American Naturalist, 106:283—292. Nei, M. and R.K. Chesser. 1983. Estimation of fixation indices and gene diversities. Annanls of Human Genetics 47:253—259.  222 Nei, M. 1987. Molecular evolutionary genetics. Univ. Press, New York.  Columbia  Ornduff, R. ed. 1967. Index to plant chromosome numbers for 1965. International bureau for plant taxonomy and nomenclature. Netherlands. Ornduff, R. ed. 1968. Index to 1ant chromosome numbers for 1966. International bureau for plant taxonomy and nomenclature. Netherlands. Pukui, N.K. and S.E. Elbert. 1992. New pocket Hawaiian Honolulu: Univ. of Hawaii Press. dictionary. Rabakanandrianina, E. 1980. Infrageneric relationships and the origin of the Hawaiian ednemic genus Lipochaeta (Compositae). Pacific Science 34(1):29-39. Ray,  J.D. 1956. The genus Lysimachia in the new world. Illinois Biological Monographs, 24(3—4), 160 pp. The University of Illinois Press, Urbana.  Richards, A.J. 1986. Allen and Unwin. Rock,  Plant breeding systems.  J.F. 1910. Some new Hawaiian plants. Torrey Botanical Club 37:300-303.  London:  George  Bulletin of the  St. John, H. 1933. Lysimachia, Labordia, Scaevola, Pluchea. Hawaiian Plant Studies 1. Occasional Papers Bernice P. Bishop Museum 10(4)1—10. St. John, H. 1971. Endemic plants of Kipahulu Valley, Maui, Hawaiian Islands, Hawaiian Plant Studies 36. Pacific Science 25:50—52. St.  John, H. 1973. List and summary of the flowering plants in the Hawaiian Islands. Pacific Trop. Bot. Gard. Mem. 1:1—519.  St. John, H. 1987. (Primulaceae), 64(1) :43—50.  Diagnoses of Lysimachia Species Hawaiian Plant Studies 159. Phytologia  St. John, H. Unpublished manuscript. Monograph of Lysimachia (Primulaceae) in Hawaii. Hawaiian Plant Studies 101. Bishop Museum, Honolulu, Hawaii. Selling, 0.11. 1947. Studies in Hawaiian pollen statistics. Part II: The pollens of the Hawaiian Phanerogams. Bernice P. Bishop Museum, Special Publication 38. Bishop Museum, Honolulu, Hawaii.  223 Skottsberg, C. 1944. Vascular plants from the Hawaiian Islands. IV. Phanaerogams collected during the Hawaiian bog survey, 1938. Acta Horti Gothob. 15:429—430. Skottsberg, C. 1953. flowering plants.  Chromosome numbers in Hawaiian Ark. Bot. 3:63-70.  Soltis, D.E., C.H. Haufler, D. C. Darrow, and G. J. Gastony. 1983. Starch gel electrophoresis of ferns: A compilation of grinding buffers, and staining schedules. American Fern Journal 73:9-27. Sun, N. 1987. Genetics of gynodioecy in Hawaiian Bidens (Asteraceae). Heredity 59:327—336. Sun, N. and F.R. Ganders. 1988. Mixed mating systems in Hawaiian Bidens (Asteraceae). Evolution, 42(3):5l6-527. Swof ford, D.L. and R.B. Selander. Biosys-l, a computer 1989. program for the analysis of alleleic variation in population genetics and biochemical systematics (Release 1.7). Illinois Natural History Survey. Tanaka, A. and N. Hizume. 1980. Karyomorphological studies on species differentiation in some species of Lysimachia. La Kromosomo 11—18—19: (515—525). Thomas, R.H. and J.A. Hunt. 1991. The molecular evolution of the alcohol dehydrogenase locus and the phylogeny of Hawaiian Drosophila. Molecular Biology and Evolution 8(5) :687—702. Van Steenis, C.G.G.J. 1972. Flora Malesiana. Wolters -Noordhoff Publishing, Goningen, The Netherlands, 177 —186. Vander Kloet, S.P. 1993. Biosystematic studies of Vaccinium Section Macroelma (Ericaceae) in Hawaii. Pacific Science 47(1) :76—85. Wagner, W.L., D.R. Herbst, and S.H. Sohmer. 1990. Manual of the Flowering Plants of Hawai’i. Honolulu, Hawaii: Univ. of Hawaii Press, Bishop Museum Press. Wawra, H. 1874. Beitrage zur Flora der Hawai’schen Inseln. Flora 57:523. Walker, G. 1990. Geology. Pp. 21-35, in Manual of the Flowering Plants of Hawai’i, eds. Wagner, W.L., D.R. Herbst, and S.H. Sohmer. Honolulu: Hawaii: Univ. of Hawaii Press, Bishop Museum Press.  224 Wendel, J.F. and N.F. Weeden. 1989. Visualization and interpretation of plant isozymes. Pp. 5-45 in Isozymes in plant bioloqy, eds. D.E. Soltis and P.S. Soltis. Portland, Oregon: Dioscorides. Weeden, N.F. and J.F. Wendel. 1989. Genetics of plant isozymes. Pp. 46—72 in Isozymes in plant bioloqy, eds. D.E. Soltis and P.S. Soltis. Portland, Oregon: Dioscorides. Weller, S.G. and A.K. Sakai. 1988. Artificial hybridization in Schiedea (Caryophyllaceae), an endemic Hawaiian genus. American Journal of Botany, 75(6):2l6. Wendel, J.F. and A.E. Percival. 1990. Molecular divergence in the Galapagos Islands-Baja California species pair, Gossypium klotzschianum and G. davidsonii (Malvaceae). Plant systematics and evolution, 171:99-115. Wendel, J.F. and R.G. Percy. 1990. Allozyme diversity and introgression in the Galapagos Islands endemic Gossypium darwiniii and its relationship to continental . barbadense. Biochemical systematics and ecology 18:517 —528. Westphal, L. and G. Wricke. 1991. Genetic and linkage analysis of isozyme loci in Daucus carota L. Euphytica 56: 259—267. Wilkinson, L. 1990. SYSTAT: the system for statistics. Version 5.01 for DOS. Evanston, ILL: SYSTAT Inc. Witter, M.S. and G.D. Carr. 1988. Adaptive radiation and genetic differentiation in the Hawaiian Silversword alliance (Compositae:Madiinae). Evolution 42(6) :1278 1287. Zimmerman, E.C. 1978. Insects of Hawaii, Vol. 9, Microlepidoptera Part I; Monotrysia, Tineoidea, Tortricoidea, Gracillarioidea, Yponomeutoidea, and Alucitoidea. The University Press of Hawaii, Honolulu.  225 Appendix 1 Table A3.1. populations.  Table of allele frequencies of Hawaiian Lvsixnachia See Table 3.1 for population codes.  Locus/ allele(N)  Population  DBIGB  DSECO  GKALR  GKALL  GHONO  KHONO  KAAPU  KMAKA  TPI—1 (29) a 0.000 b 1.000 c 0.000  (10) 0.000 1.000 0.000  (20) 0.000 1.000 0.000  (17) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (29) 0.000 0.950 0.050  (29) 0.000 0.793 0.207  (10) 0.000 0.550 0.450  TPI—2 a b c d e  (25) 0.120 0.040 0.800 0.040 0.000  (9) 0.000 0.000 1.000 0.000 0.000  (20) 0.000 0.000 1.000 0.000 0.000  (17) 0.235 0.000 0.676 0.088 0.000  (30) 0.133 0.000 0.833 0.033 0.000  (28) 0.000 0.000 0.946 0.054 0.000  (29) 0.000 0.000 0.828 0.172 0.000  (10) 0.000 0.000 1.000 0.000 0.000  TPI—3 a b  (25) 0.140 0.860  (9) 0.167 0.833  (20) 0.000 1.000  (17) 0.000 1.000  (30) 0.133 0.867  (28) 0.000 1.000  (29) 0.000 1.000  (10) 0.000 1.000  ADH—1 (22) a 1.000 b 0.000  (9) 1.000 0.000  (20) 1.000 0.000  (19) 1.000 0.000  (30) 1.000 0.000  (16) 1.000 0.000  (13) 1.000 0.000  (9) 1.000 0.000  IDH—1 a b c d e  (16) 0.000 0.125 0.875 0.000 0.000  (7) 0.000 0.000 1.000 0.000 0.000  (20) 0.375 0.275 0.000 0.350 0.000  (18) 0.083 0.194 0.028 0.694 0.000  (25) 0.040 0.160 0.140 0.660 0.000  (29) 0.000 0.310 0.638 0.052 0.000  (28) 0.000 0.036 0.962 0.000 0.000  (10) 0.000 0.000 1.000 0.000 0.000  MDH—3 a b c d  (27) 0.000 0.000 1.000 0.000  (9) 0.000 0.000 1.000 0.000  (19) 0.000 0.000 1.000 0.000  (17) 0.000 0.000 1.000 0.000  (30) 0.000 0.000 1.000 0.000  (30) 0.000 0.000 1.000 0.000  (28) 0.000 0.000 0.964 0.036  (9) 0.000 0.000 1.000 0.000  MDH—4 a b  (27) 0.018 0.981 0.000  (9) 0.056 0.944 0.000  (19) 0.000 1.000 0.000  (17) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (28) 0.089 0.911 0.000  (9) 0.000 1.000 0.000  SKD—l (20) a 0.000 b 1.000 C 0.000  (4) 0.000 1.000 0.000  (9) 0.000 1.000 0.000  (17) 0.088 0.912 0.000  (29) 0.000 1.000 0.000  (19) 0.131 0.868 0.000  (16) 0.000 1.000 0.000  (9) 0.000 1.000 0.000  Table A3.1.  continued on next page  C  226 Table A3.l. continued. Lysimachia populations.  Table of allele frequencies of Hawaiian  Locus! allele (N)  Population  DBIGB  DSECO  GKALR  GKALL  GHONO  KHONO  KAAPU  KMAKA  DIA—l (25) a 0.000 b 1.000  (5) 0.000 1.000  (20) 0.000 1.000  (15) 0.000 1.000  (30) 0.000 1.000  (14) 0.000 1.000  (24) 0.000 1.000  (10) 0.000 1.000  PGM—l (28) a 0.053 b 0.053 C 0.172 d 0.689 e 0.000 f 0.000  (8) 0.000 0.000 0.187 0.500 0.312 0.000  (20) 0.000 0.000 0.000 1.000 0.000 0.000  (19) 0.000 0.000 0.079 0.921 0.000 0.000  (21) 0.000 0.000 0.071 0.929 0.000 0.000  (27) 0.000 0.000 0.130 0.759 0.111 0.000  (22) 0.000 0.000 0.182 0.545 0.273 0.000  (6) 0.000 0.000 0.000 0.833 0.167 0.000  PGI—1 (27) a 0.923 b 0.000 C 0.000 d 0.074  (9) 0.889 0.000 0.000 0.111  (20) 1.000 0.000 0.000 0.000  (19) 1.000 0.000 0.000 0.000  (27) 1.000 0.000 0.000 0.000  (26) 0.000 1.000 0.000 0.000  (29) 0.000 1.000 0.000 0.000  (10) 0.000 1.000 0.000 0.000  Table A3.1.  continued on next page  227 Table A3.1. populations.  Table of allele frequencies of Hawaiian Lysimachia  Locus/ allele (N)  Population  KWAIA  OLIMA  OWAIN  NSPTR  FKONE  FKTHR  FKFIV  FOTWO  TPI—l (29) a 0.000 b 0.948 c 0.052  (9) 0.000 0.889 0.111  (47) 0.000 1.000 0.000  (7) 0.000 0.214 0.786  (13) 0.000 1.000 0.000  (7) 0.000 1.000 0.000  (6) 0.000 1.000 0.000  (26) 0.000 1.000 0.000  TPI—2 a b d e  (28) 0.000 0.000 1.000 0.000 0.000  (8) 0.250 0.000 0.750 0.000 0.000  (43) 0.035 0.000 0.965 0.000 0.000  (7) 0.000 0.000 1.000 0.000 0.000  (12) 0.000 0.000 0.917 0.083 0.000  (6) 0.000 0.000 1.000 0.000 0.000  (5) 0.000 0.000 1.000 0.000 0.000  (26) 0.000 0.000 0.846 0.000 0.015  TPI—3 a b  (28) 0.036 0.964  (8) 0.250 0.750  (43) 0.314 0.686  (7) 0.000 1.000  (12) 0.083 0.917  (6) 0.250 0.750  (5) 0.000 1.000  (26) 0.019 0.981  ADH—1 (20) a 1.000 b 0.000  (9) 1.000 0.000  (32) 1.000 0.000  (30) 1.000 0.000  (9) 1.000 0.000  (6) 1.000 0.000  (5) 1.000 0.000  (21) 1.000 0.000  IDH—1 a b c d e  (29) 0.000 0.000 0.931 0.000 0.069  (9) 0.000 0.444 0.556 0.000 0.000  (47) 0.000 0.681 0.319 0.000 0.000  (8) 0.000 0.125 0.875 0.000 0.000  (12) 0.000 1.000 0.000 0.000 0.000  (7) 0.000 1.000 0.000 0.000 0.000  (6) 0.000 1.000 0.000 0.000 0.000  (27) 0.000 0.000 1.000 0.000 0.000  MDH—3 a b c d  (28) 0.357 0.000 0.643 0.000  (9) 0.000 0.000 1.000 0.000  (44) 0.000 0.034 0.966 0.000  (8) 0.000 0.000 1.000 0.000  (13) 0.000 0.000 1.000 0.000  (7) 0.000 0.000 1.000 0.000  (6) 0.000 0.000 1.000 0.000  (26) 0.000 0.000 1.000 0.000  MDH—4 a b c  (28) 0.161 0.839 0.000  (9) 0.000 1.000 0.000  (35) 0.000 0.714 0.286  (8) 0.000 1.000 0.000  (13) 0.115 0.885 0.000  (7) 0.143 0.857 0.000  (6) 0.000 1.000 0.000  (26) 0.000 1.000 0.000  SKD—1 (15) a 0.000 b 1.000 C 0.000  (9) 0.000 1.000 0.000  (2) 0.000 1.000 0.000  (7) 0.000 1.000 0.000  (10) 0.000 1.000 0.000  (6) 0.000 1.000 0.000  (3) 0.000 1.000 0.000  (6) 0.000 1.000 0.000  Table A3.l.  continued on next page  C  228 Table A3.l. continued. Lysimachia populations.  Table of allele frequencies of Hawaiian  Locus! allele (N)  Population  KWAIA  OLIMA  OWAIN  NSPTR  FKONE  FKTHR  FKFIV  FOTWO  DIA—l (10) 0.000 a 1.000 b  (7) 0.000 1.000  (31) 0.000 1.000  (8) 0.312 0.688  (14) 0.000 1.000  (7) 0.000 1.000  (5) 0.000 1.000  (16) 0.000 1.000  PGM—1 a b c d e f  (22) 0.000 0.000 0.000 0.750 0.250 0.000  (8) 0.000 0.000 0.000 0.187 0.813 0.000  (45) 0.000 0.000 0.000 0.000 0.878 0.122  (4) 0.000 0.000 0.000 0.750 0.250 0.000  (13) 0.000 0.000 0.000 0.808 0.192 0.000  (6) 0.000 0.000 0.000 1.000 0.000 0.000  (3) 0.000 0.000 0.000 1.000 0.000 0.000  (27) 0.000 0.000 0.000 0.870 0.130 0.000  PGI—1 (10) 0.000 a b 1.000 C 0.000 0.000 d  (7) 0.000 1.000 0.000 0.000  (45) 0.000 1.000 0.000 0.000  (8) 0.000 1.000 0.000 0.000  (13) 0.000 1.000 0.000 0.000  (7) 0.000 1.000 0.000 0.000  (6) 0.000 1.000 0.000 0.000  (13) 0.000 1.000 0.000 0.000  Table A3.l.  continued on next page  229 Table A3.1. populations.  Table of allele frequencies of Hawaiian Lysimachia  Locus! allele (N)  Population  FOTHR  HPUKE  HPUKS  HWAIK  HLUAF  HLUAS  WMBLG  WMLIM  TPI—1 (23) 0.000 a 1.000 b 0.000 c  (30) 0.000 0.900 0.100  (6) 0.000 1.000 0.000  (24) 0.000 0.958 0.042  (28) 0.000 1.000 0.000  (12) 0.000 1.000 0.000  (13) 0.000 0.961 0.038  (11) 0.091 0.909 0.000  TPI—2 (23) a 0.000 0.000 b 0.609 c 0.130 d 0.261 e  (29) 0.569 0.000 0.293 0.017 0.121  (6) 0.000 0.000 0.583 0.000 0.417  (25) 0.700 0.000 0.300 0.000 0.000  (17) 0.000 0.000 0.912 0.088 0.000  (11) 0.364 0.000 0.454 0.182 0.000  (11) 0.000 0.000 1.000 0.000 0.000  (12) 0.000 0.000 0.792 0.000 0.208  TPI—3 a b  (23) 0.283 0.717  (29) 0.172 0.828  (6) 0.250 0.750  (25) 0.060 0.940  (17) 0.029 0.971  (11) 0.000 1.000  (11) 0.091 0.909  (12) 0.042 0.958  ADH—1 (23) a 1.000 b 0.000  (30) 0.883 0.117  (6) 1.000 0.000  (23) 1.000 0.000  (15) 1.000 0.000  (12) 1.000 0.000  (12) 1.000 0.000  (6) 1.000 0.000  IDH—1 a b c d e  (23) 0.000 0.000 1.000 0.000 0.000  (30) 0.000 0.783 0.217 0.000 0.000  (6) 0.000 0.417 0.583 0.000 0.000  (24) 0.000 0.646 0.334 0.000 0.000  (22) 0.000 0.367 0.636 0.000 0.000  (11) 0.000 0.318 0.409 0.273 0.000  (13) 0.000 0.269 0.731 0.000 0.000  (12) 0.000 0.000 1.000 0.000 0.000  MDH—3 a b c d  (22) 0.000 0.000 1.000 0.000  (30) 0.000 0.000 0.600 0.400  (6) 0.000 0.000 0.334 0.667  (24) 0.000 0.000 0.750 0.250  (27) 0.000 0.000 0.963 0.037  (11) 0.000 0.000 1.000 0.000  (13) 0.000 0.000 1.000 0.000  (11) 0.000 0.000 1.000 0.000  MDH—4 a b  (22) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (6) 0.000 1.000 0.000  (24) 0.000 1.000 0.000  (27) 0.000 1.000 0.000  (11) 0.000 1.000 0.000  (13) 0.000 1.000 0.000  (11) 0.000 1.000 0.000  SKD—1 (20) a 0.000 1.000 b 0.000 C  (22) 0.023 0.977 0.000  (6) 0.000 1.000 0.000  (24) 0.000 1.000 0.000  (26) 0.000 1.000 0.000  (12) 0.000 1.000 0.000  (7) 0.000 1.000 0.000  (17) 0.000 1.000 0.000  Table A3.1.  continued on next page  C  230 Table A3.l. continued. Lysimachia populations.  Table of allele frequencies of Hawaiian  Locus! allele(N)  Population  FOTHR  HPUKE  HPUKS  HWAIK  HLUAF  HLUAS  WNBLG  WNLIM  DIA—1 (12) 0.000 a 1.000 b  (21) 0.000 1.000  (5) 0.000 1.000  (18) 0.000 1.000  (19) 0.389 0.611  (12) 0.000 1.000  (9) 0.000 1.000  (10) 0.000 1.000  PGM—1 a b c d e f  (21) 0.000 0.000 0.000 1.000 0.000 0.000  (28) 0.000 0.000 0.000 0.393 0.607 0.000  (6) 0.000 0.000 0.000 0.917 0.083 0.000  (20) 0.000 0.000 0.000 0.725 0.275 0.000  (27) 0.000 0.000 0.000 0.889 0.111 0.000  (7) 0.000 0.000 0.000 0.500 0.500 0.000  (9) 0.000 0.000 0.000 0.833 0.167 0.000  (9) 0.000 0.000 0.000 1.000 0.000 0.000  PCI—i a b c d  (23) 0.000 1.000 0.000 0.000  (30) 0.000 1.000 0.000 0.000  (5) 0.000 1.000 0.000 0.000  (24) 0.000 1.000 0.000 0.000  (28) 0.000 1.000 0.000 0.000  (12) 0.000 1.000 0.000 0.000  (11) 0.000 1.000 0.000 0.000  (12) 0.000 1.000 0.000 0.000  Table A3.l.  continued on next page  231 Table A3.1. populations.  Table of allele frequencies of Hawaiian Lysimachia  Locus/ allele ( N)  Population  WNNN2  WNMPP  WMHPS  WNHPL  WMHPY  WMLIS  WMLIL  WMHEL  TPI—l (29) 0.000 a 0.893 b 0.107 c  (30) 0.000 0.933 0.067  (26) 0.057 0.904 0.038  (40) 0.000 0.950 0.050  (13) 0.000 0.577 0.423  (61) 0.000 0.926 0.074  (18) 0.000 0.972 0.028  (50) 0.000 0.990 0.010  TPI—2 (27) 0.018 a 0.000 b c 0.870 0.018 d 0.092 e  (29) 0.000 0.000 0.931 0.000 0.069  (24) 0.000 0.000 0.938 0.062 0.000  (38) 0.000 0.026 0.881 0.026 0.066  (10) 0.000 0.000 0.950 0.050 0.000  (60) 0.017 0.000 0.883 0.008 0.092  (18) 0.056 0.000 0.722 0.028 0.194  (33) 0.030 0.000 0.818 0.182 0.000  TPI—3 a b  (27) 0.074 0.926  (29) 0.172 0.828  (24) 0.125 0.875  (38) 0.013 0.987  (10) 0.050 0.950  (60) 0.108 0.892  (18) 0.194 0.806  (33) 0.015 0.985  ADH—1 (27) a 1.000 b 0.000  (24) 1.000 0.000  (15) 1.000 0.000  (29) 1.000 0.000  (4) 1.000 0.000  (53) 1.000 0.000  (11) 1.000 0.000  (21) 1.000 0.000  IDH—1 a b c d e  (29) 0.000 0.034 0.948 0.017 0.000  (30) 0.000 0.217 0.783 0.000 0.000  (29) 0.000 0.000 1.000 0.000 0.000  (40) 0.000 0.000 0.988 0.012 0.000  (11) 0.000 0.000 1.000 0.000 0.000  (59) 0.000 0.025 0.975 0.000 0.000  (18) 0.000 0.139 0.861 0.000 0.000  (50) 0.000 0.010 0.990 0.000 0.000  MDH—3 a b c d  (28) 0.000 0.232 0.750 0.018  (30) 0.000 0.067 0.917 0.017  (26) 0.000 0.231 0.769 0.000  (28) 0.000 0.125 0.875 0.000  (11) 0.000 0.000 0.954 0.045  (55) 0.000 0.127 0.836 0.036  (17) 0.000 0.294 0.706 0.000  (48) 0.000 0.396 0.604 0.000  MDH—4 a b c  (30) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (26) 0.000 1.000 0.000  (28) 0.000 1.000 0.000  (11) 0.000 1.000 0.000  (55) 0.036 0.964 0.000  (17) 0.029 0.971 0.000  (48) 0.000 1.000 0.000  SKD—1 (27) 0.092 a 0.870 b 0.037 c  (26) 0.000 1.000 0.000  (24) 0.000 1.000 0.000  (9) 0.111 0.889 0.000  (11) 0.000 1.000 0.000  (44) 0.079 0.920 0.000  (8) 0.000 1.000 0.000  (39) 0.051 0.923 0.026  Table A3.l.  continued on next page  232 Table A3.1. continued. Lysimachia populations.  Table of allele frequencies of Hawaiian  Locus! allele(N)  Population  WMMN2  WMMPP  WMHPS  WMHPL  WNHPY  WMLIS  WMLIL  WMHEL  DIA—1 (30) a 0.000 1.000 b  (21) 0.000 1.000  (30) 0.000 1.000  (26) 0.000 1.000  (6) 0.000 1.000  (57) 0.000 1.000  (16) 0.000 1.000  (40) 0.000 1.000  PGM—1 a b c d e f  (25) 0.000 0.000 0.000 0.620 0.380 0.000  (22) 0.000 0.000 0.000 0.636 0.364 0.000  (24) 0.000 0.000 0.000 1.000 0.000 0.000  (23) 0.000 0.000 0.000 0.652 0.348 0.000  (9) 0.000 0.000 0.000 0.778 0.222 0.000  (46) 0.000 0.000 0.054 0.663 0.283 0.000  (16) 0.000 0.000 0.000 1.000 0.000 0.000  (43) 0.000 0.000 0.000 0.698 0.302 0.000  PGI—1 (30) a 0.000 b 1.000 C 0.000 d 0.000  (28) 0.000 1.000 0.000 0.000  (23) 0.000 0.956 0.043 0.000  (21) 0.000 1.000 0.000 0.000  (13) 0.000 1.000 0.000 0.000  (50) 0.000 1.000 0.000 0.000  (20) 0.000 1.000 0.000 0.000  (50) 0.000 1.000 0.000 0.000  Table A3.l.  continued on next page  233 Table A3.1. populations.  Table of allele frequencies of Hawaiian Lysimachia  Locus/ allele (N)  Population  EMPAL  EMHPA  EMLWA  EMKPR  EMKAW  EMKAE  EMWAI  ENKOT  (28) TPI—l 0.000 a 1.000 b 0.000 c  (28) 0.000 1.000 0.000  (31) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (17) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (32) 0.000 1.000 0.000  TPI—2 a b c d e  (27) 0.000 0.000 0.944 0.000 0.056  (17) 0.000 0.000 1.000 0.000 0.000  (30) 0.000 0.000 0.950 0.000 0.050  (30) 0.000 0.000 0.967 0.033 0.000  (16) 0.000 0.000 0.969 0.031 0.000  (26) 0.000 0.000 0.981 0.019 0.000  (30) 0.000 0.000 0.950 0.000 0.050  (32) 0.000 0.000 1.000 0.000 0.000  TPI—3 a b  (27) 0.111 0.889  (17) 0.000 1.000  (30) 0.233 0.767  (30) 0.083 0.917  (16) 0.125 0.875  (26) 0.058 0.942  (30) 0.033 0.967  (32) 0.000 1.000  ADH—1 (21) 1.000 a 0.000 b  (16) 1.000 0.000  (21) 1.000 0.000  (24) 1.000 0.000  (11) 1.000 0.000  (24) 1.000 0.000  (27) 1.000 0.000  (24) 1.000 0.000  IDH—1 a b c d e  (26) 0.000 0.000 1.000 0.000 0.000  (29) 0.000 0.000 1.000 0.000 0.000  (27) 0.000 0.000 1.000 0.000 0.000  (29) 0.000 0.000 1.000 0.000 0.000  (17) 0.000 0.000 1.000 0.000 0.000  (30) 0.000 0.000 1.000 0.000 0.000  (27) 0.000 0.000 1.000 0.000 0.000  (29) 0.000 0.000 1.000 0.000 0.000  MDH—3 a b c d  (28) 0.000 0.964 0.054 0.000  (28) 0.000 0.946 0.036 0.000  (29) 0.000 0.759 0.241 0.000  (29) 0.000 0.845 0.155 0.000  (17) 0.000 0.941 0.059 0.000  (30) 0.000 0.700 0.300 0.000  (30) 0.000 0.950 0.050 0.000  (31) 0.000 1.000 0.000 0.000  MDH—4 a b c  (28) 0.000 1.000 0.000  (28) 0.000 1.000 0.000  (29) 0.000 1.000 0.000  (29) 0.000 1.000 0.000  (17) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (31) 0.000 1.000 0.000  SKD—1 (21) 0.000 a 1.000 b 0.000 c  (20) 0.000 1.000 0.000  (25) 0.000 1.000 0.000  (25) 0.000 1.000 0.000  (7) 0.000 1.000 0.000  (29) 0.155 0.845 0.000  (19) 0.000 0.789 0.210  (28) 0.000 1.000 0.000  Table A3.1.  continued on next page  234 Table A3.1. continued. Lyimachia populations.  Table of allele frequencies of Hawaiian  Locus/ allele (N)  Population  EMPAL  EMHPA  EMLWA  EMKPR  ENKAW  EMKAE  EMWAI  EMKOT  DIA—l (29) 0.000 a 1.000 b  (18) 0.000 1.000  (30) 0.000 1.000  (9) 0.000 1.000  (7) 0.000 1.000  (29) 0.000 1.000  (27) 0.000 1.000  (20) 0.000 1.000  PGM—l a b c d e f  (26) 0.000 0.000 0.000 0.673 0.327 0.000  (24) 0.000 0.000 0.000 1.000 0.000 0.000  (22) 0.000 0.000 0.000 0.727 0.273 0.000  (29) 0.000 0.000 0.000 0.759 0.241 0.000  (16) 0.000 0.000 0.000 0.938 0.062 0.000  (30) 0.000 0.000 0.000 0.600 0.400 0.000  (30) 0.000 0.000 0.000 1.000 0.000 0.000  (30) 0.000 0.000 0.000 0.833 0.167 0.000  PGI—1 (28) 0.000 a b 1.000 0.000 C d 0.000  (28) 0.000 1.000 0.000 0.000  (30) 0.000 1.000 0.000 0.000  (30) 0.000 1.000 0.000 0.000  (9) 0.000 1.000 0.000 0.000  (29) 0.034 0.966 0.000 0.000  (30) 0.000 1.000 0.000 0.000  (32) 0.000 0.969 0.031 0.000  Table A3.1.  continued on next page  235 Table A3.1. populations.  Table of allele frequencies of Hawaiian Lysimachia  Locus/ allele(N)  Population  EMKOB  ENKIP  TPI—l (8) a 0.000 b 1.000 c 0.000  0.000 1.000 0.000  0.000 1.000 0.000  0.000 1.000 0.000  0.000 1.000 0.000  0.000 1.000 0.000  0.000 1.000 0.000  0.000 1.000 0.000  TPI—2 a b c d e  (8) 0.000 0.000 1.000 0.000 0.000  (22) 0.000 0.000 1.000 0.000 0.000  (29) 0.069 0.000 0.724 0.000 0.207  (26) 0.000 0.000 0.904 0.000 0.096  (26) 0.000 0.000 0.692 0.000 0.308  (24) 0.125 0.000 0.479 0.000 0.396  (30) 0.033 0.000 0.433 0.000 0.533  (20) 0.000 0.000 0.725 0.000 0.275  TPI—3 a b  (8) 0.000 1.000  (22) 0.023 0.977  (29) 0.362 0.638  (26) 0.769 0.231  (26) 0.250 0.750  (24) 0.583 0.417  (30) 0.467 0.533  (20) 0.450 0.550  ADH—l (7) a 1.000 b 0.000  (19) 1.000 0.000  (20) 1.000 0.000  (3) 1.000 0.000  (24) 1.000 0.000  (9) 1.000 0.000  (19) 1.000 0.000  (20) 1.000 0.000  IDH—1 a b c d e  (10) 0.000 0.000 1.000 0.000 0.000  (23) 0.000 0.000 1.000 0.000 0.000  (28) 0.071 0.393 0.536 0.000 0.000  (27) 0.370 0.315 0.315 0.000 0.000  (26) 0.019 0.519 0.461 0.000 0.000  (35) 0.100 0.671 0.228 0.000 0.000  (29) 0.000 0.000 1.000 0.000 0.000  (22) 0.364 0.273 0.364 0.000 0.000  MDH—3 a b c d  (10) 0.000 0.950 0.050 0.000  (24) 0.000 0.917 0.083 0.000  (25) 0.000 0.180 0.820 0.000  (27) 0.000 0.000 1.000 0.000  (24) 0.000 0.375 0.625 0.000  (26) 0.000 0.154 0.827 0.019  (28) 0.000 0.250 0.750 0.000  (23) 0.000 0.000 0.913 0.087  MDH—4 (10) a 0.000 b 1.000 C 0.000  (19) 0.000 1.000 0.000  (25) 0.180 0.820 0.000  (27) 0.056 0.944 0.000  (24) 0.000 1.000 0.000  (26) 0.000 1.000 0.000  (27) 0.056 0.944 0.000  (23) 0.283 0.717 0.000  SKD—l (10) a 0.000 b 1.000 C 0.000  (25) 0.000 1.000 0.000  (30) 0.000 1.000 0.000  (13) 0.000 1.000 0.000  (6) 0.000 1.000 0.000  (24) 0.000 1.000 0.000  (27) 0.000 1.000 0.000  (15) 0.000 1.000 0.000  Table A3.l.  continued on next page  (23)  MWAIK  (29)  NiCOLE  (27)  MONIN  (27)  MMAKA  (30)  MMAXI  (30)  MKAWE  (20)  236 Table A3.1. continued. Lysimachia populations.  Table of allele frequencies of Hawaiian  Locus/ allele (N)  Population  EMKOB  EMKIP  NWAIK  MKOLE  MONIN  NNAKA  IV1MAXI  MKAWE  DIA—1 (10) 0.000 a 1.000 b  (24) 0.000 1.000  (20) 0.000 1.000  (20) 0.000 1.000  (24) 0.000 1.000  (30) 0.000 1.000  (27) 0.000 1.000  (21) 0.000 1.000  PGM—1 a b c d e f  (9) 0.000 0.000 0.000 1.000 0.000 0.000  (21) 0.000 0.000 0.000 0.429 0.571 0.000  (29) 0.000 0.000 0.000 0.750 0.250 0.000  (20) 0.000 0.000 0.000 0.525 0.475 0.000  (17) 0.000 0.000 0.000 0.500 0.500 0.000  (27) 0.000 0.000 0.000 0.518 0.481 0.000  (29) 0.000 0.000 0.052 0.948 0.000 0.000  (23) 0.000 0.000 0.000 0.674 0.326 0.000  PGI—1 (9) 0.000 a 1.000 b 0.000 C 0.000 d  (20) 0.000 1.000 0.000 0.000  (29) 0.000 1.000 0.000 0.000  (28) 0.000 1.000 0.000 0.000  (28) 0.000 1.000 0.000 0.000  (37) 0.000 1.000 0.000 0.000  (30) 0.000 1.000 0.000 0.000  (30) 0.000 1.000 0.000 0.000  237  Table A4.l. Pollen stainability in cotton blue of Hawaiian Lysimachia species grown in the greenhouse. See Table 4.1 for collection localities. Plant numbers are Marr collection numbers unless denoted by ““, these are S. Perlman collections. Number to left of decimal of plant number is the collection number, number to right is sibling number. Species Plant No.  Mean  N  glutinosa 254.1 255.1 255.2 2557 255.18 276.1 1304.1 1304.2 1304.3  82 63 78 85 56 94 78 87 83  1 3 3 2 1 2 1 1 1  ka1alauenis l1287.l 11659.l# 257.1 271.1 271.6 273.1  71 87 73 85 84 50  2 1 4 1 2 1  filifolia 11l49.l# 11149.2# l1l49.4#  88 74 87  3 1 2  L. waianaeensis 241.1 242.1 244.2 1297.1  77 99 9 87  1 1 1 2  L. maxima 1299.1 1308.1  35 77  2 1  L.  L.  L.  L.  % Pollen Stainability SD Mm Max  10 6 7  54 72 79  75 86 90  1  93  94  1  70  72  3  70  79  3  81  86  4  86  93  3  85  90  1  86  88  7  30  40  72  96  83  87  89 83  90 85  remyi subsp. kipahuluensis (West Maui) 432.1 85 3 12 436.102 84 1 436.104 85 2 3 436.107 55 1 436.108 89 2 0 436.201 84 2 1  Table A4.l.  continued on next page.  238  Table A4.l. continued. Pollen stainability in cotton blue of Hawaiian Lysimachia species. Species Plant No. .  remvi subsp. 372.201 374.2 377.2 377.3 380.2 386.2 386.3 387.201 387.202 387.203 388.2 388.3 388.4 393.4 393.5 393.6 393.7 395.106 395.201 395.202 395.205 399.2 399.3 399.4 399.8 400.1 1306.4 1306.7  Table A4.1.  Mean subherbacea 55 66 46 38 59 97 97 97 95 86 76 81 89 97 94 85 55 53 95 90 96 92 88 93 92 94 79 94  N 1 1 1 2 2 1 2 2 2 1 4 1 2 1 2 2 1 1 1 2 2 1 2 1 2 2 2 1  % Pollen Stainability SD Mm Max  9 3  31 57  44 62  0 0 2  97 97 94  98 97 97  11  62  87  8  83  95  4 8  91 79  97 92  6 0  85 95  94 97  1  87  89  5 2 7  89 92 74  96 95 85  continued on next page.  239  Table A4.1. continued. Pollen stainability in cotton blue of Hawaiian Lysiniachia species. Species Plant No. .  remyi subsp. 346.1 346.2 346.101 346.102 346.105 347.1 347.4 363.1 350.1 350.2 350.3 354.1 354.5 355.103 355.201 356.201 356.202 357.1 357.2 357.3 359.1 361.2 361.6 362.1 362.2 362.3 362.5 362.6 362.7 362.8 410.101 411.3 411.4 411.5 411.20 413.2 415.2 416.1 416.2 424.1 829.1 888.1 919.1  Table A4.1.  Mean remyi 82 77 78 91 93 81 87 82 94 72 93 95 94 78 96 85 67 83 91 89 75 68 69 86 91 80 90 95 76 89 86 96 96 94 88 74 90 97 94 97 90 95 92  N 2 1 2 1 2 2 2 1 2 4 1 1 2 3 2 2 1 1 1 1 3 4 1 1 2 2 2 2 2 2 1 1 3 1 1 3 1 1 2 1 1 1 1  % Pollen Stainability SD Mm Max 5  78  86  18  65  91  4 17 4  89 69 84  96 94 90  0 10  93 64  94 88  1 14 1 3  93 68 95 83  95 95 97 88  5 5  69 65  81 75  0 17 1 2 8 1  91 68 89 93 70 88  92 93 91 97 82 90  0  96  97  18  53  89  3  92  96  continued on next page.  240  Table A4.1. continued. Pollen stainability in cotton blue of Hawaiian Lysixnachia species. Species Plant No. L.  L.  L.  remyi subsp. 935.1 935.2 1302.1 1303.1 remyi subsp. 336.1 337.2 338.1 338.5 343.29 343.31 344.2 344.4 344.6 remyi subsp. 281.7 281.9 281.15 281.23 282.1 282.3 282.4 282.5 282.6 282.7 282.9 282.10 282.11 282.12 283.3 283.7 284.4 284.6 285.1 285.2 285.3 285.4 285.5 290.4 291.1 291.3 291.4 291.201  Table A4.l.  Mean  N  % Pollen Stainability SD Mm Max  remyi (continued) 98 1 87 1 80 1 88 1 caliginis 36 1 92 2 1 57 2 9 3 80 13 95 1 80 1 57 1 70 1 67 1 kipahuluensis (East Maui) 57 2 23 83 3 4 97 1 92 2 2 46 1 96 1 84 2 4 88 3 5 93 1 53 1 90 1 97 1 84 1 96 1 88 2 3 81 1 57 3 4 27 1 96 1 71 4 9 77 3 4 78 1 84 1 82 1 88 3 3 34 3 3 89 1 91 2 1  continued on next page.  91 50 72  93 64 96  41 79  73 87  90  93  81 85  87 94  86  91  52  61  62 72  82 82  84 30  91 37  90  93  241  Table A4.1. continued. Pollen stainability in cotton blue of Hawaiian Lysimachia species. Species Plant No.  Mean  N  % Pollen Stainability SD Mm Max  Plants from hybrid swarm between j. remyi subsp. kipahuluensis and L. remyi subsp. calicrinis 296.2 90 2 89 92 2 308.5 75 2 74 75 0  242  Table A4.2. Pollen stainability of artificial hybrids between species of endemic Hawaiian Lysimachia. Female Parent  Male Parent  L. glutinosa X L. 255.2 257.1 255.2 257.1 255.2 257.1  Po11n Stainability Mean N SD Mm Max  Sibling Number kalalauensis 1 2 3  75 83 66  4 3 2  11 9 10  59 62 58  85 97 73  kalal auensis X L. criutinosa 271.1 255.2 1 271.1 255.2 3 271.1 255.17 1 255.17 271.1 2  0 18 29 90  1 2 4 2  5 2 3  7 27 87  29 32 92  glutinosa X i.. filifolia 1304.1 11149.1 1  23  1  L. filifolia X L. 11149.1 255.1  glutinosa 1  89  2  1  88  90  L. glutinosa X . 1304.1 432.1  remyi subsp. 1  L.  .  r. subsp. kipahuluensis 436.107 255.2 1  L.  L.  glutinosa X . 255.2 281.7  .  1  subsp.  L. r. subsp. kipahuluensis 281.2 255.2 1 281.2 255.2 2 285.2 255.2 1 285.2 255.2 3 285.2 255.2 4 285.2 255.2 5 285.2 255.2 6 285.2 255.2 7 285.2 255.2 8 Table A4.2.  kipahuluensis 20 1  (West Maui) 13  X L. 1  kipahuluensis 48 2 (East Maui) 29 72 73 38 39 45 54 32 44  continued on next page.  (West Maui) glutinosa  (East Maui) 18 35  X L. 1 1 2 1 3 1 2 2 3  60  glutinosa 5  69  77  8  34  48  19 12 8  41 24 35  67 41 49  243  Table A4.2. continued. Pollen stainability of artificial hybrids between species of endemic Hawaiian Lysimachia. Female Parent  Male Parent  L. glutinosa X L. 255.8 359.1 359.1 255.8 255.1 424.1 255.1 424.1 254.4 1301.1  Sibling Number  r.  1 2 1 2 1  subsp.  . subsp. rernyi X L. 355.103 255.2 1 356.201 255.2 1 356.201 255.2 2 361.2 255.2 1 361.2 255.2 2 410.101 255.2 1 410.101 255.2 2  .  • glutinosa X 255.8 372.201 255.2 393.5 255.2 393.5 255.16 395.205 .  .  1 1 2 1  j. . subsp. subherbacea X 372.1 255.2 1 387.202 255.2 1 393.3 255.2 1 393.3 255.2 2 393.3 255.2 3 393.3 255.2 4 393.5 255.2 4 399.4 255.2 1 L. r. subsp. caliginis X L. 338.1 255.2 1 343.17 255.2 1 343.17 255.2 2  78 42 64 38 57  3 3 4 1 3  70 19 37 24 61 51 46  1 2 1 1 2 1 2  subherbacea 78 42 38 24  1 3 3 2  glutinosa 26 66 29 51 42 46 38 37  1 3 4 3 3 1 1 1  .  glutinosa 71 71 57  subsp.  L. r. subsp. subherbacea X 400.1 271.1 2 Table A4.2.  reinyi  glutinosa  subsp.  j. Jcalalauensis X j. . 271.1 387.202 1  % Pollen Stainability Mean N SD Mm Max  .  2 4 3  subherbacea 3 82 kalalauensis 71 1  continued on next page.  10 4 34  68 37 28  88 44 97  22  32  75  3  18  21  7  56  66  —  46  46  3 5 1  40 35 24  45 43 25  20 7 9 2  53 20 41 41  89 37 57 44  4 20 7  68 48 51  73 93 64  7  78  89  244  Table A4.2. continued. Pollen stainability of artificial hybrids between species of endemic Hawaiian Lysimachia. Female Parent  Male Parent  Sibling Number  % Pollen Stainability Mean N SD Mm Max  L. filifolia X . kalalauensis 11149.1 271.1 1 L. filifolia X . 11149.1 290.4 11149.1 290.4  .  1 2  1  41  subsp. kipahuluensis 1 56 2 81  (East Maui) 10  74  89  L. . subsp. kipahuluensis 285.2 11149.1 1 285.2 11149.1 2  (East Maui) 77 77  X j. 3 1  filifolia 5 72  82  L. r. subsp. kipahuluensis 436.108 11149.1 1 436.108 11149.1 2 436.201 11149.1 1  (West Maui) 92 87 95  X L. 2 2 1  filifolia 5 88 2 85  95 89  L. . subsp. kipahuluensis 436.101 362.1 1  (West Maui) 83  X  • r. subsp. remvi X . r. 362.1 436.201 1  subsp. kipahuluensis 73 2 11  L. r. subsp. kipahuluensis subherbacea 436.101 399.6 1  (West Maui)  .  .  1  X  .  r.  subsp.  remyi  (West Maui) 65 81 subsp.  63  2  15  53  74  L. . subsp. remyi X j. kalalauensis 410.101 271.1 1 77 410.101 271.1 2 93  2 2  9 2  71 91  84 94  j. . subsp. remyi X L. 355.103 11149.1 1 355.103 11149.1 2 355.201 11149.1 1 355.1 11149.1 1 361.2 11149.1 1 361.2 11149.1 3  3 1 1 1 4 1  13  45  70  15  63  93  Table A4.2.  filifolia  58 94 57 79 78 67  continued on next page.  245  Table A4.2. continued. Pollen stainability of artificial hybrids between species of endemic Hawaiian Lysimachia. Female Parent  Male Parent  Sibling Number  % Pollen Stainability Mean N SD Mm Max  L. . subsp. kipahuluensis 282.4 257.1 1 257.1 282.4 2 282.13 257.1 1  (East Maui) 94 80 17  L. . subsp. reinvi X i. 362.1 377.1 1 362.1 377.1 2 362.7 377.1 1 362.8 387.202 1 361.2 395.201 1 361.2 395.202 1 1302.1 393.5  subsp.  .  L. r. subsp. kipahuluensis 284.6 350.2 1 282.4 359.1 1 282.4 362.1 1 282.4 362.1 2 285.2 362.1 1 285.2 415.1 3 285.2 415.2 1 285.2 415.2 2 285.2 415.2 3 285.2 415.2 4 subsp. remyi X L. L. 346.101 282.5 1 346.101 290.1 1 361.2 282.1 362.1 282.4 1 362.3 290.4 1 362.4 290.4 2 .  Table A4.2.  r.  kalalauensis 11  subherbacea 84 2 2 58 1 1 31 1 78 53 2 11 44 2 16 79 1  (East Maui) 85 82 85 96 72 39 70 78 93 89 subsp.  X L. 1 2 1  X L. 1 1 4 1 1 1 2 1 3 1  .  continued on next page.  1 1 2 1 2 1  87  83  85  45 32  61 55  subsp.  remvi  7  74  91  8  65  76  4  89  96  (East Maui)  ..  85 81 57 91 64 95  72  11  49  65  0  64  64  246  Table A4.2. continued. Pollen stainability of artificial hybrids between species of endemic Hawaiian Lysimachia. Female Parent  Male Parent  Sibling Number  L. . subsp. kipahuluensis subherbacea 372.201 285.2 1 290.4 372.201 1 281.25 377.1 2 282.1 377.1 1 282.1 377.1 2 281.25 393.8 1 281.25 393.8 2 281.25 393.8 3 281.25 393.8 4 281.25 393.8 5 281.25 393.8 6 r. subsp. subherbacea X Maui) 372.201 282.5 1 393.4 282.10 1 393.4 282.10 2 395.201 282.1 1 372.201 285.2 1  .  L. . subsp. caliginis X L. 337.3 11149.1 338.1 11149.1 5 338.5 11149.1 2 338.5 11149.1 3 L. . subsp. caliginis X 338.1 354.5 1 340.1 354.5 1 Table A4.2.  .  % Pollen Stainability Mean N SD Mm Max (East Maui) 82 9 77 77 62 84 85 84 83 81 83 .  .  subsp.  X  .  1 1 1 2 2 3 3 5 1 1 2  2 2 2 2 1  filifolia 93 95 95 90  1 1 3 2  subsp. remyi 85 2 90 1  continued on next page.  subsp.  9 3 2 2 7  71 60 82 84 75  83 65 85 87 90  2  81  84  kipahuluensis  83 87 71 93 83  .  .  (East  5 1 5 1  79 86 68 92  87 87 75 93  4 4  91 87  98 92  1  85  86  247  Table A4.2. continued. Pollen stainability of artificial hybrids between species of endemic Hawaiian Lysimachia. Female  Parent  Male Parent  Pn11n tinabi1itv Mean N Mm SD Max  Sibling Number  L. r. subsp. caliginis X 338.1 377.1 2 338.1 386.2 1 338.1 386.2 2 336.1 387.202 1 336.1 387.202 2 340.1 387.202 1  .  .  L. r. subsp. subherbacea X j.. 377.1 338.1 1 377.1 338.1 2 377.1 338.1 3 377.1 338.1 4 L. r. subsp. remyi X j. 362.1 377.1 1 362.1 377.1 2 362.7 377.1 1 362.8 387.202 1 361.2 395.201 1 361.2 395.202 1 1302.1 393.5  .  r. subsp. subherbacea X 393.8 350.2 1 393.8 350.2 2 393.8 350.2 3 393.8 350.2 5 393.8 350.2 6 393.4 362.1 1 393.4 362.1 2 377.1 1302.1 1 377.1 1302.1 2  L.  subsp. subherbacea 40 1 82 2 5 78 87 1 45 2 6 41 35 2 34 64 3 26 33 .  subsp.  .  L. r. subsp. subherbacea X L. 377.1 11149.1 400.1 11149.1 1 400.1 11149.1 2  r.  subsp. 64 73 78 45  caliciinis 4 20 43 1 1 1  subherbacea 84 2 2 58 1 31 1 78 1 53 2 11 44 2 16 79 1 subsp. 83 85 82 86 67 89 91 73 38  remyi 2 9 2 9 1 1 1 2 1 2 1 2 1 2 6  filifolia 98 83 2 79  8  85 49 36 82 84  83  85  45 32  61 55  76 78  89 91  88 90  73 35  89 92 74 42  73  84  

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.831.1-0088322/manifest

Comment

Related Items