CHEMOTAXONOMY OF THE CUPRESSACEAE.
by
P. A. GADEK
School of Botany
University of New South Wales
February, 1986.
Thesis submitted for the degree of Doctor of Philosophy. DECLARATION.
"I hereby declare that this thesis is my own work and that, to the best of my knowledge and belief, it contains no material previously published or written by another person nor material which to a substantial extent has been accepted for the award of any other degree or diploma of a university or other institute of higher learning, except where due acknowledgement is made in the text of the thesis." ABSTRACT.
The aim of this thesis was to extend the data base of the
Cupressaceae s.s. by a survey of leaf biflavonoids, and to apply this, with other available data, to a critical reassessment of the current tribes and subfamilies.
The biflavonoids in ethanolic leaf extracts of representatives of all genera were analysed by thin layer chromatography. Compounds were identified by chromatographic comparisons with a range of standards, by colour of fluorescence after spraying with an ethanolic solution of AlCu, shifts in UV absorption spectra, as well as permethylation. TLC of permethylated raw extracts proved to be a sensitive method of detecting the range of biflavonoid skeletons present in each species.
While there was a high degree of uniformity in the biflavonoid series present in most genera, marked discontinuities were detected within
Calocedrus, Chamaecyparis and Thuja. A reassessment of these genera on a broad -range of available data led to the resurrection of two genera,
Heyderia and Callitropsis, a redefinition of Thuja as a monotypic genus and the erection of a new genus, Neothuja, to encompass the species removed from Thuja.
Outgroup comparisons indicate that the presence of the amentoflavone series is pleisomorphic in the family, while the presence of the cupressuflavone and taiwaniaflavone series, and the absence of the hinokiflavone series are apomorphs.
Marked discontinuities in biflavonoid patterns did not correlate with current tribal and subfamily groupings, but allowed the recognition of 4 groups, 3 of which are defined by synapomorphs. A consideration of available data indicated some support for these groups, but there are many gaps in the data base which must be filled before a more satisfactory taxonomy can be constructed.
A method for the localisation of biflavonoids in fresh leaf sections using aluminium chloride-induced fluorescence is described. This revealed that biflavonoids accumulated in the cuticle and the cutinised outer periclinal and anticlinal walls of the epidermis in a broad range of biflavone-containing taxa. Confirmation was gained by chromatographic analysis of epidermal peels, cuticular scrapings and middle-leaf tissue fractions in Agathis robust& It is postulated that biflavones serve a protective role against invasion of the leaf by microorganisms and/or attack by leaf-eating insects. ACKNOWLEDGEMENTS.
I am indebted to Dr C. J. Quinn for his advice, assistance and considerable encouragement which enabled this thesis to be completed.
Thanks are due also to:
Assoc. Prof. A. E. Ashford for assistance and advice on GMA
sectioning and histochemical staining;
Dr I. McFarlane for advice on the techniques of HPLC, Mass
Spectroscopy and Nuclear Magnetic Resonance Spectroscopy;
Dr L. A. S. Johnson, Director, for permission to sample specimens in
the Royal Botanic Gardens and the National Herbarium of NSW;
Mr D. Symon, Waite Institute, University of Adeliade, South
Australia, for providing specimens from the Waite Arboretum;
Dr G. D. McPherson, Herbarium, Missouri Botanic Gardens, USA, for
collections of Neocallitropsis;
and to the staff and postgraduate students of the School of Botany,
University of New South Wales. STATEMENT.
It should be noted that, according to article 29 of the International Code of Botanical Nomenclature, this thesis does not qualify as an effective and valid publication. Therefore, descriptions of new genera or new combinations contained herein are not validly published. TABLE OF CONTENTS.
Declaration.
Abstract.
Acknowledgements.
Statement.
1. TAXONOMIC HISTORY.
2. BIFLAVONOIDS.
2.1 Introduction.
2.2 Distribution and use in Taxonomy.
2.3 Distribution in the Cupressaceae.
3. TECHNIQUES.
3.1 Plant material.
3.2 Extraction and Isolation.
3.2.1 Thin Layer Chromatography.
3.2.2 High Performance Liquid Chromatography.
3.3 Identification.
3.3.1 Reference Compounds.
3.3.2 Determination of Interflavonoid Linkage.
3.3.2.1 Reaction with Aluminium Chloride.
3.3.2.2 Permethylation.
3.3.2.3 Reference Permethyl Ethers.
3.3.2.4 High Performance Liquid Chromatography.
3.3.2.5 Centrifugally Accelerated Thin Layer Chromatography.
3.3.2.6 Autofluorescence.
3.3.2. 7 Mass Spectroscopy.
3.3.2.8 Nuclear Magnetic Resonance Spectroscopy. 3.3.3 Determination of Methylation Patterns.
3.3.3.1 Partial Demethylation.
3.3.3.2 UV-Spectroscopy.
4. DISTRIBUTION OF BIFLAVONOIDS IN THE CUPRESSACEAE.
4.1 Reliability of Biflavonoid Pattern within Species.
4.2 Definition of Character-states.
4.3 Biflavonoid Patterns.
5. CHEMOTAXONOMY.
5.1 Polarity of chemical characters.
5.2 Correlation with existing generic boundaries.
5.3 Biflavonoid Patterns in relation to Taxonomy above the Generic
Level.
6. OTHER DATA SOURCES.
6.1 Introduction.
6.2 Leaf characters.
6.3 Cone morphology.
6.4 Reproductive Biology.
6.5 Wood Histology.
6.6 Chemistry
6.6.1 Tropolones.
6.6.2 Terpenes.
6.6.3 Flavonoids.
6.7 Pollen Morphology.
7. CONCLUSIONS.
8. LOCALISATION OF BIFLAVONOIDS IN LEAVES.
8.1 Introduction.
8.2 Materials and Methods.
8.2.1 Leaf Anatomy
8.2.2 Fluorescence Microscopy. 8.2.3 Extraction of Biflavonoids.
8.3 Results.
8.3.1 Leaf anatomy of Agathis robusta.
8.3.2 Fluorescence Microscopy.
8.3.2.1 Autofluorescence.
8.3.2.2 Aluminium Chloride Induced Fluorescence.
8.3.3 Chromatography of Extracts.
8.3.4 Survey of Other Taxa.
8.4 Discussion.
REFERENCES.
APPENDIX.
PUBLICATIONS. TABLES.
1. Examples of Taxonomic Groupings ••••••••••••••••••••••••••• Section 1.
2. Occurrence of Biflavonoids in the Coniferales ••••••••.••••••••••. 2.2
3. Distribution of Biflavonoids in the Cupressaceae ••••••••••••••••• 2.3
4. Chromatographic Characteristics and Sources of
Reference Samples of Parental Compounds and their
Partial Methyl Ethers •••••....••••.•..•••••.••••••••.....•••••. 3.2.1
5. HPLC of Methyl Ethers of Amentoflavone and
Hinokiflavone using a Diol column .•••.••••...•.•.••..••••.••••• 3.2.2
6. Chromatographic and Spectral Characteristics of
Permethy 1 Ethers...... • ...... • ...... 3. 3. 2. 3
7. HPLC of Permethyl Ethers using Reverse Phase (C1s)
and Normal Phase (silica) columns .•••.•••••.••••••••••••••••••• 3.3.2.4
8. Chemical Shifts of Protons in Cupressuflavone
HexaJDethy 1 Ether. . • . . . . . • . . • ...... • . • . . • • . • • . . . • . • . . • • ...... • 3. 3. 2. 8
9. Reliability of Biflavonoid Pattern ••.•.•••••••••••••••••••••••••• 4.1
10. Biflavonoid Derivatives Detected in Leaf Extracts ••••••••••••••• 4.3
11. Biflavonoid Permethyl Ethers Detected in the
Permethylated Raw Leaf Extracts •.•••••..•••..••••••••••••••••• 4.3
12. Generic Groupings suggested by Biflavones ••••••••••••••.••••.••• 5.3
13. Distribution of Character-states of Other Data Sources •••••••••• 6.2
14. Results Obtained from two-dimensional Paper
Chromatography of extracts of Agathis robusta leaves •••••••••• 8.3.3
15. Occurrence of Biflavonoid Permethyl Ethers in the
Permethylated Raw Extracts of Agathis robusta leaves ••••...... 8.3.3 FIGURES.
1. Examples of the Different Interflavonyl
Linkages of Biflavonoids ...... 2.1
2. Biosynthetic Pathways of Biflavonoids •••••••....•••••••••••••••••• 2.1 3. HPLC trace of the Permethylated Raw Extract
of Cal ocedrus decurrens...... 3. 3. 2. 4 4. H-NMR trace of Cupressuflavone Hexamethyl Ether •••••.••••••••••••• 3.3.2.8 5. Products from Partial Demethylation ••••••••••••••••••.•.••••••.••• 3.3.3.l
6. UV spectral data of Cupressuflavone 77" dimethyl ether •••••••••••• 3.3.3.2
7. UV spectral data of Amentoflavone 7"4"' dimethyl ether •••••••••••• 3.3.3.2
8a. Leaf margin of Agathis robusta as seen in
transverse sect ion...... 8. 3. 1
8b. Diagram illustrating the epidermal and hypodermal
wall systems of the adaxial surface of the leaf of
Agathis robusta ...... B. 3.1
9-14 Light microscopy of the adaxial region of
Agathis robusta leaves in transverse section •.•••••••••...••••• 8.3.l 15-20 Fluorescence microscopy of fresh transections or.
epidermal peels from the adaxial surface of
Agathis robusta leaves •••••••••.••.••••••....••••••...••.•.•• 8. 3. 2
21-28 Fluorescence microscopy of fresh sections of leaves
of other taxa...... 8. 3. 4
29-34 Colour photographs of induced fluorescence ••••.•••••••••••.••.. 8.3.4
Tokyo Tokyo
doubtful doubtful
Bot. Bot.
part) part)
or or
Meg. Meg.
in in
t t
(Bot. (Bot.
(Svensk. (Svensk.
known known
is is
t t
part) part)
• •
t t
(1943) (1943)
• •
in in
cleayana) cleayana)
NeacallitroJ)llis) NeacallitroJ)llis)
Berl, Berl,
Platycl11dus) Platycl11dus)
Juniperus Juniperus
Zucc. Zucc.
Spach) Spach)
Kurz) Kurz)
Hayata Hayata
Florin Florin
Thoau Thoau
Kot. Kot.
.. ..
(• (•
Florin Florin
("' ("'
132) 132)
L. L.
et et
( = = (
f. f.
Florin Florin
Ii Ii
& &
f. f.
Mig. Mig.
Endl. Endl.
Muell Muell
Spach Spach
Hort. Hort.
24: 24:
Auat~rus) Auat~rus)
TbuJopsi!I TbuJopsi!I
Tetraclinis Tetraclinis
'/7JuJa '/7JuJa Fokieni11 Fokieni11
Fjtzroy11 Fjtzroy11
Li Li
lliddrin1toni11 lliddrin1toni11
Libocedrus Libocedrus
Cal11traps1s Cal11traps1s
Callitris Callitris
Actinostrobus Actinostrobus
Maat, Maat,
DiselJM DiselJM
Biota Biota
Juniperua Juniperua
l'ha.recypar l'ha.recypar
Cupressus Cupressus
Arceutbos Arceutbos /llicrobiota /llicrobiota
*>SELEY *>SELEY
incompletely incompletely
F. F.
Endl. Endl.
Vent. Vent.
Ant. Ant.
L. L.
L, L,
Sieb. Sieb.
Rook. Rook.
loch loch
Beary Beary
25) 25)
Hook. Hook.
(Acta (Acta
Callitris Callitris
Calocedrua Calocedrua
Pl11tycl11dua Pl11tycl11dua
.Tuniperus .Tuniperus
L. L.
Don Don
(incl. (incl.
34: 34:
46:27) 46:27)
Tidakr. Tidakr.
(= (=
(= (=
(= (=
genera genera
(= (=
Pilgerodendron Pilgerodendron
Tetraclinaceae Tetraclinaceae
t: t:
htr11clini111 htr11clini111
f'hujopsia f'hujopsia
Fit.rroya Fit.rroya
Actinoatrobua Actinoatrobua
Octoclini• Octoclini•
t'buj11 t'buj11
Cup~au. Cup~au. Reyderia Reyderia liboc«lrua liboc«lrua
Callitria Callitria
liiddrin1tania liiddrin1tania
Foltiaia Foltiaia
Arb. Arb.
Arceutboa Arceutboa Juniperua Juniperua
Disel• Disel• Pilprodendron Pilprodendron
Papuacedrw Papuacedrw
Neocellitropaia Neocellitropaia
Ch~ypari• Ch~ypari•
Biota Biota
(1953) (1953)
Boutelje Boutelje
/llicrobiota /llicrobiota
1932 1932
1930 1930
status status
Callitro1deae Callitro1deae
Thujoideae Thujoideae
Cuprea•oideae Cuprea•oideae
Juniperoideae Juniperoideae
LI LI
& &
Arn. Arn.
(J. (J.
Florin Florin
•tatua: •tatua:
Li Li
Junipereae Junipereae
Cupreaaeae Cupreaaeae
Libocedreae Libocedreae
ThUJopaideae ThUJopaideae
Tetreclineae Tetreclineae
Herb. Herb.
Soc. Soc.
Actino•trobeae Actino•trobeae
1eneric 1eneric
GROUPINGS. GROUPINGS.
Linn, Linn,
17,29) 17,29)
Syst. Syst.
NeocallUropais) NeocallUropais)
,A,-trocedrua ,A,-trocedrua
Pwpu«:«/rua Pwpu«:«/rua
(J. (J.
= =
.. ..
( (
(Nat. (Nat.
180) 180)
(1926) (1926)
19154 19154
19153 19153
4: 4:
uncertain uncertain
Cupresaoideae Cupresaoideae
Callitroicleae Callitroicleae
Thuja Thuja
Thujopais Thujopais
liiddrin,toni11 liiddrin,toni11
Tetraclinis Tetraclinis L1bocedrua L1bocedrua Fokienia Fokienia
DiseJ DiseJ
Fitzroya Fitzroya
Juaiperw Juaiperw
Cellitro,wis Cellitro,wis
Cal]itria Cal]itria
CupruaU11 CupruaU11
C-"-ecyparia C-"-ecyparia
Actinoatrobua Actinoatrobua
Arceutboa Arceutboa
Compton Compton
Nicrobiota Nicrobiota
Komerov Komerov
Pll,CllR Pll,CllR
Petrop. Petrop.
Pil1er Pil1er
Pil1er Pil1er
part) part)
Pil1er Pil1er
in in
placed: placed:
Hort. Hort.
45:432) 45:432)
SUBFAMILY SUBFAMILY
Callitropsu Callitropsu
NicrobiotE, NicrobiotE,
not not
1923 1923
1922 1922
Pl•tycladua) Pl•tycladua)
Juniperus Juniperus
Tbujoicleae Tbujoicleae
Juniperoideae Juniperoideae
Cuprea•oide-
AND AND
(= (=
(-" (-"
1950) 1950)
J'etraclinia J'etraclinia
ffiuJOP*i• ffiuJOP*i•
ffiuja ffiuja
Actinaatrobua Actinaatrobua
Neocallitropeia Neocallitropeia
FitzroJ'll FitzroJ'll
Libocedru., Libocedru.,
lfiddrin1toni11 lfiddrin1toni11
Cllllitria Cllllitria
Diael
Nicrobiot11 Nicrobiot11 Biota Biota
Pil,erodendron Pil,erodendron
Fokienia Fokienia
CJJa..ecyparis CJJa..ecyparis
CUpressus CUpressus
( (
Arceuthoa Arceuthoa
Juniperu111 Juniperu111
Soc. Soc.
vi. vi.
Chron. Chron.
Bot. Bot.
353) 353)
196) 196)
Soc. Soc.
JANCREN JANCREN
179) 179)
Hort. Hort.
2: 2:
2: 2:
(Gard. (Gard.
* *
(Oestr. (Oestr.
Taa11. Taa11.
TRIBAL TRIBAL
Thujopaicleae Thujopaicleae
Cupreaaeae Cupreaaeae
Junipereae Junipereae
Roy. Roy.
Hort. Hort.
Bot. Bot.
Actinoetrobeae Actinoetrobeae
• •
* *
• •
Fl. Fl.
(J. (J.
( (
(J. (J.
1913a) 1913a)
Kot. Kot.
Platycladus) Platycladus)
(J, (J,
( (
Thoaas Thoaas
* *
OF OF
(Dendrol. (Dendrol.
& &
f. f.
f. f.
et et
(; (;
249) 249)
uncertain uncertain
Maat. Maat.
Kurz. Kurz.
rhujopai!I rhujopai!I
liiddrin,tonia liiddrin,tonia
~tr11clinia ~tr11clinia
Callitri!I Callitri!I
f'buja f'buja ActinOlltrobus ActinOlltrobus
Fitzroya Fitzroya
Cha.aecyparis Cha.aecyparis
Cupre11aus Cupre11aus
JuniperU11 JuniperU11
Biota Biota
Libocedl"UII Libocedl"UII
Ant. Ant.
SAXTON SAXTON
Henry Henry
Koch. Koch.
Hook. Hook.
Hook. Hook.
Cupre••o1deae Cupre••o1deae
Juniperoideae Juniperoideae
(Nov/Dec) (Nov/Dec)
14:250) 14:250) position position
(June/July) (June/July)
49:66) 49:66)
Wochenbl. Wochenbl.
264) 264)
Tetraclini!I Tetraclini!I
Fokienis Fokienis
•: •:
Jkyderia Jkyderia
Fitzro.YtJ Fitzro.YtJ
Clllocedrua Clllocedrua
Arceuthos Arceuthos
Diselat1 Diselat1
1911 1911
1892 1892
Callitroideae Callitroideae
1854 1854 1873 1873
1857 1857
1851 1851
CupreHoideae CupreHoideae
EXAMPLES EXAMPLES
part) part)
in in
2: 2:
2: 2:
ii) ii)
(Paleon. (Paleon.
Veg. Veg.
10) 10)
1. 1.
Vea:. Vea:.
part) part)
42) 42)
Jap. Jap.
Suppl. Suppl.
(1948) (1948)
294) 294)
(Pl. (Pl.
Nat. Nat.
Neoc11J]jtropai111) Neoc11J]jtropai111)
Jun1peru111 Jun1peru111
PlatycJIH/w) PlatycJIH/w)
in in
Nov. Nov.
Florin Florin
if. if.
Pl. Pl.
Nat. Nat.
(FI. (FI.
1002) 1002)
Pl. Pl.
46) 46)
(= (=
(= (=
( = = (
Con Con
Leha. Leha.
Gen. Gen.
(Hist. (Hist.
f'bujopais f'bujopais
J'etraclini• J'etraclini•
Pl. Pl.
lfiddrin,toaia lfiddrin,toaia
ffiuJ• ffiuJ•
Foltienia Foltienia
C11llitropsJ'• C11llitropsJ'•
Juniperua Juniperua
Actina.trobua Actina.trobua
Nicrobiota Nicrobiota
Di•el• Di•el•
Pillerodeadran Pillerodeadran Fitzro,,. Fitzro,,.
Libocetlrul Libocetlrul
Arceutboa Arceutboa
liot11 liot11 C•llitri• C•llitri•
C/J-«:ypari• C/J-«:ypari•
CUpreaa~ CUpreaa~
BUCHHOLZ BUCHHOLZ
(Gen. (Gen.
(Hiat. (Hiat.
(Gen. (Gen.
(1847) (1847)
Zucc. Zucc.
ex ex
Conif. Conif.
TABLE TABLE
(Syn. (Syn.
Callitri• Callitri•
Platyclatlin) Platyclatlin)
(Dec. (Dec.
(Sp. (Sp.
et et
(-
(• (•
644) 644)
Endl. Endl.
Spach. Spach.
Mit, Mit,
85B,•90) 85B,•90)
(Syn. (Syn.
Spach. Spach.
1'hujopais 1'hujopais
Thuja Thuja
Liboc«lrus Liboc«lrus
Frenela Frenela
Nfr/ddn1to11111 Nfr/ddn1to11111 l: l:
Callitri• Callitri•
Actina.trabus Actina.trabus
Cha.aecyparis Cha.aecyparis
Cupre••us Cupre••us
Biota Biota
Juniperua Juniperua
Endl. Endl.
Neoc.llitro,-ia Neoc.llitro,-ia
Linneaua Linneaua
Linneaus Linneaus
Vent. Vent.
Linneaua Linneaua
Sieb. Sieb.
ENDLICHBR ENDLICHBR
239) 239)
Endl. Endl.
Linneaus Linneaus
1944 1944
Juniperoideae Juniperoideae
C•llitroideu C•llitroideu
Thujoideae Thujoideae
Cupre••oideae Cupre••oideae
11: 11:
Preiss. Preiss.
333) 333)
25) 25)
inee inee
Thuj11 Thuj11
libocedrua libocedrua
ThujoJ)lli• ThujoJ)lli•
liiddrin1toni11 liiddrin1toni11
Cblaaecyparis Cblaaecyparis
.luniperu11 .luniperu11
Actina.trobus Actina.trobus
(."aJlitris (."aJlitris
Cuprea•us Cuprea•us
Cup~sua Cup~sua
Pl11tycladus Pl11tycladus
Biota Biota
1808 1808
1737 1737
1847 1847
1844 1844 1845 1845
1842 1842
1753 1753
Actinoatrobeae Actinoatrobeae
Thuiopeideae Thuiopeideae
Junip@r Junip@r Cuprea•ineae Cuprea•ineae 1. TAXONOMIC HISTORY.
There have been numerous taxonomic treatments of the members of
the family Cupressaceae Nager that have proposed many different
groupings at the tribe and subfamily level, although only one or two
characters, mostly of the mature ovulate cone, have been utilised. An
outline of the taxonomic history of subfamily and tribal groupings is
given below; membership of the various groupings is shown in Table 1.
Synonymy of suprageneric taxa prior to 1950 has been given by Janchen
(1950); synonymy of genera and species is to be found in Dallimore and
Jackson (1966) and Gaussen (1968). Authorities for all genera are
listed in Table 1. For other authorities not mentioned in the text or
the Table in which a binomial first appears, refer to the appendix.
Prior to the formalisation of the Cupressaceae Nager (1907),
members of the family were combined with those of the present
Taxodiaceae Neger in one family. For example, Endlicher's (1847) order
( = family) Cupressineae recognised 5 "suborders", 4 of which contained
the cupressaceous taxa and the fifth the taxodiaceous taxa. The
cupressaceous "suborders" were defined on characteristics of the
ovulate cone, particularly the form of the cone scales: Juniperinae
characterised by fleshy cones, Actinostrobeae by valvate cone scales,
Thujopsideae by imbricate cone scales, and Cupressineae Verae by
peltate cone scales. Eichler (1889), Neger (1907) and Vierhapper
( 1910) followed this arrangement of cupressaceous taxa with only minor
changes such as in the placement of Libocedrus (Table 1), simply
incorporating new genera as they were reported. Section I, p.2
Saxton (1913b,c, 1929), howevei-, used gametophyte and pi-o-embi-yo characters to divide the Cupressaceae into 2 subfamilies, so distinguishing the embi-yologically diffei-ent genera Widdringtonia,
Callitris and Actinostrobus from the remaining genera as the subfamilies Calliti-oideae and Cupi-essoideae i-espectively. His embryological survey included only 7 cupressaceous genera, the remaindei- being placed on an unexplained assessment of 'probable' or
'possible' character-states.
The detailed revision of the Gymnospermae by Pilger (1926) i-ecognised the separation of Cupressaceae from Taxodiaceae as proposed by Neger (1907). The Cupressaceae was divided into 3 subfamilies, once again on characters of the ovulate cone: Juniperoideae with fleshy cones, Cupressoideae with woody, shield-like cone scales, and
Thujoideae with cone scales that either separated or remained overlapping at maturity (" •• Schuppen klappig auseinanderweichend oder dachig deckend."). The genei-a compi-ising Saxton's subfamily
Callitroideae were submerged in the Thujoideae.
Moseley's (1943) revision of Pilger's treatment of the family re-introduced embryological characters to re-erect the Callitroideae as a foui-th subfamily; however the incompleteness of the data for these characters meant that again many genera were only tentatively placed.
The subsequent revision by Buchholz (1948) recognised these same four subfamilies, each defined (Buchholz 1946) by characters of the ovulate cone: Cupressoideae with valvate, shield-shaped woody scales;
Juniperoideae with fleshy cones; Callitroideae with valvate cone scales; and Thujoideae with "somewhat imbricated" cone scales.
Membership of the 4 subfamilies was substantially according to Moseley and Pilgei-, with only Tetraclinis moved from Thujoideae into Section 1, p.3
Callitroideae, and Microbiota moved from Juniperoideae to Thujoideae, both without comment.
Janchen (1950) also recognised 4 subgroups in the family, but treated them as tribes which he grouped into 2 subfamilies, the
Juniperoideae and Cupressoideae. The former contained a single tribe, characterised by fleshy cones; the latter included 3 tribes, separated once again primarily by characters of the ovulate cone scales:
Cupresseae with shield-like scales, which he considered anatomically distinct from those found in the remaining 2 tribes; Thujopsideae with imbricate cone scales; and Actinostrobeae with cone scales valvate or only a little imbricate.
Li (1953) proposed a reclassification of the family into 2 subfamilies, basing the distinction solely on cone scale displacement: imbricate in Cupressoideae, valvate in Callitroideae. In view of the statements of some of the previous authors cited above about the nature of cone scales in various genera, particularly of the Cupresseae, the basis for this division seems questionable; it was probably more than a little influenced by geographic considerations, since his Cupressoideae included all the essentially northern genera except for the North
African Tetraclinis, while the Callitroideae included Tetraclinis and all southern genera.
Each subfamily was further divided into 3 tribes, again defined on characters of the female cone. In the Callitroideae, the
Actinostrobeae contained genera with ternate scales, the Libocedreae those with paired or quadrate scales, and the Tetraclineae those with
"paired dissimilar" scales. In the Cupressoideae, the Cupresseae comprised genera with thick, shield-like scales, the Thujopsideae those with flat, more or less concave scales, and the Junipereae those with fleshy scales that coalesced at maturity. Section 1, p.4
Li's arrangement of the family has been generally followed by subsequent authors with only a few changes in tribal membership proposed. Gaussen (1968), in his revision of the family, moved
Neoca11itropsis from the Libocedreae to the Actinostrobeae on the basis of a proposed evolutionary link with Callitris macleayana as indicated by ovule and cone scale number. Krussman (1971) placed the northern
Calocedrus back into the Libocedreae in the southern subfamily without comment. Eckenwalder (1976a,b) maintained Li's subfamily distinction, although at the tribal level, when he proposed the re-unification of the Cupressaceae and Taxodiaceae.
Not everyone, however, has been convinced of the naturalness of
Li's arrangement. Even Boutelje (1955), who found support for some aspects of Li's treatment of the family in his study of wood anatomy, commented that the position of Fitzroya in the southern subfamily was uncertain. Florin (1963) expressed dissatisfaction with the placement of Tetraclinis in the southern subfamily. de Laubenfels (1965) suggested a closer relationship between Fitzroya and Diselma than indicated by Li's tribal groupings, and further commented that the separation of the subfamilies on the basis of valvate as distinct from imbricate cone scales seemed to be "nebulous"; he suggested that
Libocedrus would be better placed within the northern tribe Thujoideae.
Most recently, a study of leaf cuticle characters led Oladele (1983a,b) to conclude that, while there is substantial variation between genera, there is no clear separation between northern and southern genera.
Indeed, the interpretation of the valvate or imbricate nature of the cone scales in the family by various authors varies markedly: Buchholz
(1948) considered the scales to be valvate in Cupressus and
Chamaecyparis and imbricate in Diselma, Pilgerodendron and Fitzroya, whilst Li (1953) considered them to be imbricate in the first two and Section 1, p.5 valvate in the last three. Janchen (1950) described the cone scales as imbricate in both Pilgerodendron and the Libocedrus s.1. complex, while
Li considered only the species split from the latter complex as
Calocedrus ( ut Heyderia) to have imbricate cone scales, both
Pilgerodendron and the remaining taxa of Libocedrus having valvate scales.
Other traditional sources of taxonomic characters do not appear to be very useful in assessing the present groupings. de Laubenfels
(1953) considered the leaves of conifers in general to be more uniform in their external morphology than the reproductive structures;
Fitzpartick (1965) could not distinguish Oupressus (in part),
Ohamaecyparis, Thuja, Platycladus, Microbiota, Thujopsis, Oalocedrus and Fokienia by foliage morphology alone. It seems that the marked reduction of the leaves in this family has led to great uniformity.
Harrison (1966: 623) commented on anomalies in the taxonomy of this family, and called for a critical reappraisal of the whole family:
"The family Cupressaceae is beset with anomalies, some of its genera having been studied more fully than others. Several taxonomic changes have been proposed by various authors, but the whole family needs critical reappraisal to co-ordinate our increasing knowledge of its component taxa." His comments are still fully applicable to the taxonomy of the Cupressaceae s.s.
In view of the evidence cited above that the present subfamily and tribal groups do not appear to adequately reflect affinities between taxa, it is appropriate to broaden the data base for the family before attempting this reappraisal.
In recent years, flavonoid content of leaves, stems, fruit and wood has yielded useful taxonomic data in a wide range of plant taxa
(Hegnauer 1969; Harborne 1973). Biflavonoids are a distinctive group Section 1, p.6 of flavonoid pigments that are a characteristic component of the leaves of most gymnosperms (Geiger and Quinn 1975, 1982). Although few members of the Cupressaceae have been studied for leaf biflavonoids, there is already evidence of some chemical discontinuities in the family (Geiger and Quinn 1982:527). The aim of this thesis is to broaden the data base of the family by analysing the biflavonoid content of the leaves in a wide range of members, and to apply this, along with other existing data, to a reassessment of affinities within the Cupressaceae s.s. FIGURE 1. EXAMPLES OF THE ThlTERFLAVONYL LINKAGES IN DIFFERENT SERIES OF BIFLAVONOIDS.
0
Cupressuftavone
Amentoftavone
,~o,~mr R40,Jyll _ Agathisftavone 1 RO,W"' 1Y O A O OR 0 :,. I I
OR1 0
Hinokiftavone
HO Taiwaniaftavone
OH 0
Robustaftavone
Each series is illustrated by the parental biflavone (R=H) • The numbering system of individual carbon atoms is illustrated by the flavonoid nucleus in the top left. Double and triple primes identify the atoms of the second ring in the biflavonoid structure. }~GORE 2. BIOSYNTHE.'11IC PATHWAYS OF FLAVONOIDS.
Phenylalanine + t Acetyl-CoA ! 4-Coumaroyl -CoA Malonyl - CoA
OH
HO~OH_~
wOH 0 Chalcone
~t OH
HO
OH 0 Flavanone 1 / /j \ ', ,,,,-1/ / \ ...... OH / / \ '-...
HO / / \ HO'CfO ef/ I :::,.., l{o_~ OH O :::,... I OH F~aHvo~e / / / \ lsoflovone H Apigenin/ OH OH ~ OH / I HO HO HO o :::... I
OH 0 OH OH 0 Flovonol Anthocyonidin Luteolin
3! OCH3 OH
HO l HO OH
OH 0 Chrysoeriol OH 0
Amentoflavonc
(adapted from Ebel and Hahlbrock 1982)
The enzymes identified. in the biosynthesis of some flavones are: 1, Flavonoid oxidase; 2, Flavonoid 3'-hydroxylase; 3, SAM: flavonoid 31 -0-4llethyltransferase. TABLE 2. OCCURRENCE OF BIFLAVONOIDS IN THE CONIFERALES.
Podocarpaceae Ara.ucariaceae Ci:pressaceae Taxod.iaceae Cephalotaxaceae
::c,. s;~ecies investigated 20 1 24 11 meth,vlation E::iRe ccr.:round pattern :.i;r.er.to:'lavone 13 4 18 4
1 6 1 3 4' 6 2 E 1" 8 3 2 4'" 8 2 2 14' 9 11" 2 14"' 4 7a 1"4' 2 4'4"' 8 1"4"' 5 2 14'1" 74'4"' ( 10b) 2 1 (3b) 5 4'7"4"' 3 2 3
74'7"4"' 4 2,3 di hydro 4'7" 2,3 di hydro 74'4"' hi r:o>::i flavone 11 3 17 10 1 7" 5 (2°) 5 4'" 6 4 11" .," ( 1a) 7"4"' 3 ci;pressuflavone 3 7 13
1 3 6 4' 2
11" 3 5 74'7" 3 4 74'7"4"' 3 robustaflavone tai wani aflavone 1" 4'7" agRthisflavone 5 7 6 77" 4 14'" 3 11"4"'
a, unidentified dimethyl ether reported; b, incompletely identified trimethyl ether; c, incompletely identified monomethyl ether, Section 2, p. 7
2. BIFLAVONOIDS.
2.1 Introduction.
Biflavonoids are dimeric flavonoids which, in the gymnosperms, are
mainly based on apigenin (Geiger and Quinn 1982); a range of known
skeletons resulting from different interflavonyl linkages is
illustrated in Fig 1. Although very little research has been carried
out on their biosynthesis, they are thought to be synthesized according
to the general scheme for flavonoid formation (Swain 1975; Giannasi
1978; Ebel and Hahlbrock 1982; Fig. 2), possibly by oxidative coupling
of 2 chalkone or flavanone units with subsequent modification of the
central C3 units (Geiger and Quinn 1975). All classes of flavonoids
are biosynthetically closely related, with a chalcone being the common
intermediate. Methylation of hydroxyl positions is assumed to occur
after dimerization and subsequent to all other modifications of the
flavonoid ring structure (Ebel and Hahlbrock 1982).
2.2 Distribution and use in taxonomy.
Biflavonoid aglycones are a characteristic component of the leaves
of most gymnosperms, with the exception of the Gnetales and the
Pinaceae, although a biflavonoid has recently been reported from Abies
webbiania Lindl. (Chatterjee et al. 1984), and a spiro-biflavonoid from
Larbr gmelini (Rupr) Rupr (Zhaobang et al. 1985). They are also known
in at least 11 families of angiosperms. There are no reports of their
cellular localization within leaf tissue. Their chemistry and
distribution has been extensively reviewed by Geiger and Quinn ( 1975,
1982). Table 2 lists the occurrence of biflavonoids reported from the
families comprising the Coniferales. TABLE 3. DISTRIBUTION OF BIFLAVONOIDS IN THE CUPRESSACEAE.
Am mAm dAm tAm Cu mCu dCu Hi mHi dHi Ro Ref Cupressoideae Junipereae Juniperus chinensis L. + + + 1 'pfitizeriana' + +a + + + + 2 'plumosa aurea' + +a + + + 2 J. ca..unis L. + +a + 2 J. horizontalis Moench. + +a + +b + 2,3 J. JBBcropoda Bo iss + + + +c + 4,5 J. occidentalis Hook. + +a + + 2 J. oxycedrus L. + + + 6,7 J. phoenica L. + +a, d+ + + +c 2,5 J. procUBlbens S ieb. et Zucc. + +a + + + + 2 J. recurva Buch. -Ham. + + + + + +e 3 J. sabina L. + +a 2 J. squB111BtB Buch.-Ham. + +a 2 J. virginiana + +a 2 Cupresseae A Cupressus arizonica Greene + 8 c. funebris Endl. + + + +c 9 c. goveniana Gord. + + +f + +c 8,9,10 c. lusitanica Mill. + + + 9 var. bentha.i Carr + +d + +f + + 11 c. se11pervirens L. + + 9, 12 'stricta' + + 9 c. torulosa D. Don + + + 9, 12 ChBIIIBecyparis obtusa (Sieb. et Zucc.) Endl. +c 9,12 'breviranei' + +h +i 8 c. pisifera (Sieb. et Zucc.) Endl. + +h +i 8 'squarrosa' + +h +i 8 Thujopsideae Thuja plicata D. Don + + 13 T. standishii Carr. + +h +i 8 Thujopsis dolobrata Sieb. et Zucc. + +h +i 8 Platycladus orientalis (Spach) Franco +c +h +i 8,9
Callitroideae Actinostrobeae 8 Callitris colu.ellaris F. Muell. + +II ? ? 14 c. rhOlllboideae R. Br + 15
A, ut Cupressus glauca;. B, ut Callitris glauca R. Br. Am, amentoflavone; mAm, monomethyl amentoflavone; dAm, dimethyl amentoflavone; tAII, trimethyl amentoflavone; Cu, cupressuflavone; mCu, monomethyl cupressuflavone; dCu, dimethyl cupressuflavone; Hi, hinokiflavone; mHi, monomethyl hinokiflavone; dHi, dimethyl hinokiflavone; Ro, robustaflavone. a, 4' monomethyl amentoflavone; b, 74'4"' trimethyl amentoflavone; c, 7" monomethyl hinokiflavone; d, 4'" monomethyl amentoflavone; e, 77" dimethyl cupressuflavone; f, 4' monomethyl cupressuflavone; g, 7 monomethyl amentoflavone; h, 4"' monomethyl hinokiflavone; i, 77" dimethyl hinokiflavone. References: 1, Pelter et al. (1971); 2, Lamer-Zarawaska (1975); 3, Hameed et al. (1973); 4, Ilyas et al. (1977a); 5, Fatma et al. (1979); 6, Pascual Teresa et al. (1980); 7, Lebreton et al. (1978); 8, Miura and Kawano (1968); 9, Natarajan et al. (1970); 10, Taufeeq et al. (1979); 11, Taufeeq et al. (1978); 12, Murti et al. (1967); 13, Rahman et al. (1972); 14, Ansari et al. (1981); 15, Siva Prasad and Krishnamurty (1977). Section 2.2, p.8
Although reports of biflavonoid occurrence have tended to be
sporadic and/or limited to a few members of any taxon, there are
several studies which have shown the value of biflavonoids as chemical
markers in taxonomic revisions. For example, the three families
comprising the order Cycadales have been shown to be clearly
differentiated on their leaf biflavonoid profiles (Dossaji et al.
1975a), members of the Cycadaceae possessing both amentoflavone and
hinokiflavone derivatives, the Zamiaceae having amentoflavone
derivatives only, and the monotypic Stangeriaceae being devoid of
biflavonoids. The authors also found a high degree of uniformity
between species of the same genus, differences in the biflavonoid
pattern supporting the existing generic boundaries. But the occurrence
of different biflavonoids in the testa of Macrozamia (Gadek 1982) to
those reported in the leaves, and studies of biflavonoids in bark,
wood, fruits, roots, stamens and leaves of other vascular plants
(Geiger and Quinn 1975, 1982; Wannan et al. 1985) indicate that these
compounds may be specific to particular tissues, and reinforce the need
to use strictly homologous tissues in chemotaxonomic comparisions of
different taxa.
On the other hand, discontinuities in leaf biflavonoids within
Dacrydium s.1. (Podocarpaceae; Quinn and Gadek 1981) were found to
correlate strongly with morphological and anatomical discontinuities,
and assisted in the recognition of three new segregate genera (Quinn
1982).
2.3 Distribution in the Cupressaceae.
The known distribution of leaf biflavonoids amongst members of the
Cupressaceae is shown in Table 3. Only 24 species, representing 7
genera (of a possible 20) have been reliably examined. Indeed, some Section 2.3, p.9 data in the literature cannot be utilised or relied upon for comparative studies due to a number of inadequacies. Firstly, many early publications reported only the major constituents or those most easily isolated and characterised (Geiger and Quinn 1975). Secondly, misidentifications were not uncommon before advanced methods became available. A number of authors have reported on the lack of reliability of certain early procedures of structural determination,
-and in particular the identification of plant products by such physical property techniques alone as gas liquid chromatographic retention times, Rf values from paper and thin layer chromatography (TLC), colour reactions, or spot tests (Harborne 1975; Seigler 1981). Natarajan et al. (1970) found that biflavonoids purified by earlier methods of crystallization or derivitization were not single entities, and reinvestigations of these fractions by TLC revealed the presence of previously undetected biflavonoids. More recent refinements of the TLC technique have allowed even greater resolution than was obtained in the work of Natarajan et al. ( 1970). In Cupressus goveniana Gord., for example, Miura and Kawano (1968) reported only cupressuflavone, while
Natarajan et al. (1970) reported amentoflavone and hinokiflavone in addition to cupressuflavone; recently Taufeeq et al. (1979) isolated the 4' methyl ether of cupressuflavone and the 7" methyl ether of hinokiflavone in addition to the previously reported compounds.
Hence, data from many studies undertaken prior to Natarajan et al.
(1970) have not been incorporated into Table 3, while even those of
Natarajan and contemporaries are possibly incomplete.
Thirdly, even some recent reports fail to indicate clearly either the presence or identity of biflavonoids occuring in minor concentrations. Where the occurrence is noted as monomethyl or dimethyl ethers of a particular parental structure, this has been Section 2.3, p.10 incorporated into Table 3. However, some reports simply indicate the presence of biflavonoids by Rf values alone. For example,
Lamer-Zarawaska ( 1975) records the presence of biflavonoids in a range of Juniperus species by the occurrence of bands of differing Rf's in a number of solvents. While the author gives an indication of the possible structures present, the data are of little taxonomic use.
Fourthly, and of greatest concern to the taxonomist, is the lack of adequate documentation of plant sources by citing authorities, localities and voucher specimens. None of the reports in the literature cite voucher specimens and many omit authorities. Only 17 reports cite collecting localities, and only 3 of these appear to be from natural populations. Several dubious and irregular reports have been omitted from Table 3 on this basis.
Despite the limitations inherent in the data in Table 3, the distribution of biflavonoids in the family shows some significant discontinuities, both in the number of biflavonoid skeletons and in the methylation patterns.
6 of the 7 genera for which reports are available are members of
Li's subfamily Cupressoideae, being drawn from all three tribes.
Hinokiflavone is recorded from all 6 genera, although it has not been detected in all species of Juniperus or Cupressus. The patterns of leaf biflavonoids in Cupressus and Juniperus appear very similar, and contrast markedly with the reported occurrence of hinokiflavone derivatives alone in Chamaecyparis. This calls into question the present tribal groupings which place Cupressus and Chamaecyparis together in the Cupresseae, while separating Juniperus into a tribe of its own, the Juniperae.
The only recent records of. biflavonoid distribution in the subfamily Callitroideae are from two species of Callitris: Section 2.3, p.11 amentoflavone alone in C. rhomboidea R. Br. (Siva Prassad et al. 1977), and amentoflavone and its 7 methyl ether in C. glauca R.Br ( = C. columellaris F. Muell.; Ansari et al. 1981). The latter report also indicated that hinokiflavone had been detected in trace amounts by TLC.
Clearly, the family has not been adequately surveyed for any assessment of affinities to be made on the basis of leaf biflavonoids.
An extensive survey of representatives of all genera was therefore undertaken in order to produce a reliable set of data on leaf biflavonoids within the family. Section 3, p.12
3. TECHNIQUES.
3.1 Plant Material.
Extractions were made on homologous samples of sterile branchlets
and leaves obtained from both fresh and herbarium material. Samples
from herbarium collections were small, approximately 0.5 to 5 g. dry
weight, whilst samples of fresh material ranged from 10 to 150 g. dry
weight. Voucher specimens are listed in the appendix.
3.2 Extraction and Isolation.
3.2.1 Thin Layer Chromatography.
Samples were lightly crushed and extracted in 70% ethanol for
48 hours, filtered and the extract dried overnight in an oven at
30 - 40 °C. Extracts from fresh material and from large amounts
of dried material were washed in petroleum ether (bp 60-80 °C) to
remove excess oils before drying, The dried residue was taken up
in a small volume of 70% ethanol and separated into its
constituent biflavonoids by sequential one-dimensional
chromatography.
(i) Extracts from large amounts of dried material were subjected to
descending paper chromatography in n-butanol:acetic acid:water
(60:15:25) (BAW). The biflavonoids, which appeared under UV as
a single absorbing band immediately behind the solvent front,
were eluted with 5% acetic acid in 95% ethanol.
(ii) Chromatography of the eluant on semi-preparative plates of
silica-gel 60 in toluene-ethyl formate-formic acid (5:4:1) (TEF)
yeilded biflavonoid bands which appeared dark absorbing under
long wave UV; these were scraped from the plate and extracted in TABLE 4. CHROMATOGRAPHIC CHARACTERiffi'ICS AND SOURCES OF REFERENCE SAMPLES OF PARENTAL COMPOUNDS AND THEIR PARTIAL METHYL NI'BERS.
Rf in Colour of 0-!"'.?et~~flati on 3ase BPF fluorescence ::iattern compound Principle source(s) 1()0: 20:7 in AlC13 s.o yellow 77"4'4"' amentoflavone standard - ~arkham
c.76 yellow 7"4'4"' amentoflavone Araucaria cunninghamii 0.76 yellow 74'4"' amentoflavone standard - Geiger 0.75 ora!'lge 77"4' cupressuflavone Araucaria cunningharnii o. 7 5 dark yello11 7"4"' hi noki flavone standard - Geiger
1 • .., 0.64 ,yellow 74' amentoflavone standard - !larborne; D~'.S-2; Halocarpus kirkii 3.5c 0.62 yellow 7"4"' amentoflavone standard - Geiger; Lagarostrobos colensoi 3. ",3 c,.62 dark yello" 411' hinoki flavone standard - Geiger C·.62 brieht yellow apigenin standard Sigma 0.60 orange 77" cu ;,re ssuflavone Araucaria cunninghamii, !• bidwillii 0.59 yellow 4'11"' amentoflavone standard - Ha.rliorne; D!,'.S-3; Halocarpus bidwillii. o. 55 dark yellow 7" hinokiflavone standard - Geiger 0.45 orange 74'" ag-c1thisflavone Araucaria cunninghamii 0.45 orange 77" agathisflavone Araucaria bidwillii
C ,4]:J 0.35 yello,1 4' amentoflavone standard - Geiger, Harborne; Dlf.S-4 2 .4_;,_ 0.35 yellow 7" arnentoflavone standard - Geiger; DM-1 2.)C O. 33 yellow 4'" amentoflavone standard - Geiger; DMS-5; ni:-2 2.33 0.33 yellow 7 tai waniaflavone Taiwania crY1Jtomerioides 2.33 0.33 orange 7 cupressuflavone Araucaria bidwillii ~.jrt o._n dark yellow hi noki fl avone standard - Geiger 2.2 0.29 yellow 7 arnentoflavone Lagarostrobos colensoi, Halocarpus kirkii 2.1 0.27 orange 7 agathisflavone Araucaria bidwillii, !• cunninghamii
o. 18 2,3 dibydro amentoflavone standard - Geiger 1.4 0.17 yellow amentoflavone standard - Geiger 1.4 0.17 yellow tai waniaflavone Taiwania cryptomerioides 1.3 0.16 yellow robustaflavone Agathis robusta 1.2 o. 15 orange cupressuflavone Cuoressus sempervirens, Araucaria bidwillii o. 12 orange agathisflavone Araucaria bidwillii Section 3.2.1, p,13
70% ethanol.
(iii) These bands were further refined on pre-coated
aluminium-backed silica gel plates (layer thickness 0.2mm)
developed in benzene-pyridine-formic acid (BPF). The
proportions were varied according to the Rf of the bands in the
previous system. 100:10:5 gave a good separation of the faster
moving constituents, while 100:20:7 and 100:30:10 were employed
for the medium and slow constituents.
Extracts from small volumes of material were chromatographed
directly on semi-preparative or pre-coated silica-gel plates
developed in BPF ( 100:20:7) and refined as above in suitable
variants of the BPF system.
(iv) A final purification was sometimes necessary to separate
structural isomers (Geiger and Quinn 1975). This was performed
on pre-coated cellulose plates developed in freshly prepared
butanol-2N ammonium hydroxide (1:1, upper layer; BN).
Evaluation.
TLC on silica gel in BPF solvent separates the biflavonoids
into major bands, each containing those with the same number of
free hydroxyl groups. Thus hinokiflavone moves to the same level
as amentoflavone monomethyl ethers, since all have 5 free hydroxyl
groups (Fig. 1).
Separation of compounds within these major bands by TLC on
silica gel has not previously been reported (see Chexal et al.
1970; Khan et al. 1971; Kamil et al. 1977, 1981). In this study,
re-chromatography in variants of BPF often allowed the separation
of the components of the band. This is indicated in Table 4 by the
second numeral: 2-1 refers to the slowest moving component in the
second slowest band from the initial separation in BPF. It Section 3.2.1, p.14
appears that the small differences in Rf of methyl ethers within
each major band reflect the degree of structural hinderance of the
free hydroxyl groups, and resultant differences in adsorption on
silica gel.
Samples of all bands were chromatographed in the BN system in
order to check their purity. In most instances complete
separation could be obtained with BPF; where this was not so, the
entire fraction was chromatographed in BN. Low, medium and high
Rf's in this system are indicated by the letters A, B and C
respectively. The bands on cellulose were diffuse and some
components did not move from the origin. Most major components,
however, could be adequately separated by these methods (see Table
4). A final check of purity was made by permethylation (see
3.3.2.2).
The separation of some derivatives, notably taiwaniaflavone
and its methyl ethers from amentoflavone and its methyl ethers,
was not possible with these techniques. When mixtures of these
compounds were encountered an initial separation was achieved by
TLC on silica gel in BN and the resulting bands were refined by
re-chromatography on silica gel in variants of BPF as above.
3.2.2 High Performance Liquid Chromatography (HPLC).
The first published application of HPLC to flavonoid analysis
(Ward and Pelter 1974) contained a demonstration of the separation
of a mixture of three permethyl biflavonoids. Although techniques
were quickly established for analysis of flavonoid mixtures (Wulf
and Nagel 1976; Galensa and Herrmann 1980; Bankova et al. 1982;
Casteele et al. 1982; Daigle and Conkerton 1982), there has been
only one other publication on the application of HPLC to Section 3.2.2, p.15
biflavonoid analysis: viz., the report by Briancon-Scheid et al.
(1982) on the separation of mixtures of partial methyl ethers of
amentoflavone.
Since the application of HPLC to both permethyl and partial
methyl ethers of biflavonoids had been little studied, sample
preparation, solvent systems and columns used in flavonoid
analysis and by Briancon-Scheid et al. (1982) were tried.
(i) Instrumentation. The separation of biflavones was carried out
with a Waters liquid chromatograph with a programmable gradient.
Peaks were detected at 320 nm with a variable wavelength
spectrophotometric detector.
(ii) The column tested was a Hibar Lichrosorb Diol column (10 um)
equipped with a silica pre-column.
(iii) Solvents used were chloroform and tetrahydrofuran (LO grade),
degaussed through an organic filter before use.
(iv) Standards and sample preparation. A range of reference
compounds of parentals and their methyl ethers were available
(see 3.3.1). These were dissolved in tetrahydrofuran and
ultracentrifuged before injection to reduce the risk of
particulate matter being injected.
(v) Chromatographic conditions. The best separation of parental
biflavones and their partial methyl ethers was obtained using
the Lichrosorb Diol column with a 9:1 mixture of chloroform and
tetrahydrofuran at an elution rate of }ml/minute.
(vi) Identification of peaks. Peaks were collected and
chromatographed against reference compounds on silica-gel in
BPF.
Evaluation. TABLE 5. HPLC OF VARIOUS METHYL ETHERS OF AMENTOFLAVONE AND HINOKIFLAVONE USING A DIOL COLUMN.
Compound Peak No. Retention (minutes) Amentoflavone 74'4"-trimethyl ether I 3:30 Hinokiflavone 7"4"'-dimethyl ether 1 3:30 Hinokiflavone 4"7-monomethyl ether 2 5 Hinokiflavone 7"-monomethyl ether 3 6:20* Amentoflavone 7"4"'-dimethyl ether 4 7* Hinokiflavone 5 13 Amentoflavone 6 20 *, no baseline separation between peaks 3 and 4. Conditions: 9:1 ratio of chloroform:tetrahydrofuran, eluting at 1 ml/minute. Peaks were detected at 320 run. Section 3.2.2, p.16
In contrast to the study of Briancon-Scheid et al. (1982),
which only examined the separation of methyl ethers of
amentoflavone, good separation of some partial methyl ethers of
amentoflavone and hinokiflavone was obtained in this study (Table
5). However, the initial work on mixtures of dimethyl ethers of
amentoflavone and cupressuflavone indicated that this column was
not able to separate these compounds under the conditions tested.
The problem may be overcome by further experiment with changing
solvents or solvent ratios.
3.3 Identification.
The purity of refined bands was checked on silica gel and
cellulose developed in BPF and BN respectively. Initial identification
of compounds was made by chromatographic comparison with reference
compounds using Rf and colour of fluorescence under UV after spraying
with a 5% ethanolic solution of aluminium chloride.
3.3.1 Reference Compounds.
Reference samples of 4'-monomethyl amentoflavone,
74'-dimethyl amentoflavone and 4'4"'-dimethyl amentoflavone were
supplied by Professor J. B. Harborne, University of Reading;
amentoflavone and its 4'-, 4"'-, 7"-, 7"4"'- and 74'4"'- methyl
ethers, and hinokiflavone and its 7"- , 4- , and 7"4- methyl
ethers were supplied by Dr H. Geiger, Hohenheim University;
amentoflavone 74'7"4"' tetramethyl ether was provided by Dr K. R.
Markham, Chemistry Division, DSIR, New Zealand.
Further reference compounds were isolated from species in
which their occurrence had been reported (see Table 2). Leaves of
Cupressus sempervirens L. were reported by Natarajan et al. (1970) Section 3.3.1, p.17 to contain only amentoflavone and cupressuflavone. TLC comparison of an extract of this species with the reference compound of amentoflavone led to the initial identification of a compound as cupressuflavone. The same compound was also isolated from
Oupressus macrocarpa Hartweg. and Araucaria bidwillii Hooker, both of which have been reported to contain cupressuflavone (Khan et al. 1971; Quinn and Gadek 1981).
Comparison of a leaf extract of Agathis robusta (C. Moore ex
F. Muell.) F. M. Bail., which contains robustaflavone (Varshney et al. 1973), with that of Araucaria bidwillii, which does not contain this biflavone (Khan et al. 1971), led to the isolation of robustafiavone; comparison of a leaf extract of Taiwania cryptomerioides Hayata (Kamil et al. 1977, 1981), which contains amentoflavone, hinokiflavone and taiwaniaflavone based derivatives, with amentoflavone and hinokiflavone reference compounds led to the isolation of taiwaniaflavone and its monomethyl ether.
Various methyl ethers of amentoflavone, cupressuflavone and agathisflavone have been reported from Araucaria bidwillii and A. cunninghamii D. Don (Khan et al. 1971, Ilyas et al. 1977b, 1978), and this enabled the. isolation of cupressuflavone and its 7- ,
77"- methyl ethers and an unknown trimethyl ether, as well as a number of methyl ethers of amentoflavone and agathisflavone.
There was some discrepancy in the identification of some methyl ethers from A. cunninghamii between two reports by Ilyas et al.
(1977b vs 1978). The dimethyl ether of amentoflavone was identified as 4'7" in 1977, but as 4'7 in 1978; similarly the trimethyl ether of amentoflavone was identified as sciadopity~in
(74'4"'-0-methyl) in 1977, but as kayaflavone Section 3.3.1, p.18
(7"4'4"'-0-methyl) in 1978. These particular records have not
been utilised.
A full list of reference compounds used in this study, their
source and TLC characteristics, is given in Table 4.
3.3.2 Determination of interflavonyl linkage.
3.3.2.1 Reaction with Aluminium Chloride (AICb).
An initial indication of the biflavonoid series to which a
component belonged was made on the TLC plate by spraying with a
5% ethanolic solution of AIOia. AlCb forms acid-stable
complexes with 5-hydroxy flavones (Mabry et al. 1970; Markham
1982) which fluoresce when viewed under UV light (366nm)
(Markham 1982). This was initially used to heighten the
sensitivity of biflavonoid band detection. Empirical
observations later revealed that the colour of fluosescence
could also be used as an indicator of the position of the
interflavonyl link. Derivatives of amentoflavone,
robustaflavone and taiwaniaflavone fluoresce yellow on addition
of AlCh under UV, hinokiflavone derivatives dark yellow, while
cupressuflavone and agathisflavone derivatives fluoresce orange
(see Table 4; see also Geiger and Quinn 1976, 1982; Quinn and
Gadek 1981).
3.3.2.2 Permethylation.
Permethylations were carried out using dimethyl sulphate in
dry boiling acetone over fused potassium carbonate and refluxing
for at least 7 hours (Khan et al. 1971; Markham 1982). The
eluant was filtered and dried overnight in an oven at 30-40°0.
The dried residue was re-extracted in ethanol and Section 3.3.2.2, p.19 chromatographed on pre-coated silica-gel plates developed in either BPF (100:20:7) or benzene-pyridine-ethyl formate-dioxan
(5:1:2:2) (BPEFD).
Permethylations were carried out at three stages.
Initially, a part of the original ethanolic extract of a collection (raw extract) was permethylated. Since all derivatives of any particular skeleton were converted to a single permethyl ether by this process, and since these permethyl ethers are fluorescent and can be readily distinguished chromatographically (Table 6), this is a rapid and sensitive technique for determining the combination of parental series present in the extract. This often allowed detection of
minor amounts of a biflavonoid series (i.e., various derivatives of a single skeleton), the individual components of which could
not be isolated by TLC of the original extract. Secondly, a
part of each band extracted from silica-gel TLC plates was
permethylated, again in order to determine the biflavonoid
series to which the component(s) of the band belonged. Often
these bands contained derivatives of more than one biflavonoid
series. Thirdly, a part of all major bands that were refined on
cellulose and chromatographically identified against reference
compounds were permethylated to confirm the parental biflavonoid
skeleton (see 2.1, Fig. 1).
Initial identification of each permethyl ether was made by
chromatographic comparison with previously prepared reference
compounds on silica-gel plates run in BPF and BPEFD and by the
colour of the autofluorescence of each compound under long wave
UV light (Table 6). TABLE
?.,3
6.
hi
cupressuflavone
robustaflavone agathi taiwaniaflavone
dihydro amentoflavone
apie;enin
Base
unknown
noki
compound
flavone
sflavone
amentoflavone
1
*
colour
Chromatographic
of
0.65
0.48 0.50 0.53 0.42 0.44 0.46
0.37 0.41
Rf
fluorescence
fluorescence
lie-,ht
BPF
lie-,ht
light
light
light
yellow
yellow
yellow
and
orange
yellow
blue spectral blue blue blue
30
minutes
*
data
o. 0.85
0.54
o.68 0.45
o. 0.48
o. 0.40
Rf
75
after
51
58
for
BPEFD
fluorescence
lie;ht light
permethylated
removing
light
blue blue
white
blue
blue white
yellow
yellow
blue
plate
*
compounds.
from
tN
266, 263,
266,
268, 265, 264,
absorption
solvent.
~max
322 328 324 324
323 324
emission
'Xmax:
470 437 427
460 Section 3.3.2.2, p.20
3.3.2.3 Reference permethyl ethers.
Reference samples of amentoflavone and hinokiflavone (Table
4; 3.3.1) were used to produce amentoflavone hexamethyl ether
and hinokiflavone pentamethyl ether respectively.
Cupressuflavone hexamethyl ether was obtained by permethylation
of cupressuflavone extracted from Cupressus macrocarpa,
Araucaria araucana and A. bidwillii (3.3.1). A fourth permethyl
ether was obtained from A. bidwillii and an identical product
was obtained from the permethylated leaf extracts of both A.
cunninghamii and Agathis robusta. The fact that agathisflavone
or its partial methyl ethers have been reported from the leaf
extracts of all three species (Varshney et al. 1973; Ilyas et
al. 1977b), and that no other group of biflavonoids is known to
be common to all three, lead to the identification of this
product as agathisflavone hexamethyl ether. Robustaflavone
hexamethyl ether was isolated as a fifth product from the
permethylated leaf extract of Agathis robusta. No comparable
fraction could be detected in the permethylated leaf extracts of
either of the Araucaria species; this accords with the reported
distribution of this biflavonoid series (Geiger and Quinn 1982).
Permethylation of the leaf extract of Taiwania cryptomerioides
yielded three products, two of which proved to be amentoflavone
hexamethyl ether and hinokiflavone pentamethyl ether. The third
product was distinct from all the above permethyl ethers, and
was determined to be taiwaniaflavone hexamethyl ether by
reference to the previous analysis of leaf biflavonoids in this
species ( Kamil et al. 1977, 1981). Section 3.3.2.4, p.21
3.3.2.4 High Performance Liquid Chromatography (HPLC).
The method has been described in Section 3.2.2.
(i) Waters Radial Pak 018 Reverse Phase and Waters Radial Pak
Silica 5 um and 10 um columns equipped with a silica
pre-column were tested.
(ii) Solvents used were chloroform and tetrahydrofuran (LC
grade), isopropanol (AR grade), methanol and deionized water
(redistilled). All were degaussed and filtered through the
appropriate organic or aqueous filters before use.
(iii) Samples of reference permethyl ethers obtained by TLC and
eluted in methanol were re-extracted into diethyl ether (2-3
times) to remove contamination by silica-gel. The relatively
non-polar permethyl ethers moved to the upper phase of the
methanol/diethyl ether mixture. This upper phase was dried,
taken up in redistilled methanol and ultra-centrifuged before
injection into the column.
(iv) Samples of permethylated raw extracts were dried, taken up
in chloroform and loaded onto a sep-pak silica cartridge.
This was eluted successively with n-heptane, ethyl acetate,
tetrahydrofuran, and a 9:1 mixture of tetrahydrofuran and
isopropanol. The last two fractions eluted the permethyl
ethers and were combined, dried, taken up in tetrahydrofuran
and ultra-centrifuged before injection.
(v) Peaks identified at 320 nm were collected and
chromatographed against reference compounds on silica-gel in
BPF and BPEFD.
Evaluation.
Separation of permethyl ethers by the 018 reverse phase
column was attempted using a mixture of methanol and water, FIGURE 3. HPLC TRACE OF THE PERI·';El'HYLATED RAW EXTRACT OF CALOCEDRUS DECURRENS.
4
3 5
inj.
0 2 4 6 8 10 12 14 mins
A 5um Silica column was used, with 10(1)6 THF. For further information see text. See Table 7 for peak identification. TABLE7. HPLC OF VARIOUSPERMETHYL ETHER SERIES USING REVERSEPHASE (C1s) AND NORMALPHASE (SILICA) COLUMNS.
Permethyl ether C1s1 10 um Si2 5 um Si3 Series Peak no. Retention Peak no. Retention Peak no. Retention Cupressuflavone 1 11:20 3 11 3 7:20 Amentoflavone 2 17:30 4 14 5 12:20 Robustaflavone 3 17:50 -- 2 4:20 Taiwaniaflavone 4 18:30 2 6:20 4 8:20 Hinokiflavone 4 18:30 1 3:40 1 3
Conditions: 1, 70% MeOH increased linearly to 90% MeOHover 15 minutes, eluting at 1 ml/minute; 2, THF:Chloroform 50:50, increasing linearly to 100:0 over 15 minutes, eluting at 2 ml/minute; 3, 100% THF, eluting at 2.5 ml/minute. All peaks detected at 320 nm. Section 3.3.2.4, p.22 increasing the percentage of methanol linearly from 70% to 90% over 15 minutes while maintaining an elution rate of lml/minute.
The permethyl ethers tested could not be adequately separated by this method (Table 7). Cupressuflavone hexamethyl ether was well separated from the other common series, but even the amentoflavone and hinokiflavone permethyl ethers were represented as a double peak without baseline separation,
A better separation of peremthyl ethers was obtained with the 10 um Radial Pak Silica column using a mixture of tetrahydrofuran and chloroform, increasing the percentage of tetrahydrofuran from 50% to 100% linearly over 15 minutes at an elution rate of 1.5ml/minute (Table 7). A baseline correction was incorporated.
The best resolution was obtained using the 5 um Radial Pak
Silica column with 100% tetrahydrofuran (Table 7, Fig. 3).
A trial sample of a permethylated extract of approximately
70 g. of dried leaves of Calocedrus decurrens was run through the 5 um Silica column in order to isolate amentoflavone, taiwaniaflavone, cupressuflavone and hinokiflavone permethyl ethers for further analysis (Fig. 3). After injection of approximately half the extract had been completed, however, the resolution of the column changed and the backpressure increased from 150 psi to 400 psi. It was concluded that the method of sample preparation using sep-pak cartridges was inadequate for the continued use of HPLC for separation of permethyl ethers from permethylated raw extracts. The remainder of the permethylated extract of Oalocedrus decurrens was subjected to centrifugally accelerated TLC. Section 3.3.2.4, p.23
3.3.2.5 Centrifugally accelerated TLC - 'Chromatotron'.
Silica gel PF of layer thickness 1mm was used with
tetrahydrofuran as the eluting solvent at a rate of
2-4mls/minute. The remaining permethylated extract of
Calocedrus decurrens was divided into 3 equal volumes, each of
approximately 2 mls. During the initial chromatographic run,
fractions were collected at 30 second intervals. Each fraction
was subjected to TLC in BPF to determine which contained the
permethyl ethers, the degree of separation of the permethyl
ethers and the purity. The fluorescent hexamethyl ethers could
be easily detected on the plate using a UV lamp during a run,
and in the later runs, were simply collected as they were spun
off. This proved to be a quick and effective semi-preparative
technique for the chromatography of large samples of
permethylated extracts.
3.3.2.6 Autofluorescence.
Permethyl ethers autofluoresce brightly in UV, and the
colour of fluorescence varies noticeably between the different
ethers (Table 6). This has previously been described in a
subjective manner; there is also the complication that initial
fluorescence after chromatography using acid solvents is very
bright, but the colours fade with time (Chexal et al. 1970; Lin
and Chen 1975; Dossaji et al. 1975b; Ilyas et al. 1978). The
initial fluorescence is probably an effect of protonation of the
ethers in the acid conditions. Examination of plates left for
48 hours after running in BPF showed lighter and less intense
fluorescence in colours of white and blue, which are similar to
those observed immediately after using BPEFD. This suggested Section 3.3.2.6, p.24
that the evaporation of the BPF solvent had removed the acid
conditions.
Emission spectra were obtained for the common permethyl
ethers in the family using a Perkin-Elmer fluorescence
spectrophotometer. They reflect the non-protonated ethers as
the permethyl ethers were dissolved in methanol. It can be seen
that the colours observed under UV within 30 minutes of removal
of the TLC plate from BPF (i.e., the fluorescence of the
protonated forms) are a very good reflection of the differences
shown by emmission spectra (see Table 6), and provide a good
basis for the initial identification of the interflavonyl link.
3.3.2.7 Mass Spectroscopy (MS).
Facilities for MS by chemical ionization were available.
Samples of permethyl ethers of amentoflavone, robustaflavone and
hinokiflavone obtained from a leaf extract of Dacrycarpus
compactus (Wassch.) de Laub. (Podocarpaceae), agathisflavone
hexamethyl ether from Araucaria cunninghamiani, and
cupressuflavone hexamethyl ether from Juniperus virginiana L.,
provided mass spectra whose parent peak was the same as that
expected for each permethyl ether series. Permethyl ethers of
amentoflavone, agathisflavone, cupressuflavone and
robustaflavone gave peaks at 623.18 + 0.5, representing the
molecular weight plus a positive ion. The actual molecular
weight (622.18 + 0.5) is very close to the theoretical molecular
weight expected for these hexamethyl ethers (622). Similiarly,
hinokiflavone pentamethyl ether gave a peak at 609.18, the
actual molecular weight (608.18) very close to the expected
value (608), The spectra were not, however, of sufficient p,p.m.
O
1
EI'HER.
2
IIEXAMmHYL
3
4
CUPRESSUFLAVONE
OF
TRACE
5
NMR
H
1
4.
6
FIGURE
7
8 TABLE 8. CHEMICAL SHIFTS OF PROTONS IN CUPRESSUFLAVONE HEXAMETHYL ETHER. Signal ( ) ppm No. of Protons2 Assignment2 Found1 Li terature2
7.33(d) 7.3O(d) 4 2'6' and 2"'6"' protons (ortho coupling) 6.77(d) 6.7O(d) 4 3'5' and 3'"5'" protons (ortho coupling) 6.59(s) 6.6O(s) 4 33" and 66" protons 4.13(s) 4.15(s) 6 two OMe groups 3.86(s) 3.85(s) 6 two OMe groups 3.78(s) 3.75(s) 6 two OMe groups
1, see Figure 2 , from Murti et al. (1967); s, singlet; d, doublet. Section 3.3.2.7, p.25
quality to enable an analysis of structure by fragmentation
pattern.
3.3.2.8 Nuclear Magnetic Resonance Spectroscopy (NMR).
Proton NMR was available, and was used to obtain
information on the number and position of methoxy groups, and
thus some confirmation of the identification of the permethyl
ethers.
Initial tests were conducted using samples of permethyl
ethers of amentoflavone and cupressuflavone derived from
Juniperus virginiana, and taiwaniaflavone permethyl ether
derived from Calocedrus decurrens. 100MHz 1H spectra in
deuterochloroform using TMS as an internal reference (0 ppm)
were obtained in a Jeol JNM-FXl00 spectrometer at 25°C.
Evaluation.
The chemical shifts of cupressuflavone hexamethyl ether
(Table 8, Fig. 4) align with those reported in the literature
(Murti et al. 1967), confirming the identification of this
distinctive hexamethyl ether. The NMR spectra of amentoflavone
and taiwaniaflavone permethyl ethers gave peaks consistent with
the presence of 6 methoxy groups in each (3.77, 3.78, 3.86,
3.92, 3.96 and 4.09 ppm for amentoflavone permethyl ether; 3. 71,
3. 75, 3.85, 3.87, 3.90, 3.96 ppm for taiwaniflavone peremthyl
ether) confirming their identification as hexamethyl ethers.
However, further analysis of these samples, was hampered by poor
sample quality.
3.3.3 Determination of methylation patterns. Section 3.3.3.1, p.26
The determination of the pattern of methylation of isolated
methyl ethers primarily involved chromatographic comparisons with
reference compounds and chromatographic data reported in the
literature. Support for the identification was also gained from
two other techniques.
3.3.3.1 Partial demethylation.
Partial demethylation was achieved by heating with
pyridinium chloride to 140-150 oC for 5-7 hours (Beckmann et al.
1971). The crude product was extracted with ethyl acetate and
the products isolated by TLC (3.2.1).
Evaluation.
The technique was tested on a number of reference compounds
to check whether there was any re-arrangement of the molecule
during the reaction, and to check the identity of some reference
compounds. Partial demethylation of the standard of
amentoflavone 74'4"'- trimethyl ether produced 4 bands in BPF:
DMS-2 to -5 (Fig. 5). DMS-2 co-chromatographed with 74'
dimethyl amentoflavone, and DMS-3 with 4'4"' dimethyl
amentoflavone. The Rfs of DMS-4 and DMS-5 corresponded with
those of monomethyl ethers. DMS-4 co-chromatographed with a
standard of 4' monomethyl amentoflavone, and DMS-5 with 4'"
monomethyl amentoflavone.
The ease of demethylation varies with the position of the
methoxyl group and its relationship to the interflavonyl link
(Geiger and Quinn 1975). Since the demethylation method had
failed to produce the 74"' dimethyl ether, it seems likely that
the 4' position is not as susceptible to demethylation as are FIGURE 5. PRODUars OBTAINED FROM DEl'TEI'HYLA'I'IONOF 74'4"' TRIME'rHYL AI>'lENTOFLAVONE(DI-1S)
AND 7"4"1 DIMEil1HYL .AMENTOFLAVONE(DM) •
CH3 0 9CH3 /~OH 0 CH3 0 OH HO 9'H3 HO 9CH3
DMS 2 OH 0 DMS 3 OH 0 DM OH 0 l \ l OH {');:OCH3 I HO, _,.,... _o_ ' __..OH HO, ,,.,.._ - o _ {');: __..9CH3 ao ~ ~ Ho,~ _o, ~ __..OH
DMS 4 OH 0 DMS 5 OH 0 DM 2 DM 1 OH 0 Section 3.3.3.1, p.27
the other two positions, which appear to be less structurally
hindered (Fig. 5).
Partial demethylation of amentoflavone 7"4"'- dimethyl
ether produced two bands in BPF, DM-2 chromatographing with 4"'
monomethyl amentoflavone and also with DMS-5, while DM-1
chromatographed with 7" monomethyl amentoflavone.
3.3.3.2 UV-spectroscopy.
UV-absorption spectra, which are important in the
identification of most flavones (Mabry et al. 1970; Markham
1982), are considered to be of very limited value in the
identification of biflavones, as each flavonoid chromophore
responds independently to the various shift· reagents used to
determine hydroxylation patterns (Geiger and Quinn 1975). The
small number of biflavonoid spectra reported by Dossaji et al.
(1975b) show that it is possible to gain some information,
however, on the hydroxylation patterns of some biflavonoids.
Spectra were obtained using the techniques for flavonoids
described by Mabry et al. (1970) and Markham (1982). Shifts in
the neutral spectrum after addition of sodium methoxide, sodium
acetate and aluminium chloride were recorded and interpreted
according to Mabry et al. (1970) and Dossaji et al. (1975b).
Evaluation.
The neutral UV absorption spectra of biflavonoids typically
consist of two absorption maxima in the ranges 265-275 nm (band
II) and 320-325 nm (band I). In sodium methoxide, in which all
phenolic hydroxyl groups will be ionised to some extent (Markham
1982), a large bathochromic (positive) shift of band I (+40 -
+70 nm) with the same or increased intensity indicates the FIGURE 6. UV SPECTRAL TRACES OF 77" CUPRESSUFLAVONE DIMNl'HYL EI'HER
..... MeOH - \ +NaOMe -- OH 0 \ \ CH 30 \ OH OH
I CH 0 ' -' 3 OH 0
CUPRESSUFLAVONE 7 r DIMETHYLETHER '.,,. I
SPECTRAL DATA 11'max, nml 250 290 330 370 MeOH 269, 320. A,nm + NaOMe 269, 360. +NaOAc 269,320.
MeOH - +NaOAc- -
-
250 290 330 370 A,nm nml
0
348
365
DIMETHYLETHER
OH
[).max,
4•
7•
300.h,
DATA
280,300.sh, 276,
270,325
•
0
OH
NaOAc
MeOH
EI'I!ER
SPECTRAL
+
+NaOMe
HO
AMENTOFLAVONE
DIMEI'HYL
'
'
4"
7"
370
--
-
~-,
/\
VONE
I
/
MeOH
+NaOMe
330
'-
\
\
\
\
I
.AMENTOFLA
A,nm
\
\
OF
\
290
\
\
I\
I
I
I
TRACES
250
SPECTRAL
UV
'
370
\
7 • 7
--
-
MeOH
330
FIGURE
+NaOAc
A,nm
\
\
\
\
\
290
\
\
I'
I
I
I
I
I
I
'
I
250 Section 3.3.3.2, p.28 presence of free 4' and/or 4'" hydroxyl groups in biflavonyls based on a biapigenin structure (Fig. 6). A smaller shift in band I ( + 35-50 nm) with a decrease in intensity coupled with a small bathochromic shift of band II (+5 - +15 nm) with an increase in intensity indicates the presence of a free 7 (or 7") hydroxyl group. Where both 4' (or 4"') and 7 (or 7") hydroxyl groups are free, the shift in band I in the presence of sodium methoxide is again large, but of decreased intensity.
Confirmation of the presence of a free 7 (or 7") hydroxyl group is shown by a bathochromic shift of 5-15 nm in band II in the presence of sodium acetate (Fig, 7). This reagent causes significant ionisation of only the most acidic hydroxyl groups, and is therefore useful in distinguishing between the 4' and 7 hydroxyl groups (Markham 1982).
The identification of some methyl ethers of biflavones by
UV absorption spectra alone is not feasible. For example, 7 4' and 7"4'" dimethyl amentoflavone show almost identical spectral shifts: in sodium methoxide, the respective shifts were +54 and
+40 nm with decreased intensity for band I, and + 12 and + 10 nm with increased intensity for band II. In each case there is a clear indication that both a 4' and 7 hydroxyl group are unsubstituted. These two methyl ethers could, however, be clearly separated by TLC on silica gel and cellulose (see Table
4).
Neither UV absorption spectral shifts nor TLC could distinguish between 74'4'" and 7"4'4'" trimethyl amentoflavone.
Both showed small bathochromic shifts of band I in sodium methoxide (+26 and +46 with reduced intensity respectively), while the shift of band II in sodium acetate was consistent with Section 3.3.3.2, p.29
a free 7 hydroxyl group (+12 and +6 respectively). A third
trimethyl ether, 77"4'" amentoflavone, isolated from Hevea
brasiliensis (Chandramouli et al. 1971; Quinn and Gadek 1981),
could be distinguished by its sodium methoxide shift in band I
of +65 nm without reduced intensity, indicating the presence of
the free 4' hydroxyl group.
Of the two products isolated from the demethylation of
7"4'" dimethyl amentoflavone (3.3.3.1), DM-1 showed a sodium.
methoxide shift in band I of +60 nm but with decreased intensity
coupled with a shift in band II of +7 nm. The large shift in
band I is consistent with the presence of two free 4' hydroxyl
groups combined with a free 7 hydroxyl group. The second
demethylation product, DM-2, was not produced in large enough
· concentrations to be examined.
A cupressuflavone dimethyl ether extracted from Araucaria
araucana was chromatographically identical with that isolated
from A. bidwillii and A. cunninghamii; 77" dimethyl
cupressuflavone was reported as the only dimethyl ether of
cupressuflavone in the latter two species (Ilyas et al. 1977b,
1978). The compound showed a sodium methoxide shift in band I
of +40 nm but no shift of band II in sodium acetate. These
observations suggest that both 7 positions are methylated and
both 4' positions are free. They are also comparable with the
shifts reported for 77" dimethyl cupressuflavone by Dossaji et
al. (1975b): a sodium methoxide shift in band I of +58 nm and no
sodium acetate shift of band II.
The application of UV absorption spectra to the
identification of compounds isolated by TLC is limited, however,
by the microscale at which many extractions (of herbarium Section 3.3.3.2, p.30 specimens in particular) are carried out: it was often not possible to obtain spectra of sufficient quality. Hence heavy reliance was often placed on TLC comparisons with standards in the different solvent systems, as well as permethylation of the crude extract, in making identifications. TABLE 9. RELI.Al3ILITY OF BIFLAVONOID PA'l'TERN.
:E:xtract Raw Permethylate:i Biflavonoid series Am ------Cu Hi Tw Am Cu F.i 'I'w methylation pattern 4' 4111 7 11 4'" 74' - 7711 - m - m
Calli tris maclea,y~ mrsw 12864, iv.85, cult., frssh • 0 • r,:sw 9239, i.80, Comboyne, NSW • 0 • Ul~SW 7485, xii. 79, Wilson R., NSW • 0 Callitris columellaris • UNSW 9341, v.80, Parkes, NSW • 0 • u:rnw 9287, ii.BO, Euston, i;sw • 0 Callitris endlicheri • w:sw 9339, ii.BO, Molong, ;.sw • 0 • UXSW 9291, v.81, Goolgowi, NSW • 0 • Calli tri s preissii ssp. murrayensis m:S",1 9288, v.Bo, l:.ildura, "'SW • 0 • Callitris p~ ssp. verrucosa u:,sw 9290, v.80, Walpeup, Vic • 0 • PlatY.cladus orient~lis 1:sw 12.i.22, cult. • C C • • 0 • m:s·,1 10337, viiLPC, cult. • 0 0 • • 0 • :~·~te 551 , ii.82, cult. • 0 0 • • 0 • -':1huJ2 p~ ::etcalf viii.20, British Columbia • 0 • 0 C:s.lder 18552, vii.56, II • 0 Calocedrus decurrens • u;;s,1 17 582, viii.85, cult., fresh • 0 0 • • • 0 • • • • ',Yai te 1242, ii.82, cult. • 0 0 • • • 0 • • • • l)arks 24251, viii.43, USA • 0 0 • • • 0 • • Junip~ communis • • :J~~ ,3";{ 10343, vi.61, British Columbia • • 0 Waite 744, ii.82, cult. • • 0 Jun:ip~ ~xycedrus Ferguson 2924, ii.71, Spain • • 0 Vlai te 740, ii.82, cult. • • 0 Junip~ virfilniana ~'lai te 570, ii. 82, cult. 0 • " ,, • • • • ~-i:1i te 617, ii.82, cult. • 0 • 0 • • • Papuace'.!r.is ;:,~p~ 1!!TSW 4206, vi.74, New Guinea • • 0 • • U:N'SW 4213, vi.74, New Guinea • • 0 • • \~gj;onia (llipresscides Skpf 989, vi.20, South Africa • • C " • • 0 • Darren-Smith viii.45, cult. 0 0 • • • ,, " • • • Waite 128:ia, ii.82, cult. • • 0 • • • • • ~etcria cedarbergensis
',faite I) (I 1275a, ii.82, cult. • • ,, • • • • • Waite 1231, ii.82, cult. • • • ll • • • • Neocalli tropsis pancheri Hartley 15068, Yio79, Kew Caledonia • • 0 0 (1 0 • • 0 i:cPherson 4601, vi.82, II II • • 0 0 0 • • 0
Am, amentoflavone; Cu, cupressuflavone; Hi, hinoki flavone; Tw, t ai wani aflavone.
t;· major band; o, minor band, 11, trace detected. by TLC only. TABLE 10. BIFLAVONOID DERIVATIVES DETECTED IN LEAF EXTRACTS.
Interflavonyl link 3'8" 8 8" 6-0-4'" 3'3" 3'6"
Taxon Compound 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Junipereae Juoiperus beraudiana + • m + • t J. c~uois + m J. drupacea + + + J. excelsa + + + J. oxycedrus + + J. procera + 1B + t J. virginiana + t t + m Cupresseae Cupressus lusitanica + • + C. seapervirens + m + Cha.aecyparis fon,osansis + + m + m m C. la"'8oniana 'Erects' + + m m m C. nootkatensis + + m m m + m C. thyoides + + m m + Fokienia hodginsii + m + + + t Thujopsideae Thujopsis dolobrata m t m + + + t m m Thuja koraiensis m m m m m T. occidentalis + m m m m + T. plicata + m T. standishii m t Platycladus orientalis + m m + Calocedrus decurrens + m m + II + m Libocedreae Neocallitropsis pancheri + + m m t Widdringtonia cedarbergensis + t + + t + ff. nodiflora (a) + + m t + ff. nodiflora (b) + m + + + m t Disel.a archeri + + + + t + Papuacedrus papuana + + m P. torricellensis + + t Pilgerodendron uviferi• + + t Libocedrus plUIIOSB + m L. yateensis + m Austrocedrus chilensis + + Actinostrobeae Actinostrobus acuainatus + t A. pyra.idalis + t Callitris canescens + t t C. coluaellaris + t C. endlicheri + t C. 1Mcleayana + t C. neocaledonica + + + + C. oblongs + + m C. preissii" + t C. sulcata + m t Fitzroya cupressoides + .. t + Tetraclinaceae Tetraclinis articulata + m +
*, see Table 9. 1, lllleJltoflavone; 2, 4'" aonomethyl amentoflavone; 3, 4' monoaethyl amentoflavone; 4, 4'4"' dimethyl aaentoflavone; 5, 4'7 dimethyl amentoflavone; 6, 4'"7" dimethyl aaentoflavone; 7, trimethyl aaentoflavone; 8, cupressuflavone; 9, 77" dimethyl cupressuflavone; 10, hinokiflavone; 11, 7" monoaethyl hinokiflavone; 12, taiwaniaflavone; 13, monoaethyl taiwaniaflavone; 14, robustaflavone; 15, monomethyl robustaflavone; 16, dimethyl robustaflavone; +, aajor band; m, minor band; t, trace detected by TLC only. TABLE ll. BIFLAVONOID PEIIMETHYL ETHERS DETECTED IN PEIIMETHYLATED LEAF EXTRACTS.
Taxon HAm HCu PHi HTw HRo HAg Ul
Junipereae Juniperus bel711Udi BDB + + + t J. cslifornics + + + J. chinensis + + + t J. ca.aunis + + + t J. conferts + + + t J. deppeBDB + + J. drupsces + + + t J. excelss + + m t J. foetidissiIItB + m m J. .anosperaa + + + t t J. oxycedrus + + m t J. procers + + + t t J. virginisns + + + t t Cupresseae Cupressus srizonics + + C. lusitsnics + + t C. seapervirens + + Ch811Becypsris f orwossnsis + + t C. lswsonisns 'Erecta' + + t C. nootkstensis + m + m C. obtusB + + C. pisifers 'Squarrosa' + + t C. thyoides + + t Fokienis hodginsii + + t Thujopsideae Thujopsis dolobrsts + + m Thujs korsiensis + + T. occidentslis + • + t T. plicsts + + t T. stsndishii + + Plstyclsdus orientslis + m + t Cslocedrus decurrens + + + + t C. for.a.sans + + C. .-scrolepis + +
Libocedreae Neocsllitropsis psncheri + + m Widdringtonis cedsrbergensis + + + If. nodiflors (a) + + m I(. nodiflors (b) + + + Disel.a srcheri + + + t Pspuscedrus pspusns + + P. torricellensis + + Pi lgerdendron uvi ferum + + Libocedrus bidwillii + + L. pluaoss + m L. }'llteenis + m Austrocedrus chilensis + + Actinostrobeae Actinostrobus scu.instus + A. pyra.idslis + m Csllitris csnescens + C. colu.ellsris + C. endlicheri + C. .-scles}'llns + C. 111Uelleri + C. neocsledonics + m C. oblongs + C. preissii* + C. sulcsts + Fi tzro}'ll cupressoides + + m Tetraclineae Tetrsclinis srticulsts + + t *• see Table 9. HAm, amentoflavone hexamethyl ether; HCu, cupressuflavone hexaaethyl ether; PHi, hinokiflavone pentaaethyl ether; HTw, taiwaniaflavone hexamethyl ether; HRo, robustaflavone hexamethyl ether; HAg, agathisflavone hexamethyl ether; Ul, unknown product;+, major band; m, minor band; t, trace detected by TLC only. Section 4, p.31
4. DISTRIBUTION OF BIFLAVONOIDS IN THE CUPRESSACEAE.
Using the methods described above, representatives of all genera
of the Cupressaceae were investigated for leaf biflavonoids. The
distribution of the various biflavonoids amongst the species is given
in Table ·10, and the distributions of permethyl ethers in permethylated
raw extracts from a broader range of taxa are given in Table 11. In
both tables, the taxa are arranged according to Li (1953).
4.1 Reliability of biflavonoid pattern within species.
The plant material utilised in this survey comes from a variety of
sources, either cultivated specimens or natural populations, fresh or
dried collections and occasionally quite old herbarium specimens. The
influence of these variables on the biflavonoids detectable within
species was first investigated. Table 9 shows the results of
extractions of different collections of a number of species.
Extracts from fresh material of Callitris macleayana was found to
contain an identical complement of biflavonoids as that extracted from
two recent herbarium collections. Similarly, extracts from fresh and
63 year old herbarium material of Platycladus orientalis gave identical
results (Table 9).
Two extractions from herbarium material of natural populations of
Thuja plicata, one 65 years old and the other 29 years old, differed
only in the detection of a minor band of 4"' monomethyl amentoflavone
in the older collection. Hence, biflavonoids appear to be very stable
for long periods in dried specimens.
Comparisons of extracts of cultivated material and natural
populations were tested in a number of species. There was no Section 4.1, p.32 difference between collections from two natural populations (Wilson
River and Causeway Road, NSW) and cultivated material (Nat. Bot. Gdns,
Canberra) of Callitris macleayana, amentoflavone and its 4"' monomethyl ether only being detectable in all extracts. Similarly, an extract taken from a collection from a natural population of Calocedrus decurrens in Darlington, U.S.A., contained the same complement of biflavonoids as those from cultivated specimens in the Waite Arboretum,
South Australia, and at Cooma, NSW. Extracts of collections of cultivated and natural populations of both Juniperus communis and J. oxycedrus also contained identical complements of biflavonoids (Table
9, Appendix).-
The biflavonoids isolated from different natural populations of
Callitris columellaris, C. endlicheri, C. macleayana and Papuacedrus papuana proved identical in each case. In Callitris preissii, they were found to be identical in both subspecies. Differences were detected, however, in the concentration or detectability of minor or trace components in the patterns of Neocallitropsis pancheri, Thuja plicata and Juniperus virginiana (Table 9).
The variation that is apparent between different collections of
Widdringtonia cupressoides (Table 9) is therefore exceptional. There is some doubt, however, about the identification of this material (see
4.3), since, with one exception, it is taken from immature cultivated specimens.
With this one notable exception, there is a complement of biflavonoids, particularly with regard to the major constituents, which is characteristic of a species; this is referred to subsequently as the
"biflavonoid pattern" of the species. This conclusion accords with those of previous investigators on the reliability of the biflavonoid pattern within species of conifers (Quinn and Gadek 1981; Geiger and Section 4.1, p.33
Quinn 1982), the major biflavonoid constituents appearing to be under
direct genetic control, and almost invariably unaffected by ecological
or temporal factors. The data in Table 10 is based, as far as
possible, on collections from natural populations, and on at least two
collections of each species (see appendix), except where material was
not available or where there was a marked uniformity in the pattern
obtained across an entire genus.
4.2 Definition of character-states.
The occurrence of chemical constituents is most commonly assessed
and recorded qualitatively, that is, as present or absent from a
particular taxon. In this survey the presence of an operative pathway
for the biosynthesis of each biflavonoid series can be considered a
character and scored as 'present' or 'absent' for all taxa.
Permethylation of the raw extract of each taxon proved to be an
effective technique for the rapid scoring of this first set of
characters. These data (Table 11) show that the distribution of each
biflavonoid series is different amongst the taxa.
A series of subcharacters can be superimposed relating to the
particular derivatives of each series (partial methyl ethers or base
compound) which are present (Table 10). In addition, the extent to
which each derivative is accummulated in each species has been
subjectively assessed as major (a predominant component), minor
(readily isolated and characterised), or trace ( detectable, but
identified by TLC only).
4.3 Biflavonoid Patterns.
Considering the biflavonoid patterns in Tables 10 and 11 in
sequence, starting with members of the tribe Cupresseae, it can be seen Section 4.3, p.34 that all three species of Cupressus surveyed here contain cupressuflavone and amentoflavone as the major biflavonoids. A total of 7 species of the genus Cupressus have now been examined in various studies of the biflavonoid content of leaves. All are reported to contain cupressuflavone, and all but one, amentoflavone (Natarajan et al. 1970; Taufeeq et al. 1978; Lebreton et al. 1978; Taufeeq et al.
1979; Miura and Kawano 1968). The report of cupressuflavone alone in the leaves of C. arizonica (Miura and Kawano 1968) is an incomplete report, as it contains both amentoflavone and cupressuflavone (Table
11). These authors also reported cupressuflavone alone in C. goveniana. Two subsequent analyses of that species (Natarajan et al.
1970; Taufeeq et al. 1979) have revealed the presence of the amentoflavone and hinokiflavone series in addition to cupressuflavone.
The hinokiflavone series, however, is of variable occurrence in the genus, having been reported only from C. goveniana, C. lusitanica, C. funebris and C. torulosa (Natarajan et al. 1970; Taufeeq et al. 1978,
1979). Previous reports of its absence from C. sempervirens (Natarajan et al. 1970; Lebreton et al. 1978) were confirmed in this survey; neither was this series found in C. arizonica (Tables 10, 11). Hence, the genus Cupressus is typified by a leaf biflavonoid pattern having major bands of amentoflavone and cupressuflavone, with bands of hinokiflavone and some monomethyl ethers also often present.
All four species of Chamaecyparis analysed contain amentoflavone and a range of its partial methyl ethers (at least some of the latter as major bands), and also hinokiflavone. C. nootkatensis alone contains cupressuflavone. C. thyoides is distinguished by the presence of a major band of the 4' methyl ether of amentoflavone, rather than the 4"' methyl ether found in the remaining three species.
Chamaecyparis is therefore chemically heterogeneous. Section 4.3, p.35
The third genus of the Cupresseae, the monotypic Fokienia, contains bands of amentoflavone 4"'7" and 4'7 dimethyl ether, as well as amentoflavone and hinokiflavone.
Of the four species of Thuja (tribe Thujopsideae) examined, only
T. occidentalis contains cupressuflavone. T. koraiensis contains hinokiflavone, amentoflavone and a range of its partial methyl ethers in more or less equal concentrations. T. standishii and T. plicata each contain detectable bands of amentoflavone and a single monomethyl ether only (Table 10), although the hinokiflavone series is clearly detectable in the permethylated extract of both (Table 11).
The monotypic genus Platycladus shows some similarity in biflavone content to T. occidentalis, containing the same major biflavones
(amentoflavone and hinokiflavone) as well as minor bands of cupressuflavone and 4"' monomethyl amentoflavone.
The other monotypic genus in this tribe, Thujopsis, displays a distinctive biflavonoid pattern in which the more highly methylated amentoflavone derivatives are the major constituents, and both robustaflavone and its monomethyl ether are detectable as minor components.
Calocedrus decurrens is distinguished by the presence of major bands of cupressuflavone and the 3'3" linked taiwaniaflavone, as well as a minor band of 4"'7" dimethyl amentoflavone. Neither cupressuflavone nor taiwaniaflavone or their derivatives were detected in the permethylated extracts of the other two species of this genus
(Table 11). Thus the tribe Thujopsideae shows marked heterogeneity,
both within and between genera.
All the species of the monogeneric Junipereae surveyed contain a
major band of amentoflavone as well as a band of cupressuflavone; few
partial methyl ethers were detected. Hinokiflavone derivatives were Section 4.3, p.36
detected, at least in the permethylated extract, in all but one species
(Tables 10 and 11).
Twelve species of Juniperus have previously been examined for leaf
biflavonoids (Table 3), in several cases by more than one worker
(Lebreton et al. 1978; Lamer-Zarawska 1975; Fatma et al. 1979; Hameed et al. 1973; Ilyas et al. 1977a; Pelter et al. 1971; Pascual Teresa et al. 1980). Apparent contradictions in these reports appear to be
mainly due to incomplete analyses of the biflavonoid content being
reported. For example, Lamer-Zarawska (1975) recorded cupressuflavone, amentoflavone and 4' monomethyl amentoflavone in J. communis, Pascual
Teresa et al. (1980) recorded cupressuflavone and hinokiflavone (Table
3), while the present study revealed all three series of biflavones
(Tables 10, 11). Amentoflavone has been reported in all 12 species,
and cupressuflavone in all but four of them: J. sabina, J. squamata, J.
occidentalis and J. virginiana (Table 3). The last species was
included in the present survey, and was found to contain major bands of
both cupressuflavone and amentoflavone. Since the data for the other
three species is drawn from the same report (Lamer-Zarawska 1975), it
seems probable that a careful re-examination of them would also reveal
cupressuflavone to be a major constituent of the biflavone fraction.
It appears, then, that the cupressuflavone series is a characteristic
component of the leaves of the genus Juniperus. Typically, it is
present as the parental compound; there is only one report of a partial
methyl ether (77" dimethyl cupressuflavone) constituting the major
cupressuflavone component (J. recurva; Hameed et al. 1973).
Callitris (tribe Actinostrobeae), with 12 species, is the largest
genus in the subfamily Callitroideae. Nine species have now been
examined for biflavonoid content. Only one species does not conform to
a pattern of amentoflavone as the major band with possible minor bands Section 4.3, p.37 of the 4"' and 4"'7" methyl ethers: C. neocaledonica is exceptional in having major bands of all three compounds as well as the 4' monomethyl ether (Table 10). Only 2 species have been investigated previously
(Table 3). The only recent report of hinokiflavone from Callitris is for C. glauca (? = C. columellaris F. Muell; Ansari et al. 1981), where it was detected as a minor component. Examination of natural populations of Australian material of this species revealed no trace of hinokiflavone derivatives in either raw extracts or permethylated extracts (Tables 10, 11). This is in agreement with Siva Prasad and
Krishnamurti's observations on C. rhomboidea (1977), and the present survey of seven other species (Tables 10, 11). It can be concluded, therefore, that the occurrence of hinokiflavone in concentrations that can be detected by the methods used in this survey is not a feature of the genus.
Actinostrobus species contain the same pattern of biflavonoids as
Callitris; again hinokiflavone could not be detected. An unidentified compound (biflavonoid?) was detected in minor concentrations in the permethylated extracts of Actinostrobus pyramidalis and Callitris neocaledonica (Table 11), but insufficient material was available to characterise it.
The pattern of the third member of the Actinostrobeae, Fitzroya cupressoides, contrasts markedly with the patterns of the other two genera, being distinguished by the presence of a major band of 77" dimethyl cupressuflavone (Table 10), and by the detection of the
hinokiflavone series in the permethylated extract (Table 11).
The monotypic Diselma archeri contains the richest array of
biflavonoids in the tribe Libocedreae, and in the subfamily (Table 10):
amentoflavone, hinokiflavone and cupressuflavone methyl ethers, as well
as a trace of robustaflavone (Table 11). Both this genus and Section 4.3, p.38
Widdringtonia contain 77"- dimethyl cupressuflavone as the major cupressuflavone derivative.
Four collections representing two species of Widdringtonia have been examined. A chemical discontinuity within the genus is detectable, one collection of W. cupressoides and W. juniperoides (=
? W. cedarbergensis Marsh) containing 7" monomethyl hinokiflavone but not 4"'7" dimethyl amentoflavone, while in the remaining two collections of W. cupressoides the situation is reversed. The delimitation of the species within this genus is complex, relying on cone characters and distribution (Coates Palgrave 1983:59), and verification of the identity of sterile material, of which the voucher specimens are comprised, is impossible. The fact that three of the collections came from immature cultivated specimens increases the likelihood of misidentification. On the other hand, w. cupressoides is
highly variable over its large latitudinal range, being a forest tree at its northern limit and a lignotuberous shrub towards the south (C.
J. Quinn pers. comm.). It may well be that variations in biflavonoids also exist. Further material of all 3 species has been obtained from
natural populations, but is still in the process of analysis.
The biflavonoid patterns of the genera Austrocedrus, Libocedrus
s.s., Pilgerodendron and Papuacedrus are very homogeneous, the only
difference between them being the absence of hinokiflavone as a
detectable band from Austrocedrus and from both species of Libocedrus
s.s. examined (Table 10). However, hinokiflavone pentamethyl ether was
detected in the permethylated raw extracts of all four genera (Table
11).
The final member of this tribe, Neocallitropsis pancheri, contains
a distinctive array of biflavonoids based on amentoflavone,
hinokiflavone and taiwaniaflavone. Section 4.3, p.39
The monotypic Tetraclinis articulata, which is the sole member of
the tribe Tetraclineae, has amentoflavone and cupressuflavone as its
major biflavonoids. Dimethyl ethers of either series are not
detectable, while hinokiflavone derivatives are present in trace amounts detectable only by permethylation of the raw extract (Table
11).
Several methyl ethers reported exclusively from the Cupressaceae
were not identified in the present survey. The 4' monomethyl cupressuflavone reported only from Cupressus lusitanica var. benthamii
(Taufeeq et al. 1978) and C. govaniana (? = goveniana Gord.) (Taufeeq et al. 1979) was not isolated in this study, but a trace of what must
have been a monomethyl ether of cupressuflavone was detected during
permethylation of the band of the 4"' monomethyl amentoflavone from C.
lusitanica var. lusitanica. Similarly, the 77" dimethyl hinokiflavone
reported exclusively from various Cupressaceae (Table 3; Miura and
Kawano 1968) was not detected, nor was it reported in a re-examination
of Platycladus orientalis (Natarajan et al. 1970) or in more recent
studies of other species of the family. Section 5, p.40
5. CHEMOTAXONOMY
5.1 Polarity of chemical characters.
Geiger and Quinn (1975:736-737) concluded that the presence of
amentoflavone and hinokiflavone was ancestoral in the gymnosperms in
general. Amentoflavone derivatives occur in all the taxa studied in
this survey, and are of general occurrence in all other families of
conifers with the exception of Pinaceae. This certainly supports the
conclusion that the presence of this series is plesiomorphic in the
Cupressaceae. Hinokiflavone derivatives appear in representatives of
all genera except Callitris and Act.inostrobus. In view of its wide
occurrence in other conifer families (Table 2) and particularly its
occurrence in all but one of the genera of the Taxodiaceae (Geiger and
Quinn 1982; Gadek unpubl.), which is generally accepted as the most
closely related family and therefore the sister group of the
Cupressaceae (Eckenwalder 1976a), it must be concluded that the
presence of this series also is a plesiomorph in the family and that
its absence from Callitris and Actinostrobus is an apomorph involving a
loss of function in the biochemical pathway involved (i.e., an
evolutionary reversal).
The cupressuflavone series, which occurs in 10 of the 19 genera in
the Cupressaceae, displays a very interesting distribution in the
Coniferales (Table 2), being characteristic of Araucariaceae and also
occurring in all three species of a single genus (Lepidothamnus Phil.)
of the Podocarpaceae (Quinn and Gadek 1981). Its restricted
distribution in the latter family favours the view that it has evolved
independently in that family. Its absence from almost half the genera
in the Cupressaceae, when taken together with its total absence from Section 5, p.41 the sister group, the Taxodiaceae, suggests a similiar conclusion for this series in the Cupressaceae. Hence, it seems most likely that the ability to synthesise the cupressuflavone series has arisen at least 3 times in the Coniferales, and a fourth time in the Angiosperms (Geiger and Quinn 1982:526). On this basis, then, the presence of cupressuflavone is an apomorph within the Cupressaceae.
The taiwaniaflavone series is known to occur in only 2 very different species in the Cupressaceae. Despite its occurrence in
Taiwania cryptomeriodes in the sister group, its very limited distribution in these 2 families (see also 5.3), and the lack of any record of its occurrence from any other group of vascular plants suggest that its occurrence is best regarded as an apomorph in both families.
The robustaflavone series is not accumulated to any great extent by any member of the Cupressaceae (Table 10), mostly being detected in only trace amounts from the permethylated raw extracts (Table 11) and consequently identified entirely on Rf and autoflourescence. These traces are not to be explained, however, as the result of a structural rearrangement during permethylation, for instance, by a Wessely-Moser rearrangement as may occur during demethylation or dehydrogenation
(Geiger and Quinn 1975), since they were repeatedly detected in some taxa but never in other taxa (Table 11). The recorded occurrence of this series is extremely limited in the Coniferales (Table 2), although it is possible that similiar trace amounts may be detected by a wider survey using the techniques employed here. Whether or not the presence of the robustaflavone series is considered plesiomorphic in the
Cupressaceae, its sporadic occurrence even in trace amounts in Thuja (2 species of 4), Chamaecyparis (3 species of 4), and Juniperus (10 species of 12), would seem to indicate that there have been several Section 5, p.42
evolutionary reversals or homoplasies in the character. For all these
reasons, this character is less promising as an indicator of affinities
above the generic level.
In view of the widespread occurrence of the base compound of both
the amentoflavone and hinokiflavone series in the conifers as well as
in the Psilotales, Cycadales and angiosperms (Geiger and Quinn 1982),
where it is commonly the only member of either series recorded, it is
concluded that the presence of partial methyl ethers represents a
derived character state. This accords with the proposed biosynthetic
pathway for methylation of flavones in, for example, parsley
(Petroselinum hortense Hoffm.), where methylation occurs subsequent to
the formation of the base compound (Ebel and Hahlbrock 1982:652; see
Fig. 2).
The occurrence of identical partial methyl ethers (e.g., 77"
dimethyl cupressuflavone in Diselma, Widdringtonia and Fitzroya) should
be taken as an indicator of closer affinity than the occurrence of
different derivatives having the same degree of methylation (e.g., the
presence of 4"' monomethyl amentoflavone in Fitzroya instead of the 4'
monomethyl ether found in Diselma and Widdringtonia); the latter is
evidence of independent specialisation.
· 5.2 Correlation with existing generic boundaries
Whilst there is a high degree of uniformity in the biflavonoid
series present within most genera, discontinuities exist in Calocedrus,
Chamaecyparis and Thuja. Chamaecyparis nootkatensis and Thuja
occidentalis are each characterised by the presence of the
cupressuflavone series in addition to the series that are present in
the remaining species of these genera, while Calocedrus decurrens
contains both the cupressuflavone and taiwaniaflavone series in Section 5.2, p.43 addition to those found in the remaining species of that genus (Table
10). Considering the high degree of uniformity in all other genera,
these discontinuities suggest that the delimitation of these genera
should be critically reappraised.
In all three cases, previous authors have commented on the
distinctiveness of the particular species on other characters. Florin and Boutelje (1954) noted that in Calocedrus decurrens the ultimate
branchlets are less strongly dorsiventral, with the facial leaves
undifferentiated and the abaxial stomatal bands of the marginal leaves
not reduced. This was contrasted with the stronglr dorsiventral
ultimate branchlets with distinctive upper and lower facial leaves, and abaxial stomatal bands of the marginal leaves extremely reduced in the
two Asian species (C. macrolepis and C. formosana). This distinction
was reinforced by observations on wood anatomy (Bannan 1944; Boutelje
1955): C. decurrens is distinguished from the remaining two species in
having ray parenchyma with conspicuous nodular thickeninge rather than
fine thickenings on the tangential walls and markedly more thickened
horizontal walls, deep indentures as distinct .from indentures very rare
and then only shallow, and croesfield pits distinct and small (mean
diameter 3um) rather than indistinct and larger (mean diameter >
4.5um). Erdtman and Norin (1966:251) also noted great differences in
the terpene constituents of C. decurrens and C. formosana, although
their survey did not include C. macrolepis.
In the case of Thuja occidentalis, Phillips (1941) reported its
wood to be atypical of the genus in two respects: viz., tangential
walls of ray parenchyma nodular rather than smooth, and transverse
walls of vertical parenchyma smooth instead of conspiuously nodular.
This species is also distinct in the genus in having only slightly Section 5.2, p.44 winged seeds in contrast to the markedly winged seeds of the other species (Dallimore and Jackson 1966; Gaussen 1968).
The position of Chamaecyparis nootkatensis has been questioned by several authors. Phillips (1941) and Bannan (1952) commented on the frequent occurrence in this species of short rays composed entirely of ray tracheids, which they regarded as a very distinctive feature, and the nodulation of the tangential walls of the ray parenchyma. Bauch et al. (1972) described the torus structure in the intertracheal pits of the spring wood as disc-shaped, resembling those found in Pinus and quite unlike those found in the remainder of the genus Chamaecyparis.
A study of cuticular characters led Alvin et al. (1981) to conclude that C. nootkatensis occupied a position remote from the other species in the genus. Erdtman and Norin (1966) drew attention to the fact that this was the one species of Chamaecyparis to contain the 015 tropolone, nootkatin, which also occurs in Cupressus and Juniperus, and concluded that "a careful reconsideration of the systematic position of this species is indicated." (p.250). von Rudloff (1975:162) stated that there are large chemical differences amongst the Chamaecyparis species, and that C. nootkatensis "has unique terpenoid and non-terpenoid foliage oil components". Added to these differences are the facts that the pollen is one-celled rather than two-celled at pollination, the proembryogeny is quite unlike that reported for other species (Owens and Molder 1975), the foliage leaves are devoid of the white stomatal lines that characterise all other species of the genus, and the ovulate cones ripen in the second year rather than in the first (Dallimore and
Jackson 1966). Finally, C. nootkatensis is the only species in the family known to hybridise outside its genus (Dallimore and Jackson
1966). Section 5.2, p.45
Thus, in every case where there is a discontinuity in the
biflavonoid series present within a genus, there are correlated
discontinuities in other characters which suggest that the generic
boundaries may be inappropriate. Apart from these three exceptions
where the generic boundaries appear questionable, there is a strong
correlation between presently recognised genera and the distribution of
different biflavonoid series. This indicates that the biflavonoid
pattern is providing accurate information on affinities at the generic
level.
5.3 Biflavonoid patterns in relation to taxonomy above the generic level.
It is apparent from Section 4.3 that there are marked
discontinuities in biflavonoid pattern within the family, and that
while these generally coincide well with existing generic boundaries,
with the noted exceptions of Calocedrus decurrens, Chamaecyparis
nootkatensis and Thuja occidentalis, they in no way align with the
tribal or subfamily boundaries proposed by Li (1953). The distribution
of the cupressuflavone series in particular cuts straight across the
tribal and subfamily boundaries, occurring in both subfamilies and 5 of
the 6 tribes. There is a remarkable similarity between the biflavonoid
patterns of Tetraclinis from the subfamily Callitroideae and Cupressus
and Juniperus of the tribes Cupresseae and Junipereae respectively in
the subfamily Cupressoideae, while other members of the Cupresseae
possess a very different pattern.
The distinctive occurrence of the 77" dimethyl cupressuflavone in
Fitzroya (Actinostrobeae) suggests a closer relationship to Diselma and
Widdringtonia (Libocedreae), which also contain this biflavone as the
major cupressuflavone derivative, rather than to Callitris and TABLE 12. GROUPINGS SUGGESTED BY BIFLAVONES.
Group Subfamily, Tribe1 Taxa Apomorphic State +cu -Hi +Tw
lA Cup, C Cupressus + Cup, J Juniperus + Call, T Tetraclinis + 1B Cup, Tj Calocedrus decurrens + + lC Call, A Fitzroya + Call, L Oisel/118 + Call, L h'iddringtonia + 1D Cup, Tj Platycladus + Cup, C ChBlllaecyparis nootkatensis + Cup, Tj Thuja occidentalis +
2 Call, A Actinostrobus + Call, A Callitris +
3 Call, L Neocallitropsis +
4 Cup, Tj Thujopsis Cup, Tj Thuja s.s Cup, C Fokienia Cup, C ChBlllaecyparis s.s. Cup, Tj Cal ocedrus s. s. Call, L Libocedrus Call, L Papuacedrus Call, L Aus trocedrus Call, L Pilgerodendron
1 from Li (1953); Cup, Cupressoideae; Call, Callitroideae; J, Juniper.eae; C, Cupresseae; Tj, Thujopsideae; L, Libocedreae; A, Actinostrobeae; T, Tetraclineae. +cu, cupressuflavone series present; -Hi, hinokiflavone series absent; +Tw, taiwaniaflavone series present. Section 5.3, p.46
Actinostrobus (Actinostrobeae), in which the cupressuflavone series is unknown.
The presence of taiwaniaflavone and its partial methyl ethers as minor fractions in Neocallitropsis raises the possibility of an affinity with Calocedrus decurrens, athough the absence of even trace amounts of cupressuflavone from the former clearly distinguishes it.
The fact that taiwaniaflavone also occurs in the Taxodiaceae raises the suggestion of a closer relationship between the Cupressaceae and
Taxodiaceae than is presently recognised, although re-combining the two families has recently been proposed by Keng (1975) and Eckenwalder
(1976a, b). This compound is not, however, a common constituent in either family. I have been unable to detect. it in leaf extracts of some other taxodiaceous taxa: viz., Cryptomeria japonica, Sciadopitys verticillata, Taxodium distichum var. imbricarium, Athrotaxis selaginoides, Sequoia sempervirens, Cunninghamia konishii or C. lanceolata. Nor has cupressuflavone, which occurs in a number of cupressaceous genera, been detected in any Taxodiaceae (see also Geiger and Quinn 1975, 1982). It seems likely, then, that the taiwaniaflavone series has evolved separately in each of the three species in which it has been detected, although this hypothesis is open to testing by a detailed analysis of other data sources in order fully to reassess affinities.
The biflavonoid patterns can be used to construct the following groups within the family, based on synapomorphs (Table 12):
Group 1. Characterised by the presence of the cupressuflavone
series.
Subgroup A. Cupressus, Juniperus and Tetraclinis - the
biflavonoid pattern of these genera comprises major bands of
amentoflavone and cupressuflavone; the more highly methylated Section 5.3, p.47
biflavones are typically absent.
Subgroup B. Oalocedrus decurrens - distinguished by the presence
of taiwaniaflavone as well as a major band of cupressuflavone.
Subgroup C. Fitzroya, Widdringtonia and Diselma - characterised
by the presence of the cupressuflavone fraction as the 77"
dimethyl ether, and also possessing major amounts of
amentoflavone and its partial methyl ethers. The monotypic
Fitzroya is distinguished from the other two genera by the
pattern of amentoflavone derivatives.
Subgroup D. Platycladus, Thuja occidentalis and Chamaecyparis
nootkatensis - characterised by the presence of minor amounts
of cupressuflavone in combination with hinokiflavone and
amentoflavone and its partial methyl ethers.
Group 2.
Oallitris and Actinostrobus - characterised by the absence of
detectable amounts of hinokiflavone derivatives, so that the
amentoflavone series alone is present.
Group 3.
Neocallitropsis - characterised by minor bands of
taiwaniaflavone, along with major bands of amentoflavone and a
monomethyl amentoflavone, and minor bands of hinokiflavone, a
monomethyl hinokiflavone, and a dimethyl amentoflavone.
Group 4.
The remaining taxa constitute a rather heterogeneous
assemblage sharing no derived character in their biflavonoid
patterns. All are characterised by the presence of Section 5.3, p.48 amentoflavone and varying partial methyl ethers, while the hinokiflavone series is detectable at least in the permethylated extracts. Th ujopsis dolobrata and
Chamaecyparis thyoides are characterised within the group by the presence of the 4' monomethyl amentoflavone in contrast to the more usually occuring 4"' monomethyl ether, and the former species is further characterised by the minor occurrence of robustaflavone and its monomethyl ether. Fokienia hodginsii and Chamaecyparis formosansis share the 4'4"' dimethyl amentoflavone in addition to the more usual 4'7 dimethyl ether.
Chemically, Libocedrus, Papuacedrus, Austrocedrus and
Pilgerodendron comprise a highly uniform group characterised by major amounts of amentoflavone and its 4"' monomethyl ether, as well as the hinokiflavone series detectable at least by permethylation. TABLE 13. DISTRIBUTION OF CHARACTER-STATES FROM OTHER DATA SOURCES.
. "'.., "':::: ....,:i; a: "' . .., ::'s ~ ....."' g ~ :::: "' v .::: ."'.., "' lrans. trac. 2 2 2 1 0 2 2 0 0 0 0 0 Ray Par·enchyma: thick hori. + + 0 + + 0 + + 0 0 0 + + + 0 0 + 0 + 0 abundant pit. + + 0 + + + 0 + + + 0 0 0 + 0 + + + 0 + nod. tang. 2 2 0 2 2 2 0 0 2 2 0 0 0 0 0 0 1 1 0 0 1 0 Axial Parenchyma: thick hori. 2 2 0 2 2 2 0 2 0 2 0 0 0 2 2 2 2 2 1 1 indentures 2 2 0 2 1 1 0 2 1 1 0 0 0 1 1,2 1 1,2 1 1 0 0 1 Tropolones: thuja··type + + + 0 + 0 0 0 + + 0 + + 0 + + 0 noot.-type + + 0 0 0 0 0 + 0 0 0 0 0 0 0 0 0 0 0 Diterpenoid Resins: call. ac. 0 0 0 0 0 0 0 0 0 0 + 0 0 0 0 0 lamb. ac. 0 0 0 + 0 0 0 + 0 0 0 0 0 0 0 0 tor. ac. + 0 + 0 0 0 0 0 0 0 0 0 0 0 + + dehydroab. ac. 0 0 0 + 0 0 0 0 0 0 0 0 0 0 0 0 agathic ac. + + 0 0 0 0 0 0 ? 0 0 0 0 0 0 un. ac. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 + un. cam. 0 0 0 0 0 0 + 0 0 0 0 0 0 0 0 0 ferruginol + + + 0 + + ? + + + ? + 0 + + + totarol + + + 0 0 0 0 + + + 0 0 0 + 0 0 abietol 0 0 0 ? 0 0 0 0 0 0 0 0 + 0 0 0 acetoxy. 0 0 + 0 0 0 0 0 0 0 0 0 0 0 0 0 tol. acetate 0 0 + 0 0 0 0 0 0 0 0 0 0 0 0 0 ' .. ' signifies that the condition in the genus is unknown. leaves/node, in whorls of 2, 3, or 4; dimorph. leaves, presence (+) of dimorphic leaves; trans. trac., the presence of highly ramified (2)' thickened (1) or unthickened (0) pit borders of transfusion tracheids; ray parenchyma, thick hori., thickened (+) horizontal walls; abundant pit.' abundant pitting (+) of the horizontal walls; nod. tan., distinct (2), indistinct (1) or no (0) nodular thickenings of the tangential walls; thick hori., distinctly thickened (2)' finely beaded (1) or unthickened and smooth (0) horizontal walls of the axial parenchyma; indentures, indentures distinct (2)' indistinct (1) or absent (O); thuja-type, the presence (+) of thuja-type tropolones; noot.-type, the presence (+) of nootkatin-type tropolones; Diterpenoid resins, (+) present, (0) absent, (?) peak detected but not identified; call. ac., callitrisic acid; lamb. ac. 1 lambertianic acid; tor. ac., torulosic acid; dehydroab. ac., dehydroabietic acid; agathic ac., agathic acid; un. ac., unidentified acid; un. cam., unidentified compound; acetoxy., 12B-acetoxysandaracopimaric acid; tor. acetate, torulosyl acetate. Section 6, p.49 6. OTHER DATA SOURCES. 6.1 Introduction. The existence of a single synapomorph in two otherwise different genera is as likely to be the result of homoplasy than to be the result of its inheritance from a common ancestor, and may not, therefore, necessarily indicate a phylogenetic affinity between the two genera. A recent common ancestory should result in similarities in a range of character-states, due to the inheritance without modification of some of the ancestoral conditions by both descendant taxa. Hence, the phylogenetic relationship between genera or groups of genera is only reflected by constructing groups based on the presence of a number of synapomorphs. In view of the lack of agreement between the groupings of Li (1953) based on ovulate cone characters, and those based on biflavonoid pattern, it is appropriate to survey other data sources to see what support exists for either groupings. 6.2 Leaf characters. All the members of the Cupressaceae have an opposite or whorled adult phyllotaxis (Table 13). Neocallitropsis, the sole member of group 3, is distinguished by its alternating whorls of 4 leaves. The group 2 genera, Oallitris and Actinostrobus, are characterised by a ternate phyllotaxis, but this state also arises in the monotypic Fitzroya and in certain species of Juniperus (both in group 1). Section 6.2, p.50 Patterns of the vascular traces in vegetative shoots have been studied in 14 genera (Namboodiri and Beck 1968; Lemoine-Sebastian 1972; Fillman 1978). Although one pattern observed by Fillman occurs only in Callitris and Actinostrobus, in most cases no pattern or combination of patterns is characteristic of a particular genus or group of genera. Indeed, there is an extreme irregularity in the patterns among leaves at the same node as well as at different nodes, and between trees and populations (Fillman 1978). This character does not appear to be useful in assessing affinities of the genera. Dimorphic leaves (i.e., adult leaves differentiated into lateral and facial forms) occur in 13 genera (Table 13). This character-state does not occur in any other coniferous family and so is clearly an apomorph in the Cupressaceae. It characterises all but Pilgerodendron in group 4, and occurs in 6 of the 10 taxa assigned to group 1 (only occurring in a single species of Cupressus). Fitzpatrick (1965), de Laubenfels (1965) and Oladele (1983a, b) all concluded that many genera cannot be distinguished on leaf type alone. This remarkable degree of similarity in vegetative morphology has led a number of authors to consider aspects of leaf anatomy, particularly of the epidermis. Both Florin (1931) and de Laubenfels ( 1965) examined the structure and distribution of stomata on the leaves of many conifers, while Oladele ( 1983a, b) has re-defined and expanded the range of cuticular and stomata! characters, and scored them for all genera of the Cupressaceae except Libocedrus s.s.. Oladele's data on the cuticular characters (1983a,b) were subjected to a numerical analysis by Alvin et al. (1981). This study supported the earlier conclusions Section 6.2, p.51 of Florin (1931) and de Laubenfels (1965), that these characters, and particularly the structure of the stomata} apparatus, allowed the discrimination of natural taxonomic units of generic rank, but were generally insufficient to allow groupings of genera into subfamilies. Leaf histogensis and anatomy was extensively studied at the turn of the century (see Napp-Zinn 1966), and was re-examined for a number of cupressaceous taxa by Al-Sharifi (1952). The basic anatomy of the leaves is recorded as similar in almost all the taxa thus far investigated. The distribution of ramified thickenings associated with pits in the transfusion tracheids, however, appears to hold some promise as an aid to determining affinities of the genera. A highly developed form of this character consisting of trabecula-like protrusions from the pit border has been recorded in Cupressus and Juniperus, both of group IA, Platycladus from group ID, as well as in Actinostrobus from group 2 (Al-Sharifi 1952; Klemm 1886). My own observations have revealed the same character-state in the leaves of Tetraclinis, the third member of group lA. An intermediate form, in which the thickenings consist of peg-like protrusions, has been reported in Calocedrus decurrens and Widdringtonia, both genera assigned to group 1, while normal bordered pits without protrusions have been reported from Chamaecyparis, Thuja, Thujopsis and Libocedrus s.J., all of which are assigned to group 4, and Fitzroya, group IC (Klemm 1886; Al-Sharifi 1952). Although the character requires further study, it is most interesting that all three genera assigned to group lA possess this most unusual feature. Section 6.3, p.52 6.3 Cone morphology. The morphology of the mature female cone has received a great deal of attention in the past. An analysis of the relationship of the cone scales to each other, in terms of their imbricate or valvate nature, and their arrangement in the mature cone was used by Li (1953) as the basis for his subfamily and tribal divisions (see Section 1). The arrangement of the cone scales usually follows the phyllotaxis of the genus, and this in particular led Li to group Fitzroya with Callitris and Actinostrobus, all having ternate cone scales. de Laubenfels (1965) considered, however, that on the basis of other cone characters, such as the morphology of the columella and seeds, that Fitzroya had an affinity with Diselma rather than to any other genus or group of genera. This conclusion is certainly more in agreement with the biflavonoid data, on which Fitzroya, Diselma and Widdringtonia have been assigned to group 1B. Lemoine-Sebastian (1969, 1971) studied the vascularization of the mature ovulate cone in representatives of all genera except Diselma. Despite the authors' optimism in the introduction to the 1969 paper that the vascularisation would assist in defining the phyletic relationships between the genera ("C'est un caractere evolutif interessant qui peut aider a preciser les rapports phyletiques entre les genres."), the data presented do not bear this out. The conclusion that neither Li's two subfamilies nor his tribes possess fundamentally different types of vascularisation ("En resume on remarquera que lee deux sous-familles, et a plus forte raison lee diverses tribus, ne possedent pas des types de vascularisation fondamentalement differents, ••• "), and the fact that many or all types of Section 6.3, p.53 vascularisation identified by the author often occurred in the larger genera indicates that these types have arisen several times throughout the family, and do not, therefore, reflect affinities of genera at the tribal or subfamily level. Indeed, the author presents some evidence, within both Cupressus and Chamaecyparis, of a relationship between some vascularisation types and the size of the cone (1969:23). It may be that useful information could be gained from characters of the cone at pollination, such as ovule number and arrangement, orientation and morphology. Few data on these aspects of the cone are available (e.g., Li's (1972) study of the ovulate cones of the Formosan members of Chamaecyparis). 6.4 Reproductive Biology. Although reproductive biology has been extensively studied in conifers in general, only a few Cupressaceae have been studied in full (Konar and Oberoi 1969; Singh 1978). The work of Owens and Molder (1974, 1975, 1980) on Chamaecyparis nootkatensis and Thuja plicata, giving accounts from cone initiation to seed maturity, are notable exceptions. Both Sterling (1963) and Singh (1978) comment on the paucity of information for any part of the life cycles of many genera in the family; this includes Neocallitropsis, Diselma, Fokienia, Microbiota, Pilgerodendron, Fitzroya, Widdringtonia, Thujopsis, and Tetraclinis. Although much has been made by various authors (e.g., Saxton 1913a, b; Moseley 1943) of the taxonomic significance of various stages of the reproductive cycle of members of the family, it is clear that the limited information precludes a comprehensive assessment. Section 6.4, p.54 Saxton (1913a) and Moseley (1943) referred to the following character-states being shared by Callitris, Actinostrobus and Widdringtonia: archegonial complex lateral in position, absence of a prosuspensor in embryogeny, a non-tiered proembryo that fills the archegonium, lack of an archegonial jacket, ovulate cone scales all fertile, and male nucleus equal in size to the female nucleus. Although these states appear to characterise Callitris and Actinostrobus (Saxton 1913a; Baird 1953), not all are applicable to Widdringtonia and several require comment. Lateral archegonial complexes are known in Tetraclinis, Fitzroya and Juniperus, although they are quite rare (Saxton 1913b, Moseley 1943). Singh and Oberoi (1962) report an archegonial complex as sometimes forming at the chalazal end of the gametophyte in Platycladus. Lateral complexes have also been reported in a group of taxodiaceous genera, Sequoia, Sequoidendron, Arthrotaxis (Buchholz 1946:33). While the occurrence of lateral complexes certainly seems likely to be a derived state, it is obvious they have arisen more than once in the conifers (see Boyle and Doyle 1953). Hence, the taxonomic significance of this character-state is uncertain. The critical stages of proembryo development in Widdringtonia are unknown, and the absence of a prosuspensor and presence of a non-tiered proembryo was postulated on the basis of later stages by Moseley (1943:121). Moseley also makes reference to " •• a slightly differentiated jacket •• around a whole basal group of archegonia" (p.119), but later tabulated the genus as the jacket being absent or obscured. Clearly, the character-states given by Moseley require re-examination. Saxton (1910a, b), for example, clearly indicated that Widdringtonia was quite distinct Section 6.4, p.55 from Callitris on many reproductive characters, such as number of megaspore mother-cells, morphology of prothallus, and numbers of archegonia in a complex. Saxton (1913b) compared aspects of the reproductive biology of Tetraclinis with those of Cupressus, Juniperus, Widdringtonia, Callitris and Actinostrobus. He considered the affinities of Tetraclinis, on the basis of the development of both gametophytes, the structure of the female cone and ovules, and the ovule arrangement with its single functional megaspore cell surrounded by tapetal tissue, to be with Juniperus and Cupressus, rather than Callitris and Actinostrobus (p.599). He placed Widdringtonia intermediate between the two groups (p.601: "Thus Widdringtonia is the most primitive of the Callitroideae, and differs least from the Cupressoideae, while Callitris and Actinostrobus are more specialised."). These conclusions are certainly more in agreement with the biflavonoid data, which places Tetraclinis with Juniperus and Cupressus in group lA, and Callitris and Actinostrobus in group 2. 6.5 Wood Histology Many species of Cupressaceae have an economic importance in forestry, and consequently characters from wood histology have been scored in representatives of a large number of genera. Early surveys by Peirce (1937), Phillips (1941) and Greguss (1955) have been complemented by studies on the north American members of the family (Bannan 1941, 1942, 1944, 1952, 1954), on the members of Libocedrus s.l. (Boutleje 1955; Patel 1968), and on Callitris (Venning 1979). Section 6.5, p.56 Whilst disjunctions in the distribution of wood character-states in three genera have previously been noted (Section 5.2), several genera show marked delimitation from the family on the presence of some wood characters. Thuja s.1. and Thujopsis are distinctive in the absence of a torus from the pit membrane of intertracheary pits (Bauch et al. 1972). The presence of callitroid thickenings across the pit border of intertracheary pits has been reported in 6 species of Oallitris, although it is absent in another 6 (Venning 1979). This type of wall thickening has also been reported in Actinostrobus, and occurs occasionally in a few species of Juniperus (Phillips 1941) and Dacrydium (Meylan and Butterfield 1980) although it is not typical of the species. Cupressaceous taxa possess both axial (vertical) and radial (ray) parenchyma, as do the woods of Taxodiaceae and Podocarpaceae but not Araucariaceae and Taxaceae (Peirce 1937; Phillips 1941). Using a combination of states from the parenchyma cells, Greguss (1955) considered that the cupresaceous genera could be divided into "three very distinct groups". The first contained genera that had smooth, thin ray cell walls and horizontal walls of axial parenchyma, and consisted of Actinostrobus, Oallitris, Neocallitropsis, Tetraclinis and Widdringtonia. The second group was reported to possess somewhat thicker horizontal ray cell walls which were smooth or sparsely pitted, while the tangential walls were still smooth and thin; it consisted of Platycladus, Ohamaecyparis s.1., Fokienia, Microbiota, Thuja occidentalis and Thujopsis. The final group had thick and pitted horizontal walls of both axial and ray parenchyma, the tangential walls of ray parenchyma had prominent bead-like thickenings, and indentures Section 6.5, p.57 (hollows at the corners of ray cells where the horizontal and tangential walls meet) were usually distinct. This group contained Oupressus, Diselma, Fitzroya, Juniperus, Oalocedrus decurrens and Pilgerodendron. Both Phillips (1941) and Greguss (1955) record the character-states in the Taxodiaceae to be: tangential walls of the ray parenchyma thin and smooth, while the horizontal walls are smooth to pitted; horizontal walls of the axial parenchyma smooth in 7 of the 10 genera and indentures absent in all 10. The states of thick horizontal walls with abundant pitting and distinctly nodular tangential walls of ray parenchyma, pitted or thickened horizontal walls of axial parenchyma, and distinct indentures are considered here to represent derived conditions in the Cupressaceae. Greguss's third group contains those genera with these derived conditions. The grouping of Cupressus, Juniperus, Diselma, Fitzroya and Oalocedrus decurrens ties in well with the biflavonoid data, all genera belonging to group 1, but the position of Pilgerodendron is anomolous. However, Phillips (1941) and Boutelje (1955) disagree with the states recorded for Pilgerodendron, recording instead that this genus lacks both indentures and the beadlike thickenings of the tangential walls of the ray cell, and has axial parenchyma with smooth horizontal walls (see Table 13). Many genera have not been closely re-examined since Greguss (1955), and those that have show that previous reports on the nature of the axial and ray parenchyma must be treated with caution. The New Zealand species of Libocedrus, which were originally recorded as having smooth horizontal walls of axial parenchyma and thin unpitted horizontal walls of ray parenchyma Section 6.5, p.58 (Phillips 1941; Greguss 1955), were found by Boutelje (1955) and Patel ( 1968) to have fine nodular thickenings on the horizontal walls of the axial parenchyma and, in L. bidwillii in particular, to have thick and abundantly pitted horizontal walls of ray parenchyma (Table 13). Again, although Cupressus and Juniperus are similar in many features of their wood structure, both Phillips (1941) and Peirce (1937) recorded the tangential walls of the ray parenchyma in Cupressus to be smooth, and in Juniperus distinctly nodular. A re-examination of seven American species of Cupressus by Bannan (1954) found that only three possessed "more or less smooth" walls, three displayed "a distinctly nodular condition", while in the 7th, C. macnabiana Murr., they were smooth in one collection, sporadically nodular in a second, and distinctly nodular in two others. More recent studies under SEM of the ray parenchyma reveal these nodules to be a product of the size and distribution of pit areas and the degree of thickening of the walls (Quinn, pers. comm.). Clearly, a re-definition of the characters of axial and ray parenchyma is needed on the basis of SEM studies of all species. SEM studies have revealed the presence of a vestured or warty layer in the tracheids of various conifers, as well as in a wide range of angiosperms (Liese 1965; Ohtani et al. 1983, 1984). There is marked variation in the size, density and distribution of the warts. In Callitris columellaris they are visible under the x40 objective of the light microscope, while in Cupressus sempervirens they are resolvable only at x3000 magnification under SEM; they may be confined to the pit chamber and/or canal, extend out from the pits to varying degrees or cover the entire inside surface of the cell (Meylan and Butterfield 1980). Insufficient Section 6.5, p.59 species have been surveyed within the family to determine whether these differences characterise individual species or coincide with generic boundaries. 6.6 Chemistry. The chemical data surveyed here has much the same limitations as the previous biflavonoid data (2.2). A general compilation and overview of chemical data is given by Erdtman and Norin (1966), and indicates that few chemical groups have been studied in representatives of all or even most genera of the family. 6.6.1 Tropolones. These unsaturated, non-benzenoid aromatics are confined to heartwoods. They have received attention because of their toxicity to a wide variety of wood-destroying fungi (Gardner 1962), and consequently most genera of the family have been surveyed. The presence of tropolones appears to be a derived character in the family, as they are not reported in any other coniferous family (Erdtman and Norin 1966). They are present in two forms in the Cupressaceae, the thujaplicin and nootkatin types, each the product of a separate biosynthetic pathway (Erdtman and Norin 1966). Their distribution is shown in Table 13. Only two genera, Cupressus and Juniperus, contain tropolones of both types. In the remainder, nootkatin occurs only in Chamaecyparis nootkatensis, while thujaplicin types occur in a number of genera from groups 1 and 4. These data, however, may not be accurate, as Gardner (1962:324) indicates that the presence of tropolones varies greatly both within and between trees. For example, the thujaplicin content in heartwood samples of Thuja Section 6.6, p.60 plicata varies from 0% to 1.2%. It is apparent, therefore, that extensive sampling is needed in order to determine the nature of the variation in this character and especially to check absences recorded in the literature. 6.6.2 Terpenes. Terpenes are the main constituents of the extractives of the wood of most Cupressaceae, and are a large component of the volatile-oil fraction of the foliage (Erdtman and Norin 1966; von Rudloff 1975). a) Wood terpenes. These data are reviewed by Erdtman and Norin (1966). They report that the cis-farnesyl sesquiterpenes (terpenes composed of 3 isoprene units) derived from Cupressus, Juniperus, Widdringtonia and Tetraclinis, all group 1 genera, are different from those reported from Callitris (group 2) and Neocallitropsis (group 3). Gough (unpublished, see Table 13) surveyed the wood resin of many species in the family for diterpenoid compounds. The results show that many genera contain unique components. Callitris alone contains callitrisic acid, although Actinostrobus, the other member of group 2, was not examined. Calocedrus decurrens contained dehydroabietic acid, which was not reported from any other taxon, as well as lambertianic acid, which was also found in Platycladus. Austrocedrus contains an unidentified diterpenoid acid, while Widdringtonia contains a number of unidentified compounds not found elsewhere in the family. The group lA genera, Cupressus, Juniperus and Tetraclinis, are the only ones to contain Section 6.6.2, p.61 agathic acid. Hence, there is some support. in these data for the generic groups proposed on biflavonoids. b) Foliage terpenes. The north American representatives of the family are particularly well represented in published reports of foliage terpenes (von Rudloff 1975; Carman and Sutherland 1979; Adams et al. 1980, 1981, 1983). Gough and Welch (1978) refer to an unpublished study that includes both northern and southern genera, which purports to show that the diterpenoid phenols sempervirol and totarol are only found in northern members of the family. However, several published reports have shown that there is variability of essential oil components within genera. von Rudloff (1975) reported the components of Juniperus occidentalis were different to those found in J. scopulorum Sarg., J. horizontalis and J. virginiana, while those found in J. communis were very different again. Carman and Sutherland (1979) found variability within foliage samples of Cupressus macrocarpa. Thus there appears to be little promise of essential oil components assisting in recognising the affinities of genera. von Rudloff (1975:167) concluded that although some distinctive terpenes appear to exist at the genus level in the family, 11 •• conifer leaf oil analysis is best applied in chemosystematic studies at the species and subspecies level. 11 • 6.6.3 Flavonoids Reports of flavonoids in gymnosperms are sporadic, but show that they are found in many families (Erdtman and Norin 1966; Harborne et al. 1975; Harborne and Mabry 1982). Lebreton et al. (1967, 1978) and Lebreton (1982) have attempted to apply a Section 6.6.3, p.62 technique for quantitatively determining the total flavonoid content of leaves to the delimitation of suprageneric taxa in the Cupressaceae. A re-classification of the family into 3 subfamilies based upon the results from only seven genera was proposed (Lebreton 1982). His conclusions are extraordinary, however, in that they are not supported by his own data. He draws attention to a chemical similarity between Cupressus and Juniperus and a marked difference between Cupressus and Chamaecyparis, yet retains the tribal arrangement that groups the latter pair and separates the former. Tetraclinis shows no compounds in common with Callitris, yet again both are placed in his subfamily Callitroideae. Hence his data are in better agreement with the biflavonoid groupings proposed in 5.3 than with his own classification. 6. 7 Pollen Morphology. A survey of the pollen morphology of representatives of most genera of the Cupresaceae was undertaken by van Campo (1953). Erdtman (1965) reviews the literature and provides a generic description of all genera based on reports of analysis by light microscopy, although the descriptions of some genera (e.g., Tetraclinis, Fitzroya) are incomplete. There are few reports of observations with electron microscopy (Ueno 1959, 1974; Yamazaki and Takaoka 1962; Venning 1979; Owens and Molder 1980; Pocknall 1981c). The family is revealed to be markedly stenopalynous (van Campo 1953; Erdtman 1965). Observed differences in pollen size, which were used by van Campo (1953) and Gaussen (1968) as a taxonomic character, may be influenced by method of preparation, Section 6. 7, p.63 sample size, mounting medium and maturity of the pollen sampled. Faegri and Iverson (1975) emphasise the caution which must be exercised when comparing data on different pollen samples. Pollen size for different genera in the family often overlap; this is especially true of the largest genus, Juniperus, which includes species at both ends of the range for the family (van Campo, 1953; Erdtman 1965). The sculpture and structure of the exine of pollen has been used in many taxonomic appraisals of angiosperm groups. Although the gymnosperm exine deviates from the typical angiosperm exine in two important respects, namely the nature of the infratectal layer and the presence of a multilayered or striate nexine (Gullvag 1966; van Campo and Lugardon 1973; Walker 1976), SEM studies of sculpture and structure have proved useful in taxonomic studies of Dacrydium s.l. and Phyllocladus (Pocknall 1981a, b). Roscher (1974) has recently suggested that criteria do exist for classifying non-saccate gymnosperm pollen using characters recorded from electron microscopy of the wall and orbicle structure, and of the pseudopore. Future studies of this type may contribute to the elucidation of affinities in the Cupressaceae. Section 7, p.64 7. CONCLUSIONS. The chromatographic techniques utilised here, particularly those involving permethylation of the raw extract, are simple, reliable and highly sensitive. The biflavonoid patterns obtained in this way are clearly producing useful information on affinities at the generic level within the Cupressaceae s.s. Marked discontinuities involving both the biflavonoid series and their derivatives, although generally coinciding well with existing generic boundaries, do not align with the tribal or subfamily boundaries proposed by Li (1953). The distribution of the cupressuflavone series in particular cuts straight across tribal and subfamily boundaries, occuring in both subfamilies and 5 of the 6 tribes. The presence of the cupressuflavone and taiwaniaflavone series of biflavones, and the absence of the hinokiflavone series appear to be apomorphs in the family. Synapomorphs allow the recognition of 3 suprageneric groups; the remaining genera constitute a rather heterogeneous assemblage sharing no derived characters in their biflavonoid patterns. The distribution of derivatives allows the recognition of 4 subgroups within the group 1 genera. There are few other data sources which allow an assessment of tribal and subfamily groupings, as most have only been sporadically investigated in the family. Even in those characters that have been widely surveyed, there is certainly no support shown for the subfamilies and tribes of Li (19{53). However, the distribution of states of some of the characters discussed above do support the existence of affinities between taxa that have been proposed on the basis of biflavonoid patterns. This is particularly true of the group 1 genera. Cupressus and Juniperus display common derived states in Section 7, p.65 their chemistry (nootkatin-type tropolones, identical diterpenoids including agathic acid), wood morphology (thick horizontal walls with abundant pitting and distinctly nodular tangential walls of ray parenchyma, thickened horizontal walls of axial parenchyma, and distinct indentures), and leaf anatomy (strongly developed trabeculae in the transfusion tracheids). This suggests that these two genera constitute a robust taxonomic group. Agathic acid and trabeculate transfusion tissue also occur in Tetraclinis, while all 5 wood characters are found in Calocedrus decurrens. All but one of this set of wood characters are found in Diselma, Fitzroya, Platycladus and Chamaecyparis nootkatensis, and all but 2 in Thuja occidentalis. Nootkatin-type tropolones also occur in Chamaecyparis nootkatensis. Hence, there is already considerable evidence to support the conclusion that the presence of the cupressuflavone series in the Cupressaceae is a monophyletic character-state. The group 2 genera are united on their ternate phyllotaxis and several reproductive specializations (absence of prosuspensors, lateral archegonial complexes, non-tiered proembryo, lack of archegonial jacket, and equal sized male and female gamete nuclei); hence there is some support for the view that the absence of the hinokiflavone series is a specialisation within the family. Clearly, then, while it is apparent that many characters are inadequately surveyed across the family, the distribution of some character-states shows a correlation with the distribution of biflavonoid characters. As yet there are too man gaps in the available data to enable a more satisfactory suprageneric taxonomy to be constructed. Additional data sources may be required to help define relationships within the group 4 genera especially, since these appear to retain a relatively unspecialised complement of biflavonoids. Even within this group, however, similarities in biflavonoid patterns of Section 7, p.66 Libocedrus, Austrocedrus, Papuacedrus and Pilgerodendron suggest a very close affinity between them. Hence, the biflavonoid pattern has proved to be a valuable data source for recognising affinities above the generic level within the Cupressaceae s.s. While biflavonoid patterns show a high degree of uniformity within the majority of genera, there are marked discontinuities in three of them which have been shown to be correlated with discontinuities in a range of other characters (5.2). It is concluded, therefore, that each of these genera is polyphyletic, and that a redefinition of generic boundaries is required. HEYDERIA K. Koch Heyderia K. Koch, Dendrol. 2(2):178 (1873). Type species: Heyderia decurrens (Torr.) K. Koch. Libocedrus decurrens Torrey, Smithen. Inst. Contrib. Know1. 5(1):7 (1853). Oalocedrus decurrens (Torr.) Florin, Taxon 5:192 (1956). Monotypic. Evergreen trees 18-46m, resinous and aromatic. Bark 20-25mm thick, deeply and irregularly furrowed into shreddy ridges. Branchlets flattened, distichously divided, terminating in dense fan-like sprays. Leaves scale-like, decussate, strongly flattened, opposite in four rows, decurrent, 3-12mm long, closely appressed except for a short pointed tip; dimorphic, differentiated into facial and lateral pairs, lateral pairs obtusely keeled, two stomata! bands on the underside equally developed, the leaves overlapping the facial leaves; facials leaves alike, slightly convex, each with a shallow median furrow. Cones. solitary, terminal; male and female cones usually on different branches of the same tree; male cones oblong-elliptical,. 6mm long, with 6-16 sporophylls; female cones woody, ripening in the first year, pendulous throughout development, ovate, 2-2.5cm Section 7, p.67 long, composed of 6 paired scales; at maturity the lower scales ovate-triangular, reflexed, about 6mm long; the middle pair the length and width of the cone and with a small triangular reflexed process near the apex; the upper pair erect, fused together, with three minute processes at the apex; middle pair only fertile. Seeds usually 2 per fertile scale; each seed with 2 unequal wings, the longest about 2cm long. Habitat mountain soils in mixed coniferous forests, at 366-2134m. Range From western Oregon south to southern California and to extreme western Nevada, USA. NOTES. This genus was resurrected by Li (1953) to incorporate the three northern hemisphere species formerly contained in Libocedrus a.I. Nomenclatural priority dictated that the valid name for the genus sensu Li was Calocedrus (Florin 1956). As redefined here, the genera reflect the original definitions, Heyderia having been based on the north American Libocedrus decurrens and Calocedrus on the Asian C. macrolepis (Kurz in J. Bot. 11:196 1873). Heyderia differs from Calocedrus in having its ultimate branchlets stouter, less strongly dorsiventral; the facial leaves not dorsiventrally differentiated; the abaxial stomata} bands of the lateral leaves not equally developed; wood ray parenchyma with conspicuous indentures and nodular thickenings on the tangential walls; distinct but small sized crossfield pitting; female cones pendulous; mature cones larger (2-2.5cm vs. 0.6-1.5cm); thick and furrowed bark; and the presence of biflavones of the cupressuflavone and taiwaniaflavone series in the leaves. CALLITROPSIS Oersted Callitropsis Oersted, Vidensk. Meddel. Dansk Naturhist. Foren. Kjobenhavn 1864:32 (1864). Type species: Callitropsis nootkatensis (D. Don) Oersted. Cupressus nootkatensis D. Don in Lamb., Descr. Genus Pinus 2:18 (1824). Section 7, p.68 Chamaecyparis nootkatensis (D. Don) Spach, Hist. Nat. Veg. Phaener. 11:333 (1842). Monotypic. Evergreen tree to 30m and 1.2-(6)m in girth. Bark thin, fibrous and shreddy, with long narrow fissures. Branches smooth, horizontal or slightly drooping. Branchlets flattened, pendulous, fern-like horizontal sprays with alternate pinnae arranged in 2 ranks. Leaves dull to bright yellow-green, closely appressed in opposite pairs, equal in length, 3mm long, pointed at the apex, usually without gland dot or glandular ridge and devoid of abaxial stomata! lines; laterals keeled, facials flattened. Cones terminal, on same tree; male cones yellow, pollen usually one-celled; female cones ripening in 2 seasons, at maturity small, globose, 12mm in diameter, 4-6 scales, each with a triangular sharp-pointed process. Seeds 2-4 per scale, flattened, winged, without resin tubercles. Habitat wet mountain soils, at sea level in the north to 2134m in the south. Range Pacific coast of North America from south and south east Alaska to the mountains of western Oregon and extreme north western California, USA. NOTES. This genus is distinct from Chamaecyparis, in which the type was previously placed, on the following characters: branches and branchlets more pendulous; leaves rarely glandular, and without white stomata! lines; pollen usually one-celled at pollination; female cones ripening in two seasons; cone scales 4-6 per cone (cf. 6-11); wood with short rays composed entirely of ray tracheids; tangential walls of ray parenchyma nodular; torus of intertracheary pits disc-shaped; presence of nootkatin-type tropolones in wood extractives; unique terpenoid and non-terpenoid foliage oil components; and the presence of the cupressuflavone series of biflavonoids in the leaves. The existence of hybrids, albeit cultivated, between Callitropsis nootkatensis and a number of Cupressus species suggests a close affinity with the genus Cupressus, although the genera are morphologically and chemically very distinct. Section 7, p.69 THUJA L, emend Gadek. Thuja L., Sp. Pl. 1002 (1753), Lectotype: Thuja occidentalis L. Monotypic. Evergreen tree to 21m, trunk often branched and prominently buttressed. Bark reddish-brown, fibrous, fissured into narrow ridges, the thin outer bark scaling off in small rolls. Branches horizontal, turning upwards at the ends. Branchlets much divided, flattened, in horizontal sprays, Leaves yellow-green above, pale blue-green beneath, overlapping, flattened, 1.5-3mm long; lateral leaves laterally compressed and strongly keeled, facial leaves dorsiventrally flattened, with a conspicuous resin gland which is smaller or absent from the facial leaves of lateral shoots. Monoecious; female Cones elliptical to oblong, 8-l0mm long, erect from a short upwardly curved stalk at pollination, sometimes pendulous when mature, composed of 4-5 pairs of thin, leathery, blunt-pointed cone scales, the second and third pair larger and fertile. Seeds 2 per fertile scale, flattened and small, each with a narrow or emarginate wing. Habitat typically in swamps. Range North America from south east Manitoba to Nova Scotia and Maine, south to New York and locally to North Carolina, and west to Illinois. NOTES. This genus as emended here is monotypic, retaining only the lectotype, T. occidentalis. The remaining 4 species are placed in the new genus Neothuja (v.i.). Thuja differs from Neothuja in having conspicuously glandular leaves, yellowish to blue-green underneath without the white stomata! lines evident in the lower facial leaves of the members of the latter genus; emarginate or slightly winged seeds (cf. markedly winged seeds); nodular tangential walls of ray parenchyma (cf. smooth) and unthickened transverse walls of axial parenchyma (cf. thickened) in the wood; and the presence of the cupressuflavone series of biflavones in foliage extracts. Section 7, p. 70 NEOTHUJA P. A. Gadek, gen. nov. Type species: Neothuja standishii (Gord.) Gadek Thujopsis standishii Gord., The Pinetum Suppl.: 100 (1862). Thuja standishii (Gord.) Carr., Traite general des Conifers, ed. 2:108 (1867). Evergreen, monoecious shrubs to large trees. Bark thin, fibrous or papery. Branches horizontal. Branchlets flattened, in horizontal sprays, often drooping. Leaves small, appressed, eglandular or with an obscure gland or furrow, and bearing white stomata! lines on the underside of the branchlet. Female cones small, conical, 0.6 to 1.3 cm long; cone scales 4-6 pairs, thin and tapering with the points turned outward when mature, middle pairs fertile. Seeds 2-3 per fertile scale, flattened, each with a marked but thin membraneous wing on either side extending and joined beyond apex of the seed. Range from China, Korea and Japan to western North America. New combinations: Neothuja koraiensis (Nakai) Gadek. N. plicata (D. Don) Gadek. N. sutchuensis (Franch,) Gadek. NOTES. Two subgroups may be recognised, one containing N. plicata from western North America, and the other containing the three Asian species. The American species has distinctive aromatic foliage when crushed, a feature not found in the other members, Section 8, p. 71 8. LOCALISATION OF BIFLAVONOIDS IN LEAVES. 8.1 Introduction The cellular and subcellular localisation of biflavonoids has not previously been described, although they are known to be a characteristic component of the leaves of many gymnosperms, and have been recorded from such diverse parts of vascular plants as the bark, heartwood, roots, stamens, fruits and testa (Geiger and Quinn 1975; Gadek 1982). Information on the localisation of biflavonoids in gymnosperm leaves should contribute to an understanding of the functional significance of their accumulation in these organs, which at present remains the subject of pure speculation. It is not surprising that the subcellular localisation of biflavonoids has not previously been described, since most preparative techniques employed in histochemistry or microscopy involve soaking in aqueous or alcoholic solutions that will leach many phenolics from plant tissue. Recently, techniques employing fluorescence microscopy have successively been used to localise bound phenolic acids in plant cell walls (Fulcher et al. 1972; Harris and Hartley 1976, 1980; Hartley and Harris 1981). Flavonoids have been localised in the UV absorbing regions of petals by induced fluorescence using the intense colour change of certain flavonoids produced by alkaline conditions (Brehm and Krell 1975). This latter method involved misting thin sections prepared by freeze microtomy with a 1% aqueous aluminium chloride solution and then observing the induced yellow fluorescence of the flavonoids under UV microscopy. Most biflavonoids occuring naturally possess a free 5 hydroxyl group and so will also form an acid-stable complex with aluminium chloride which has a strong deep yellow Section 8.1, p.72 fluorescence (see 3.3.2.1). This section reports on the application of a development of these techniques to the localisation of the biflavonoid fraction in the leaves of a wide range of gymnosperms in which they are known to be accumulated. Leaves of Agathis robusta were particularly suitable material in which to test a technique for the localisation of this fraction. Previous studies on the leaves of this and several other species of the genus Agathis have revealed that biflavonoids constitute by far the major proportion of the flavonoid fraction (Khan et al. 1972; Varshney et al. 1973). Furthermore, unlike the scale leaves of many conifers, the broad dorsiventral leaves of Agathis species are relatively easily separable into epidermal and m'esophyll fractions, allowing confirmation of the observations by TLC analysis of extracts from these fractions. For these reasons, the study concentrates on A. robusta, but includes a survey of aluminium chloride induced fluorescence in representatives of all orders of gymnosperms, as well as the Psilotales. 8.2 Materials and Methods. Unless otherwise stated, all studies were carried out on mature leaves of specimens cultivated in the grounds of the University of New South Wales; voucher specimens have been lodged with UNSW and are listed in the appendix. 8.2.1 Leaf anatomy. General leaf anatomy was studied in sections of paraffin wax-embedded leaves. The sections were stained with safranin/fast green, Sudan black B in 70% ethanol (Pearse 1968), or 0.5% toluidine blue O in acetate buffer pH 4.4 (Feder and O'Brien 1968). In addition, dewaxed unstained sections mounted in 95% ethanol were examined between Section 8.2, p.73 crossed polarising filters. Similar unstained sections were also examined by fluorescence microscopy under the conditions cited below for fresh sections. Fresh leaf sections (see below) were stained with the periodic acid - Schiff's reaction (PAS), using a 30 minute alderhyde blockade in 2,4-dinitrophenyl hydrazine in 15% acetic acid (O'Brien and McCully 1981). In controls, the periodic acid oxidation step was omitted. Thin sections were cut from leaf pieces fixed for 24 hours in 3% glutaralderhyde on ice (0.025M potassium phosphate buffer, pH 6.8), dehydrated through a methyl cellosolve series on ice and embedded in glycol methacrylate (GMA) (O'Brien and Mccully 1981). Blocks were polymerised under UV light in a nitrogen atmosphere at 25°C. Sections were cut at 2um using glass knives on a Reichert ultramicrotome and stained with Sudan black B, toluidine blue O or PAS reagents (as above), or amido black 10B (Fisher 1968). As well, GMA sections were examined between crossed polarising filters, and by fluorescence microscopy under the conditions cited below. 8.2.2 Fluorescence microscopy. Transverse sections of fresh material were either hand cut or prepared on a Reichert sledge microtome from leaves mounted in carrot tissue, transferred by a barely moist paint brush directly to a dry slide and either mounted in Leitz fluorescence-free immersion oil or in 5% aluminium chloride in 95% ethanol (AlCh). The sections were observed under a Leitz Orthoplan microscope equipped with a Ploempak epifluorescence illuminator. A broad band UV filter block was used, consisting of exciter filter, BP 340-380, beam splitter, RKP 400, and barrier filter, LP 430. AlCh forms an acid-stable complex with flavones and flavonols that possess hydroxyl groups at C-3 or C-5 Section 8.2.2, p. 7 4 positions, and induces a strong bathochromic shift in band I absorbance (Mabry et al. 1970). Biflavonoids from the amentoflavone, cupressuflavone, agathisflavone, robustaflavone and hinokiflavone series have been recorded from the leaves of members of the Araucariaceae (Geiger and Quinn 1982). In the presence of AlCh, all these biflavonoids are characterised by a band I absorbance with a maximum in the range of the exciter filter (in fact 345-355 nm) and show strong yellow to orange fluorescence. Using the BP 340-380 exciter filter, a sample of amentoflavone showed no yellow fluorescence in the absence of AlCh but showed intense deep yellow fluorescence when AlCh was present. Although AlCh will also chelate with a range of other phenolics, areas in which biflavonoids are accumulated will be characterised by an intense yellow induced fluorescence under the above system of illumination. That this AlCh induced fluorescence is due to the presence of biflavonoids must, however, be confirmed by extraction of the particular tissue fraction and identification of the compounds present. 8.2.3 Extraction of biflavonoids. Ethanolic extracts of whole leaves, adaxial epidermal peels, middle-leaf tissue (i.e., leaf with both epidermises removed) and waxy scrappings from the adaxial surface were prepared by soaking in 70% ethanol for 24 hours. Eluants were dried, taken up again in a small volume of 70% ethanol and subjected to two-dimensional chromatography on paper using tertiary butyl alcohol:acetic acid:water (3:1:1, TBA) followed by 15% acetic acid, and one-dimensional chromatography on aluminium-backed precoated silica gel plates using BPF. Papers were viewed under UV before and after spraying with AlCh. The raw extract of each leaf tissue fraction as well as ethanolic extracts of the major FIGURE LEGENDS. 8a) Leaf margin of Agathis robusta as seen in transverse section. c, small mesophyll cells adjacent to vascular bundle (see text); e, epidermis plus cuticle; f, thick-walled non-lignified cells (? fibres); h, hypodermis; p, parenchmya; ph, phloem; pm, palisade mesophyll; r, resin canal; sm, spongy mesophyll; t, transfusion tracheids; x, xylem. Scale: 250 um. 8b) Diagram illustrating the epidermal and hypodermal wall system of the adaxial surface of the leaf of Agathis robusta, as seen in transverse section. e, epidermal cell; h, thin-walled hypodermal cell; hf, hypodermal fibre. The histochemically recognizable regions are: 1, epicuticular wax; 2, lightly stained in Sudan black B but unstained with all other stains - interpreted as the cuticle proper; 3, very lightly stained as for lignified walls but is intensely black with Sudan black B - interpreted as cutinised cell wall; 4, reacts with various stains which normally stain unlignified cell walls but is unstained by Sudan black B - interpreted as uncutinised cell wall. Scale: 20 um, 8a Section 8.2.3, p. 75 bands from unsprayed plates were permethylated (3.3.2.2). 8.3 Results. 8.3.1 Leaf anatomy of Agathis robust& The leaves of Agathis robusta are broad, flat and elliptical-lanceolate, with downcurved margins (Fig. 8a), and have a hard leathery texture, The lamina is devoid of a midrib, being traversed by many longitudinal veins that diverge from the base. The leaf has a dorsiventral anatomy with a single layer of large palisade mesophyll cells, although many of these are irregularly subdivided by one or more transverse walls. The vascular bundles are embedded in the upper part of the spongy mesophyll and alternate with one, sometimes two, resin canals (Fig, 8a), Large heavily lignified astrosclereids with radiating arms, similar to those described by Kausik (1976) for Agathis dammara (Lamb.) Rich., occur throughout the spongy mesophyll, their arms sometimes projecting into the palisade mesophyll. An arc of cells (fibres?) with very thick non-lignified walls lies both above and below the vascular bundle, and these cells retain a prominent protoplast even in leaves in their second year on the tree, Transfusion tissue similar to that described in A. dammara (Kausik 1976) can be discerned lateral to the xylem and phloem. The adaxial epidermis contains a small number of stomata (Hyland 1977) and is separated from the mesophyll by a prominent hypodermis of 1-2 layers of non-lignified fibres, in which the lumen is almost occluded, and thinner-walled living cells (Fig, 8b). Fibres predominate over thin-walled hypodermal cells by 3:2 in these layers and show a strong tendency to be clustered, The abaxial epidermis contains numerous stomata and its associated hypodermis is much less regular, with Figures 9-14. Light microscopy of adaxial region of Agathis robusta leaves in transverse section. Scale: 20 um. Figure 9. GMA-embedded section of a mature leaf under half-crossed polarizing filters showing crystals as white areas. Figure 10. Fresh section of a mature leaf stained with PAS reaction. Epicuticular deposits on the surface of the leaf are PAS-positive. Anticlinal flange and cuticle proper (region 2 in Fig. 8b) are PAS-negative, while the cutinised wall (region 3) stains light pink. Figure 11. Wax-embedded section of an immature leaf (see text) stained in safranin/fast green. The inner region of the outer periclinal . wall of the epidermis (region 4 in Fig. 8b) is stained dark green. The cuticle proper (region 2 in Fig. 8b) is devoid of crystals. Figure 12. Wax-embedded section of an immature leaf showing a stoma stained with safranin/fast green. The raised florin rings are clearly visible (arrows). Small crystals in the cuticle appear to be absent around the stoma. Figure 13. GMA-embedded section of a mature leaf stained with toluidine blue. The cuticle proper and cutinized wall (regions 2 and 3 respectively in Fig 8b) are unstained. Three pits can be seen in the anticlinal wall of an epidermal cell. Note that the section has separated between the cuticle proper and the epicuticular wax deposits (arrowed). Figure 14. GMA-embedded section of a mature leaf stained with Sudan black B. The innermost (uncutinised) layer of the outer periclinal and anticlinal walls of the epidermis (arrowed) are unstained (region 4 in Fig. 8b), while the cuticle proper and the anticlinal flange (region 2) are less intensely stained than the cutinised cell walls (region 3). 11 12 13 14 'I Section 8.3. 1, p. 76 thin-walled cells outnumbering non-lignified fibres by 2:1. The downcurved leaf margin contains a mass of hypodermal tissue, with 15-30 fibres (Carr and Carr 1977), many of which are lignified by the second year. As described by Stockey and Taylor (1981), the stomata are sunken to the hypodermal level, opening into a pit, the sides of which are formed by four or sometimes five subsiduary cells. A protuberance of the subsiduary cells projects upwards around the sides of this pit (Fig. 12) to form a pronounced 'florin ring' (Fig. 3 in Stockey and Taylor 1981). The stomatal pit is almost occluded by a wax plug (Fig. 3 in Stockey and Taylor 1981). Examination of the adaxial epidermis with a number of histochemical stains allowed the recognition of four distinct layers in the outer periclinal wall of the epidermal cells. Three of these extended into the anticlinal walls (Fig 8b). A narrow inner layer of the outer periclinal wall (region 4) was intensely staining, turning purple with toluidine blue (Fig. 13), green with safranin/fast green (Fig. 11), red with the PAS reaction (Fig. 10) and black with amido black; it was completely unstained, however, by Sudan black B (Fig. 14). This layer was continuous around each epidermal cell and it is interpreted as uncutinised cell wall. Abutting this layer was a broader relatively unstained region (region 3) which extended inwards along the full length of the anticlinal walls (Fig. 8b, 11, 13). This region was only faintly stained with the PAS reaction (Fig. 10), amido black and toluidine blue (Fig. 13), but was deeply stained with Sudan black B (Fig. 14). These characteristics of region 3 are interpreted as indicating a cutinised cell wall. In GMA sections the boundary between regions 3 and 4 appeared sharp but with minute convolutions at the interface. Section 8.3.1, p.77 Overlying the cutinised wall was a narrow layer of material that was completely unstained with PAS, toluidine blue or amido black, and was only lightly stained with Sudan black B (region 2). This layer extended into the central region of the anticlinal wall as an unstained flange that was quite distinct in sections stained with Sudan black B (Fig. 14). Region 2 was usually more pronounced in leaves in their first year on the tree. It appears to be the 'cuticle proper' of Von Mohl and Roelofsen (see Holloway 1982). The thin surface layer (region 1) stained with Sudan black B and was PAS-positive in fresh material but not in GMA sections. It seems to correspond to the epicuticular wax layer of Martin and Juniper (1970). The cutinised wall and layers exterior to it are collectivey referred to here as the cuticle. Fine crystals, presumably of calcium oxalate (see Cookson and Duigan 1951), were a prominent feature of the cutinised wall and all walls of the epidermal cells on both surfaces of the leaf (Figs 9 and 11). However, the cuticle proper appear to be devoid of crystals, and, in the immediate vicinity of the stomata, these crystals were seen to be less frequent and sometimes absent from the epidermis as a whole (Fig. 12). Similar crystals were also prominent in the walls of the astrosclereids, and were sparsely scattered throughout the walls of the mesophyll. 8.3.2 Fluorescence Microscopy. 8.3.2.1 Autofluorescence Fluorescence microscopy of fresh sections mounted in oil showed intense red-orange fluorescence from the chloroplasts, particularly in the palisade mesophyll, and strong blue-white fluorescence uniformly throughout the protoplasts of some smaller cells in the spongy Figures 15-20. Fluorescence microscopy of fresh, transections or epidermal peels from the adaxial surface of Agathis robusta leaves. Note that these black-and-white photographs of fluorescence tend to over-represent some colours (e.g., light blue). Scale: 20 um. Figure 15. Transverse section (TS) of mature leaf, mounted in oil. The cuticle, outer periclinal and anticlinal walls of the epidermis are non-fluorescent. Epicuticular wax, inner periclinal wall of the epidermis and the primary walls of the hypodermis fluoresce light blue. Figures 16-20. Material mounted in AlCb. Figure 16. TS of immature leaf. The anticlinal flange appears contiguous with a non-fluorescent surface layer (arrows), the cuticle proper (region 2 in Fig. 8b), Figure 17. TS of mature leaf. The non-fluorescent anticlinal flange (arrow) in the centre of each anticlinal wall of the epidermis is well shown. The fluorescence .in the cuticle appears granular because of the non-fluorescent crystals embedded in it. Figure 18. TS of mature leaf. Inner periclinal walls fluoresce blue in contrast to the intense deep yellow of the cuticle and anticlinal walls. The anticlinal flange (arrow) is non-fluorescent. Figure 19. Surface view of the epidermis of a mature leaf at the edge of a near paradermal section through the outer periclinal wall. The cuticular flange is non-fluorescent but is crossed by fluorescent pit areas (arrowed). The outer region of the cutinised cell wall (top of figure) fluoresces deep orange while the deeper layers fluoresce yellow. Figure 20. TS of stomata! chamber. The cuticle is continuous around the chamber, fluorescing deep yellow, while the guard cell protoplasts show an intense light blue fluorescence. Section 8.3.2, p. 78 mesophyll located immediately below the palisade layer and adjacent to the vascular bundles (Fig. 8a). In addition, there were abundant droplets of material which fluoresced light blue throughout the spongy mesophyll cells. The thickened walls of the guard cells, astrosclereids and xylem vessel elements showed strong light blue fluorescence, while the thin walls of the spongy mesophyll and the vascular tissue showed faint light blue fluorescence. The walls of the palisade mesophyll cells were non-fluorescent. In some sections, the contents of the resin canals showed strong light blue flourescence. The primary walls of both the non-lignified fibres and the thin-walled hypodermal cells showed a slightly deeper blue fluorescence, as did the inner periclinal wall of the epidermis (Figs 15, 29). Similarly, the epicuticular wax layer, when present on the sections, fluoresced light blue (Figs 15, 29). But both the anticlinal and the outer periclinal wall of the epidermis, including the cuticle, were non-fluorescent, 8.3.2.2 Aluminium chloride induced fluorescence. In sections mounted in AlCh, an intense deep yellow fluorescence appeared in the outer periclinal walls (Figs 8b, 16-18, 30). The anticlinal walls appeared three-layered in the transverse sections, the central region, which extended into the cuticle, showing no fluorescence (Figs 17, 18, 30); it was clearly continuous with a non-fluorescent surface layer (region 2) of the cuticle (Fig. 16) in sections of 1-year-old leaves, Region 2 was not clearly distinguishable from region 3 by its fluorescence in sections of older leaves, but the flange remained non-fluorescent. The non-fluorescent flange of the anticlinal walls could also be discerned in epidermal peels mounted in AlCh, where the fluorescent cell walls of adjacent epidermal cells were separated by a narrow non-fluorescent band (Fig. Section 8.3.2, p. 79 19). Small 'bridges' of fluorescence were visible connecting adjacent cells, no doubt corresponding to the pit areas on the anticlinal walls (Fig. 13, 19). The fluorescence of the cuticle often appeared particulate (Fig. 17). Examination of the same section mounted in AlCh, firstly between crossed polarising filters and then using fluorescence epi-illumination, revealed non-fluorescent crystals in the cuticle that were responsible for the overall particulate appearance of the fluorescence. The cuticle lining the stomatal chamber, although much thinner than elsewhere, still showed marked deep yellow fluorescence (Fig. 20); this fluorescent layer extended through the stomatal pore and faded out on the inner side of the guard cells. The raised walls of the subsiduary cells (florin rings) often contained large crystalline bodies of intense deep yellow fluorescence. The wax plugs of the stomata were non-fluorescent. A light yellow fluorescence was visible in the primary wall of the hypodermal fibres, especially in those parts close to the epidermis. This fluorescence was sharply defined and clearly distinguishable from the deep yellow of the cuticle and the outer periclinal and anticlinal walls of the epidermis. The protoplasts of the small spongy mesophyll cells adjacent to the palisade layer and vascular bundles referred to above fluoresced an intense light straw colour under these conditions. Epidermal peels often tended to 'stain up' from the inner surface, i.e., from the epidermal cells. This probably reflects the rate of penetration of the stain and its hindrance by epicuticular wax on the leaf surface, rather than a concentration gradient of flavonoids in the cuticle, as scratches across the outer surface of the cuticle became fluorescent immediately on staining. There was no marked leaching of the yellow fluorescence into the mounting medium as occurred with the TABLE 14. RESULTS OBTAINED FROM TWO-DIMENSIONAL PAPER CHROMATOGRAPHY OF EXTRACTS OF Agathis robusta LEAVES. Spot Rf in Rf in Colour under: Presence in extracts No. TBA HOAc UV +Nib +AlCh Whole Cuticular Epidermal Middle leaf scrapings peel leaf 1 S0-100 0-10 Dark Red DarkA + + red 2 90-100 0-20 Dark Dark Intense + + + yellow 3 40-70 60-70 Dark Dull Faint + ? yellow yellow 4 40-80 70-90 Light Blue- Light + + purple green purple 5 60-80 90 None Bright None + + + purple 6 20 90-100 White Bright White + + + white 7 20-80 90-100 Light Bright Light + + + blue blue blue A . spot 1 obscured by reaction of spot 2. TBA, tertiary butyl alcohol:acetic acid:water; HOAc, 15% acetic acid; + Presence; - Absence; ? trace. Section 8.3.2, p.80 substances showing light blue fluorescence in sections of other species (v.i.). De-waxed sections of wax-embedded material, when mounted in AlCla, did not show the intense deep yellow fluorescence observed in fresh sections; presumably the flavonoid component of the cuticle and epidermal cell walls had been leached out during fixation and dehydration. GMA-embedded thin sections did, however, show some residual fluorescence. 8.3.3 Chromatography of Extracts. The results of the two-dimensional separations of the extracts of the various layers of the leaf are given in Table 14. Seven spots were repeatedly identified in the whole leaf extract. Spot 1 represented the chlorophylls, and its absence from the extracts of the adaxial epidermal peels indicates that they were reasonably free of palisade mesophyll tissue. Spot 4 appears to be characteristic of the cuticular scrapings but, although it was identified in the whole leaf extract, it was not detected in the extract of the adaxial epidermal peel. Spot 2 was present in all but the middle leaf extract, and in each case was by far the most prominent spot. It had a high Rf in TBA, a low Rf in 15% acetic acid, and under UV was dark-absorbing both with and without ammonia fumes, but showed strong deep yellow fluorescence after being sprayed with AlCh. These characteristics indicate that this spot contained the biflavone fraction. Analysis of the whole leaf extract revealed a complex mixture of agathisflavone and amentoflavone and their partial methyl ethers, robustaflavone and partial methyl ethers of cupressuflavone. The following partial methyl ethers were identified: 7-monomethyl amentoflavone, 4'-monomethyl amentoflavone, 74'-dimethyl amentoflavone, TABLE 15. OCCURENCE OF BIFLAVONOID PERMETHYL ETHERS IN THE PERMETHYLATED RAW EXTRACTS OF Agathis robusta LEAVES. Permethyl ether Presence in extracts after permethylation Series Whole Cuticular Epidermal Middle leaf scrapings peel leaf Amentoflavone + m + ? Cupressuflavone + + + Agathisflavone + + + t Robustaflavone m m m Hinokiflavone t ? +, major band; m, minor band; t, trace detected; ?, trace unable to be positively identified. Section 8.3.3, p.81 a trimethyl amentoflavone (either 7"4'4"' or 74'4"'), 7-monomethyl agathisflavone, 77"-dimethyl agathisflavone, 7-monomethyl cupressuflavone, 77"-dimethyl cupressuflavone and 77"4'-trimethyl cupressuflavone. In addition, there was an undetermined dimethyl eupressuflavone and two partial methyl ethers of robustaflavone. There was also a trace amount of hinokiflavone or its methyl ether, as revealed by the presence of a trace of hinokiflavone pentamethyl ether in the permethylated raw extract (Table 15). Comparison of the one-dimensional separations in BPF revealed that both the epidermal and cuticular extracts contained the full range of biflavonoid bands obtained from the whole leaf extract. The results of the permethylations of each of the raw extracts are given in Table 15. Apart from the trace of hinokiflavone pentamethyl ether, all the permethyl ethers obtained from the whole leaf extract were also obtained from both the epidermal and cuticular extracts. Minor traces of two permethyl ethers were also found in the middle leaf extract. 8.3.4 Survey of Other Taxa. A survey of the distribution of AlCla-induced deep yellow fluorescence in the leaves of other taxa known to contain biflavones yielded the observations set out below. Since histochemical studies were not carried out in all cases, the term 'outer wall' is used to describe the full width of the outer periclinal wall of the epidermis. Psilotales. The epidermis of both the scale leaves and stems of Psilotum nudum showed deep yellow induced fluorescence throughout the outer wall and along almost the entire length of the anticlinal walls (cf. Figs 21, 22, 23). The inner periclinal wall fluoresced light blue. Figures 21-28. Fluorescence microscopy of fresh sections of other taxa. Scale: 20 um. Figure 21. TS of stem of Psilotum nudum, in oil. The epicuticular wax layer fluoresces light blue but the outer wall is non-fluorescent. Figure 22. As for Fig. 21, in AlCb. The outer periclinal and anticlinal walls of the epidermis fluoresce an intense deep yellow. The inner periclinal walls fluoresce light blue. Note that the reduced exposure does not register the initial blue autofluorescence because of the much greater intensity of the induced fluorescence. Figure 23. TS of Callitris muelleri leaf, in AlCb. The outer two-thirds of the 'outer wall' and the outer end of the anticlinal walls of the epidermis fluoresce an intense deep yellow. Figure 24. TS of Ginkgo biloba leaf, in AlCh, A thin surface layer of the 'outer wall' of the epidermis fluoresces intense yellow. The walls of the tracheids fluoresce light blue. Figure 25. TS of Pinus radiata leaf, in oil. The primary walls of the epidermis and hypodermis fluoresce light blue, while the cutucle is non-fluorescent. Figure 26. As for Fig. 25, in AlCb. Epidermal cell walls and the primary walls of the hypodermal cells fluoresce an intense light yellow but the cuticle and anticlinal flanges remain non-fluorescent. Light yellow fluorescence is also visible in the protoplasts of the epidermis, (See note on exposure in legend to Fig. 22.) Figure 27. TS of Cedrus deodara leaf, in oil. The thin layer of epicuticular wax can be seen to fluoresce light blue on the surface of the non-fluorescent cuticle. Some faint light blue fluorescence can be seen in the primary walls of the epidermal cells. Figure 28. As for Fig. 27, in AlCh. All walls of the epidermis fluoresce bright yellow and this fluorescence extends into the primary anticlinal walls of the hypodermis. The outer region of the 'outer wall' (?cuticle) remains non-fluorescent. m - - - ,... ..:. ·'-"-'-"-- ; --~--s..; -~ - ----== Section 8.3.4, p.82 Cycadales. Both surfaces of the strongly dorsiventral pinnae of Cycas revoluta showed bright yellow induced fluorescence in the inner quarter of the outer wall and deep yellow-brown induced fluorescence in the outer three-quarters. The pinnae of Lepidozamia peroffskyana showed deep yellow induced fluorescence in the outer three-quarters of the outer wall. Ginkgoales. The deciduous, fan-shaped leaves of Ginkgo biloba bore a superficial layer of intense yellow induced fluorescence (Fig. 24) which, while clearly present on both surfaces, was much more pronounced on the adaxial surface. These leaves were notable for a strong blue fluorescence, visible both with and without AlCb, that appeared to occur widely in the mesophyll and rapidly leached into the mounting medium. It was not possible to localise the source of this material. Coniferales. The flat leaves of Podocarpus elatus and Decussocarpus falcatus (Podocarpaceae) showed a similar distribution of deep yellow induced fluorescence to that described for Agathis robusta, i.e., throughout almost the full width of the outer wall and extending some distance in along the anticlinal walls of the epidermal cells. The appressed scale-leaves of Diselma archeri, Callitris muelleri, Callitris preissii subsp. murrayensis, Cupressus arizonica and Juniperus virginiana (Cupressaceae) all showed an intense deep yellow induced fluorescence in the outer two-thirds of the outer wall, with a broad wedge of fluorescence projecting inwards towards each anticlinal wall (Figs 23, 32). In the Callitris species the fluorescence was noticeably less concentrated in the elongated decurrent leaf base than in the scale itself. In Diselma archeri, large non-fluorescent inclusions were a feature of the fluorescent layer. :"igures 29-34. Fluorescence microscopy of fresh leaf sections. Figure 29. TS of the adaxial epidermis of a mature leaf of Agathis robusta mounted in oil showing autofluorescence. Figure 30. TS of the adaxial epidermis of mature leaf of Agathis robusta mounted in AlCb. The inner periclinal walls fluoresce blue in contrast to the intense deep yellow of the cuticle and cutinised outer periclinal and anticlinal walls. The anticlinal flange is non-fluorescent. The uncutinised cell wall (region 4) fluoresces a bright yellow. Figure 31. TS of the adaxial epidermis of the leaf of Juniperus conferta in AlCh showing layering of the outer wall into an outer superficial layer fluorescing bright yellow, and two distinct underlying layers of deep yellow fluorescence. Figure 32. TS of a mature leaf of Diselma archeri in AlCb showing intense deep yellow fluorescence in the outer two-thirds of the outer wall of the abaxial epidermis and all of the outer wall of the adaxial epidermis. Large non-fluore~cent inclusions can be seen in the abaxial (exposed) outer wall. Florin rings surrounding stomata can be seen in the adaxial epidermis. Figure 33. TS of Lepidozamia peroffskyana leaf mounted in AlCh showing deep yellow fluorescence in the outer wall. Figure 34. TS of the stem of Psilotum nudum mounted in AlCla showing a deep yellow fluorescence throughout the outer wall and along almost the entire length of the anticlinal walls. Section 8.3.4, p.83 The spreading awl-shaped leaves of Juniperus conferta (Cupressaceae) showed a marked differentiation between the two leaf surfaces. The outward-facing abaxial surface, which is devoid of stomata, possessed a superficial layer that fluoresced bright yellow and often tended to separate from the leaf during sectioning (Fig. 31). Beneath this were two distinct layers of less intense deep yellow fluorescence, and an inner non-fluorescent layer. All other walls of the epidermal cells were devoid of such fluorescence. The inward-facing adaxial surface, which bore numerous stomata, possessed a single, thin superficial layer of orange fluorescence covering an otherwise non-fluorescent epidermis. The outer wall of the deciduous leaves of Taxodium distichum (Taxodiaceae) bore a thin superficial layer of bright yellow fluorescent material. The needles of both Pinus radiata and Cedrus deodara (Pinaceae) showed a strong light yellow induced fluorescence in the inner half of the ouiter wall, in the inner periclinal and anticlinal walls of the epidermis and in the primary walls of the hypodermal fibres (cf. Figs 25 and 26, 27 and 28). In addition, there was an even light yellow fluorescence in the protoplasts of the epidermal cells. The outer half of the outer wall was in each case totally without fluorescence, and in Pinus radiata a narrow non-fluorescent wedge (? cuticular flange) was visible between the anticlinal walls of the epidermal cells (Fig. 26). Taxales. The strongly dorsiventral, spreading leaves of Taxus baccata var. fastigiata (Taxaceae) showed intense deep yellow induced fluorescence throughout the full width of the outer wall and also for a short distance in along the anticlinal walls. Gnetales. The broad flat leaves of Gnetum latifolium var. minus (Gnetales) showed no induced fluorescence in the epidermal walls, but Section 8.3.4, p.84 there was a bright yellow fluorescence within the protoplasts of scattered epidermal cells; this was particularly evident in epidermal peels. Examination of fresh sections of Oycas revoluta, Lepidozamia peroffskyana, Oallitris muelleri, Podocarpus elatus and Taxus baccata mounted in Sudan black B indicated that the extent of cutinisation of the epidermal walls in each case correlated strongly with the distribution of the AlCh-induced depp yellow fluorescence. 8.4 Discussion. The outer periclinal wall of the epidermis of Agathis robusta is interpreted as comprising a narrow outer cuticle proper and a broad layer of cutinised epidermal cell wall, overlying a narrow inner layer of non-cutinised epidermal cell wall. This interpretation is consistent with that of a generalised cuticle as described by Holloway ( 1982). The observation of a sharp but minutely convoluted boundary between the uncutinised and cutinised regions of the wall suggests a complex interdigitation between the two layers similar to that postulated for Ohamaecyparis lawsoniana by Oladele (1982). This is supported by scanning electron microscope studies of the inner surface of the cuticle of Agathis robusta after digestion of wall material (see Figs 8 and 9 in Stockey and Taylor 1981). The staining reactions of the central layer of the anticlinal epidermal walls indicate it essentially consists of cuticular waxes. The region is therefore interpreted as the 'cuticular peg' (Holloway 1982) or 'anticlinal flange' (Oladele 1982). Such anticlinal flanges have been reported to extend the full depth of the epidermis in Oallitris endlicheri (Oladele 1982), and are visible as projections of considerable size on the inner Section 8.4, p.85 surface of the digested cuticles of Agathis robusta illustrated by Stockey and Taylor (1981). Small crystals are scattered throughout the leaf tissue, with marked concentrations in the cuticle and epidermal cell walls. They are reduced or absent around stomata. Concentrations of crystals, stated to be calcium oxalate, have previously been reported in Agathis cuticles (Cookson and Duigan 1951) and are known to occur in the cuticles of other gymnosperms (Johnsen 1963; Alvin and Boulter 1974; Oladele 1982). The biflavones identified in the leaves of Agathis robusta during this study agree well with those reported from other species in the family (Khan et al. 1971, 1972; Ilyas et al. 1977b, 1978). Overall, the same four biflavonoid series extracted from Araucaria rulei (Ilyas et al. 1977b) are present in Agathis robusta, viz., amentoflavone, cupressuflavone, agathisflavone and robustaflavone, and in addition a minor trace of hinokiflavone was detected. By far the major proportion of the biflavonoid fraction consisted of partial methyl ethers of these series. Observations on the AlCls-induced fluorescence in fresh leaf sections showed that the overwhelming proportion of the biflavonoid content is localised in the outer periclinal wall of the epidermis. This was confirmed by the isolation of all the major biflavonoid bands detected in the whole leaf extract from extracts of cuticular scrapings and adaxial epidermal peels, and a failure to extract significant amounts of biflavonoids from middle leaf tissue. The trace amounts of biflavonoids detected in the permethylated middle leaf extract may be due to the presence of small amounts of these compounds in other regions of the leaf. There is evidence to suggest that synthesis of some flavonoids is closely associated with chloroplast activity Section 8.4, p.86 (Saunders and McClure 1976a, b), and if this also applies to biflavonoids their presence in the middle leaf fraction would not be surprising. However, there is evidence to indicate that other flavonoids are synthesised and stored entirely within the epidermis (McClure 1975; Hrazdina et al. 1980). It is also possible that the trace of biflavonoids detected in the middle leaf extract is simply due to contamination by epidermal tissue. The close association between the epidermis and the bundles of hypodermal fibres made the clean separation of adaxial epidermal peels difficult, and of abaxial peels virtually impossible. Hence the inclusion of small fragments of epidermis in the middle leaf tissue fraction cannot be ruled out. The absence of biflavonoids from the anticlinal flange is surprising, since in previous studies no difference in staining properties between the cuticle proper and the flange have been detected (Holloway 1982), Although the flange is continuous with a cuticle proper that is also non-fluorescent in immature leaves, it does not appear to accumulate biflavonoids with age as the cuticle proper appears to do. This suggests that there is a differentiation between the cuticle proper and the flange in Agathis robust& Oladele (1982) concludes that the "almost ubiquitous occurrence of anticlinal flanges in well developed plant cuticles may indicate they are a consequence of some physical relationship between neighbouring epidermal cells." This implies complete equivalence in structure of the two regions. Further, no such differentiation between the flange and the cuticle proper was seen in any of the other species studied. The existince of this differentiation in Agathis robusta must therefore be seen as a specialization: a more detailed investigation of the fine structure of the epidermis and cuticle in this species is needed to clarify the point. Section 8.4, p.87 The survey of other gymnosperm species known to contain biflavonoids revealed that, while there were individual differences in the extent of AlCh-induced fluorescence in the outer periclinal and anticlinal walls of the epidermis, all were notable for the heavy concentration of this fluorescence in the 'outer wall'. The variation in the distribution of fluorescence is interpreted as simply reflecting the extent of cutinisation of the walls. Sudan black B staining in several species indicated that the pattern of lipid deposition was strongly correlated with the distribution of the induced fluorescence. It is concluded, therefore, that the biflavonoid content of gymnosperm leaves is characteristically concentrated in the cuticle. This is also true for Psilotum nudum, a member of the primitive order of rootless plants, the Psilotales, which is also known to contain biflavonoids (Wallace and Markham 1978). In contrast, those species belonging to the Pinaceae and Gnetales showed no AlCu-induced fluorescence in the cuticle; in all cases it was localised in the walls and protoplasts of the epidermal and hypodermal cells. This difference in distribution of induced fluorescence correlates with the known absence of biflavonoids in these taxa. The flavonoid content of the leaves of the Pinaceae has not been studied in great detail (Niemann 1979; Parker et al. 1979), and no attempt was made in this investigation to extract and identify the compounds responsible for the induced fluorescence. It is not possible to resolve the identity of individual flavonoids present in a complex mixture of flavonoids by using AlCb-induced fluorescence alone. It is per haps not surprising to find lipophilic biflavonoid aglycones associated with the waxy cuticles of gymnosperm leaves. Free flavonoid aglycones have now been isolated from a wide range of plants (Wollenweber and Dietz 1981) and, whereas glycosidic flavones tend to Section 8.4, p.88 be water soluble and thus are commonly located in cell vacuoles (McClure 1975; Hrazdina et al. 1980, 1982; Tissut and Ravanel 1980), flavonoid aglycones are scarcely soluble in water and tend to become more lipophlic with increasing methylation (Wollenweber and Dietz 1981). Biflavone glycosides are unknown in the gymnosperms, although they are a minor component of the biflavonoid fraction of both Psilotum nudum and Tmesipteris tannensis (Spreng.) Bernh. (Wallace and Markham 1978). The occurrence of flavonoid aglycones in leaf tissue, therefore, tends to be associated with secretary structures or with the production of other lipophilic plant products: e.g., the secretions from glandular trichomes of some ferns and primroses (Chance and Arnott 1981; Wollenweber and Dietz 1981; Wollenweber et al. 1981; Wollenweber 1982), epicuticular leaf waxes of some Eucalyptus species (Wollenweber and Dietz 1981; Wollenweber and Kohorst 1981), or in the external phenolic resin of Primula and Malus leaves (Baker 1982). The marked accumulation of biflavonoids in the leaves of so many gymnosperms leads us to consider their adaptive significance. The function of the epidermal accumulation of some flavonoids, particularly those confined to vacuoles (e.g., the colour determining chalcones and anthocyanins), is related to the visual perception of flowers, fruits or spores by pollinators or dispersal agents (Harborne 1980). The u.v. absorbing properties of flavonoids are also well known (Mabry et al. 1970), and while these may also contribute to visual perception, Caldwell (1968, 1971), Lowry et al. (1980, 1983) and others have suggested that a significant role of epidermal flavonoids is, or has been, in affording some protection against damaging wavelengths of natural u.v. irradiation. Flavonoids (including biflavonoids) certainly absorb strongly those wavelengths (260-280 nm) that are most effective in producing nucleotide and protein damage (McClure 1975; WHO Section 8.4, p.89 1979). Present intensities of these wavelengths in the natural irradiation are quite low (WHO 1979; Lowry et al. 1980), but they are thought to have been much higher at the time of the evolution of the first land plants and this has been postulated as the reason for the presence of flavonoids in the epidermis of such a broad range of the extant vascular flora (Lowry et al. 1980, 1983). Exposure of plants to wavelengths in the region of 260 - 280 nm is also known to be very effective in inducing flavonoid accumulation (Caldwell 1968, 1971; McClure 1975). Further, flavonoids are relatively transparent to the longer wavelengths of light essential for photosynthesis (Lowry et al. 1980). Lowry et al. (1980) considered that epidermal flavonoids also served secondary roles in early land plants, particularly as defence mechanisms against predation. It certainly seems difficult to explain the almost universal occurrence of large concentrations of biflavonoids in the leaves of gymnosperms simply as the retention of an ancestorial u.v. screen that was adaptive to a previous, but no longer existing, environment. The occurrence of complex mixtures of biflavonoids based on a range of different skeletons is also difficult to account for on this basis, since there is virtually no difference in their absorption spectra. Both these features favour a strong adaptive role in modern plants. The reported occurrence of biflavonoids in, for example, the heartwood (Chen et al. 1975; Cotterill et al. 1977), bark (Waterman and Crichton 1980) and fruits (Chen et al. 1974; Lin and Chen 1974a, b) of higher plants also indicates these compounds have now assumed other primary roles. Particular flavonoids are known to have antimicrobial properties (Harborne 1977, 1980), although the basis of this activity is not understood. Swain (cited in Wollenweber and Dietz 1981) has postulated Section 8.4, p.90 that lipophilic flavonoids, particularly methylated flavonoids, offer protection against microorganisms because of their ability to penetrate membranes. A similar explanation has been postulated by O'Neill and Mansfield (1982) for isoflavones with antifungal properties. The latter authors report that complete methylation of the hydroxyl groups removed the antifungal activity; this may account for the rarity of biflavonoid permethyl ethers in nature (Geiger and Quinn 1982). Biflavonoids may also perform a protective role as a deterrent to leaf-eating organisms. There is evidence to support such a role for some flavonoids (Harborne 1980). The ability of living organisms to evolve to overcome such defences might explain the variety of biflavonoids found in many species (Geiger and Quinn 1982). Some such defense mechanism seems probable as the primary role of the biflavonoids in the gymnosperms on the grounds of their localisation in the leaf cuticle, the diversity of their structures and their occurrence in such a wide range of taxa. It is possible that the induced fluorescence in the epidermis and hypodermis of the Pinaceae and Gnetales may by found to be due to compounds performing a similar role to that of biflavonoids. 91 REFERENCES. Adams, R. P., Rudloff, E. von, Zanoni, T. A., and Hogge, L. (1980). The terpenoids of an ancestral/advanced species pair of Juniperus. Biochem. Syst. Ecol. 8, 35-7. Adams, R. P., Rudlff, E. von, and Hogge, L. (1981). The south-western USA and northern Mexico one-seeded junipers: their volatile oils and evolution. Biochem. Syst. Ecol. 9, 93-6. Adams, R. P., Rudloff, E. von, and Hogge, L. (1983). Chemosystematic studies of the western North American junipers based on their volatile oils. Biochem. Syst. Ecol. 11, 189-193. Al-Sherifi, K. H. (1952). 'Histological Studies on the Shoot Apices and Leaves of Certain Cupressaceae.' Ph.D. Thesis, University of California. Alvin, K. L., and Boulter, M. C. (1974). A controlled method of comparative study for Taxodiaceaous leaf cuticles. Bot. J. Linn. Soc. 69, 277-86. Alvin, K. L., Dalby, D. H., and Oladele, F. A. (1981). Numerical analysis of cuticular characters in Cupressaceae. In 'The Plant Cuticle', eds D. F. Cutler, K. L. Alvin, and C. E. Price, pp. 379-96. (Academic Press: London.) 92 Ansari, F. R., Ansari, W. H., Rahman, W., Okigawa, M., and Kawano, N. (1981). Flavonoids from the leaves of Callitris glauca R. Br. (Cupressaceae). Indian J. Chem. 20B, 724-5. Baird, A. M. (1953). The life history of Callitris. Phytomorphology 3, 258-84. Baker, E. A. (1981). Chemistry and morphology of plant epicuticular waxes. In 'The Plant Cuticle', eds D. F. Cutler, K. L. Alvin, and C. E. Price, pp. 139-66. (Academic Press: London.) Bankova, S.S., Popov, S. S., and Marekov, N. L. (1982). High-performance liquid chromatographic analysis of flavonoids from propolis. J. Chromatogr. 242, 135-43. Bannan, M. W. (1941). Wood structure of Thuja occidentalis. Bot. Gaz. 103, 295-309. Bannan, M.W. (1942). Wood structure of the native Ontario species of Juniperus. Am. J. Bot. 29, 245-51. Bannan, M. W. (1944). Wood structure of Libocedrus decurrens. Am. J. Bot. 31, 346-51. Bannan, M. W. (1952). The microscopic wood structure of north American species of Chamaecyparis. Can. J. Bot. 30, 170-87. Bannan, M. W. (1954). The wood structure of some Arizonan and Californian species of Cupressus. Can. J. Bot. 32, 285-314. 93 Bauch, J., Liese, W., and Schultze, R. (1972). The morphological variability of the bordered pit membranes in gymnosperms. Wood Science and Technology 6, 165-84. Beckmann, S., Geiger, H., and de Groot-Pfteiderer, W. (1971). Biflavone und 2,3-dihydrobiflavone Metesequoia glyptostroboides. Phytochemistry 10, 2465-73. Boutelje, J. (1955). The wood anatomy of Libocedrus Endl., s. lat., and Fitzroya J. D. Hook. Acta Horti Bergiani 17, 177-215. Boyle, P., and Doyle, J. (1953). Development in Podocarpus nivalis in relation to other podocarps. I. Gametophytes and fertilization. Sc. Proc. R. Dublin Soc. 26, 179-205. Brehm, B. G., and Krell, D. (1975). Flavonoid localization in epidermal papillae of flower petals: a specialized adaptation for ultraviolet absorption. Science 190, 1221-3. Briancon-Scheid, F., Guth, A., and Anton, R. (1982). High-performance liquid chromatography of biflavones from Ginko biloba L. J. Chromatogr. 245, 261-67. Buchholz, J. T. (1946). Gymnosperms. In 'Encyclopedia Britanica', pp. 22-34. (Encyclopedia Britannica: London.) Buchholz, J. T. (1948). Generic and subgeneric distribution of the coniferales. Bot. Gaz. 110, 80-91. 94 Caldwell, M. M. (1968). Solar ultraviolet radiation as an ecological factor for alpine plants. Ecol. Monogr. 38, 243-67. Caldwell, M. M. (1971). Solar UV irradiation and the growth and development of higher plants. In 'Photophysiology', ed. A. C. Griese, pp. 131-75. (Academic Press: New Yo.rk.) Carman, R. M., and Sutherland, M. D. (1979.) Cupressene and other diterpenes of Cupressus species. Aust. J. Chem. 32, 1131-42. Carr, S. G. M., and Carr, D. J. (1977). Diagnostic anatomical characters of the leaves of three species of Agathis. Appendix in Hyland ( 1977). Brunonia 1, 103-15. Casteele, K. V., Geiger, H., and Sumere, C. F. van (1982). Separation of flavonoids by reversed-phase high-performance liquid chromatography. J. Chromatogr. 240, 81-94. Chance, G. D., and Arnott, H. J. (1981). SEM of frond farina in Pityrogramma triangularis. Scanning Electron Miscrosc. 1981, 273-8. Chandramouli, N., Natarjan, S., Murti, V. V. S., and Seshadri, T. R. (1971). Structure of heveaflavone. Indian J. Chem. 9, 895-6. Chatterjee, A., Kotoky, J., Das, K. K., Banerji, J., and Chakraborty, T. (1984). Abiesin, a biflavonoid of Abies webbiana. Phytochemistry 23, 704-5. 95 Chen, F., Lin, Y,, and Liang, C. (1974), Biflavonyls from drupes of Rhus succedanea (Anarcardiaceae). Phytochemistry 13, 276-8. Chen, F., Lin, Y,, and Hung, S. (1975), Phenolic compounds from the heartwood of Garcinia multiflora. Phytochemistry 14, 300-3. Chexal, K, K., Handa, B. K,, and Rahman, W. (1970), Thin-layer chromatography of biflavones on silica gel. J. Chromatogr. 48, 484-92. Coates Palgrave, K, (1983). 'Trees of Southern Africa.' (C. Struik: Capetown.) Cookson, I. C., and Duigan, S. L. (1951). Tertiary Araucariaceae from south-eastern Australia, with notes on living species. Aust. J. Sci. Res. (B) 4, 415-49. Cotterill, P. J., Scheinmann, F., and Puranck, G. S. (1977). Phenolic compounds from the heartwood of Garcinia indica. Phytochemistry 16, 148-9. Daigle, D. J., and Conkerton, E. J. (1982). High-performance liquid chromatography of 34 selected flavonoids. J. Chromatogr. 240, 202-5. Dallimore, W., and Jackson, A. B. (1966). 'Handbook of Coniferae and Ginkgoaceae'. Revised by S. G. Harrison. 4th edition. (Edward Arnold: London.) 96 Dossaji, S. F., Mabry, T. J., and Bell, E. A. (1975a). Biflavanoids of the Cycadales. Biochem. Syst. Ecol. 2, 171-5. Dossaji, S. F., Mabry, T. J., and Wallace, J. W. (1975b). Chromatographic and uv-visible spectral identification of biflavanoids. Rev. Latinoamer. Quim. 6, 37-45. Ebel, J., and Hahlbrock, H. (1982). Biosynthesis. In 'The Flavonoids: Advances in Research', eds J. B. Harborne and T. J. Mabry, pp. 641-80. (Chapman and Hall: London.) Eckenwalder, J. E. (1976a). Re-evaluation of Cupressaceae and Taxodiaceae: a proposed merger. Madrano 23, 237-300. Eckenwalder, J. E. (1976b). Comments on 'A new classification of the conifers'. Taxon 25, 337-9. Eichler, A. W. (1889). Gymnospermae. In 'Die Naturlichen Pflanzenfamilien', eds A. Engler and K. Prantl, pp. 6-127. (Engelmann; Leipzig.) Endlicher, S. (1847). 'Synopsis Coniferarum'. (Apud Scheitlin and Zollikofer: Sangalli.) Erdtman, G. (1965). 'An Introduction to Palynology III'. (Almqvist and Wiksell: Stockholm.) Erdtman, H. and Norin, T. (1966). The chemistry of the order Cupressales. Fortschritte d. Chem. org, Naturst. 24, 206-87. 97 Faegri, K., and Iversen, J. (1975). 'Textbook of Pollen Analysis'. 3rd edition. (Munksgaard: Copenhagen.) Fatma, W., Taufeeq, H. M., Shaida, W. A., and Rahman, W. (1979). Biflavonoids from Juniperus macropoda Boiss and Juniperus phoenicea Linn. (Cupressaceae). Ind. J. Chem. 17B, 193-4. Feder, N., and O'Brien, T. P. (1968). Plant microtechnique: some principles and new methods. Am. J. Bot. 55, 123-42. Fisher, D. B. (1968). Protein staining of ribboned epon sections from light microscopy. Histochemie 16, 92-6. Fitzpatrick, H. M. (1965). Conifers: keys to the genera and species, with economic notes. Royal Dublin Soc., Proc. A, 2, 67-129. Florin, R. (1931). Untersuchungen zur stammesgeschichte der coniferales und cordaitales. K. Sv. Vet. Akad. Handl. Ser. 3, 10, 1, 1-588. Florin, R. (1963). The distribution of conifer and taxad genera in time and space. Acta Horti Bergiani 20, 121-312. Florin, R., and Boutelje, J. (1954). External morphology and epidermal structure of leaves in the genus Libocedrus, s. lat. Acta Horti Bergiani 17, 8-37. Fulcher, R. G., O'Brien, T. P., and Lee, J. W. (1972). Studies on the aleurone layer. I. Conventional and fluorescence microscopy of the 98 cell wall with emphasis on phenol-carbohydrate complexes in wheat. Aust. J. Biol. Sci. 25, 23-34. Gadek, P. A. (1982). Biflavonoids from the seed testa of Cycadales. Phytochemistry 21, 889-90. Galensa, R., and Herrmann, K. (1980). Analysis of flavonoids by high-performance liquid chromatography. J. Chromatogr. 189, 217-24. Gardner, J. A. F. (1962). The Tropolones. In 'Wood Extractives', ed. W. E. Hillis, pp. 317-29. (Academic Press: New York.) Gaussen, H. (1968). Les Gymnospermes. Actuelles et Fossiles. X. Les Cupressacees. Trav. du. Laboratoire forestier de Touleuse, Tome II, Sect. 1, Vol. 1, part 2, fasc. X, Chapt. 13. Geiger, H., and Quinn, C. J. (1975). Biflavonoids. In 'The Flavonoids', eds J. B. Harborne, T. J. Mabry and H. Mabry, pp. 692-742. (Chapman and Hall: London.) Geiger, H., and Quinn, C. J. (1982). Biflavonoids. In 'The Flavonoids: Advances in Research', eds J. B. Harborne and T. S. Mabry, pp. 505-34. (Chapman and Hall: London.) Giannasi, D. E. (1978). Systematic aspects of flavonoid biosynthesis and evolution. Bot. Rev. 44, 399-429. Gough, L. J. (unpublished). A comparative study of some resins from the family Cupressaceae. 99 Gough, L. J., and Welch, H. J. (1978). Nomenclatural transfer of Chamaecyparis obtusa (Siebold and Zucc.) Endl. 'Sanderi' (Cupressaceae) to Thuja orientalis L. 'Sanderi' on the basis of phytochemical data. Bot. J. Linn. Soc. 77, 217-221. Greguss, P. (1955). 'Identification of Living Gymnosperms on the Basis of Xylotomy'. (Akademiai Kiado: Budapest.) Gullvag, B. M. (1966). The fine structure of some gymnosperm pollen walls. Grana Palyn. 6, 435-49. Hameed, N., Ilyas, M., Rahman, W., Okigawa, M., and Kawano, N. (1973). Biflavones in the leaves of two Juniperus plants. Phytochemistry 12, 1494-5. Harborne, J. B. (1973). Flavonoids. In 'Phytochemistry', ed. L. P. Miller, pp. 350-379. (Van Nostrand Reinhold: New York.) Harborne, J. B. (1975). Biochemical systematics of flavonoids. In 'The Flavonoids', eds J. B. Harborne, T. J. Mabry and H. Mabry, pp. 1056-95. (Chapman and Hall: London). Harborne, J. B. (1977). Flavonoids and the evolution of the Angiosperms. Biochem. Syst. Ecol. 5, 7-22. Harborne, J. B. (1980). Plant phenolics. In 'Encyclopedia of Plant Physiology. Vol. 8. Secondary Plant Products', eds E. A. Bell and B. V. Charlwood, pp. 329-95. (Springer-Verlag: Berlin.) 100 Harborne, J. B., and Mabry, T. J., eds. (1982). 'The Flavonoids: Advances in Research'. (Chapman and Hall: London.) Harborne, J. B., Mabry, T. J., and Mabry, H., eds. (1975). 'The Flavonoids'. (Chapman and Hall: London.) Harris, P. J., and Hartley, R. D. (1976). Detection of bound Ferulic acid in cell walls of the Gramineae by ultraviolet microscopy. Nature ( London) 259, 508-10. Harris, P. J., and Hartley, R. D. (1980). Phenolic constituents of the cell walls of monocotyledons. Biochem. Syst. Ecol. 8, 153-60. Harrison (1966). see Dallimore and Jackson (1966). Hartley, R. D., and Harris, P. J. (1981). Phenolic constituents of the cell walls of dicotyledons. Biochem. Syst. Ecol. 9, 189-203. Hegnauer, R. (1969). Chemical evidence for the classification of some plant taxa. In 'Perspectives in Phytochemistry', eds J. B. Harborne and T. Swain, pp. 121-138. (Academic Press: London.) Holloway, P. J. (1981). Structure and histochemistry of plant cuticular membranes: an overview. In 'The Plant Cuticle', eds D. F. Cutler, K. L. Alvin and C. E. Price, pp. 1-32. (Academic Press: London.) Hrazdina, G., Alscher-Herman, R., and Kish, V. M. (1980). Subcellular localization of flavonoid synthesizing enzymes in Pisum, Phaseolus, Brassies, and Spinacia cultivars. Phytochemistry 19, 1355-9. 101 Hrazdina, G., Marx, G. A., and Hoch, H. C. (1982). Distribution of secondary plant metabolites and their biosynthetic enzymes in pea (Pisum sativum L.) leaves. Plant Physiol. 70, 745-8. Hyland, B. P. M. (1977). A revision of the genus Agathis (Araucariaceae) in Australia. Brunonia 1, 103-15. Ilyas, M., Ilyas, N., and Wagner, H. (1977a). Biflavones and flavonol-O-glycosides from Juniperus macropoda. Phytochemistry 16, 1456-7. Ilyas, M., Seligmann, O., and Wagner, H. (1977b). Biflavones from the leaves of Araucaria rulei F. Muell. and a survey of biflavonoids of the Araucaria genus. z. Naturforsch. 32c, 206-9. Ilyas, N., Ilyas, M., Rahman, W., Okigawa, M., and Kawano, N. (1978). Biflavones from the leaves of Araucaria excelsa. Phytochemistry 17, 987-90. Janchen, E. (1950). Das system der koniferem. Osterr. Akad. Wiss., Math. Naturwiss. Kl., Sitzungsber., Abt. 1, Biol. 158, 155-262. Johnsen, T. N., Jr (1963). Anatomy of scalelike leaves of Arizona junipers. Bot. Gaz. (Chicago) 124, 220-4. Kamil, M., Ilyas, M., Rahman, W., Hasaka, N., Okigawa, M., and Kawano, N. (1977). Taiwaniaflavone: a new series of naturally occurring biflavones from Taiwania cryptomerioides. Chem. and Incj. 102 Kamil, M., Ilyas, M., Rahman, W., Hasaka, N., Okigawa, M., and Kawano, N. (1981). Taiwaniaflavone and its derivatives: a new series of biflavones from Taiwania cryptomerioides Hayata. J. C. S. Perkin I, 1981, 553-9. Kausik, S. B (1976). A contribution to the foliar anatomy of Agathis dammara, with a discussion on the transfusion tissue and stomata} structure. Phytomorphology 26, 263-76. Keng, H. (1975). A new scheme of classification of the conifers. Taxon 24, 289-92. Khan, N. U., Ansari, W. H., Usmani, J. N., Ilyas, M., and Rahman, W•. (1971). Biflavonyls of the Araucariales. Phytochemistry 10, 2129-31. Khan, N. U., Ilyas, M., Rahman, W., Mashima, T., Okigawa, M., and Kawona, N. (1972). Biflavones from the leaves of Araucaria bidwillii Hooker and Agathis alba Foxworthy' (Araucariaceae). Tetrahedron 28, 5689-95. Klemm, P. ( 1886). Ueber den bau der beblatterten zweige der cupressineen, Jahrbuch fur Wissenschaftliche Botanik 17, 498-541. Konar, R. N., and Oberoi, Y. P, (1969). Recent work on reproductive structures of living conifers and taxads - a review. Bot. Rev. 35, 89-116. Krussmann, G. (1971). 'Handbuch der Nadelgeholze'. (Verlag Paul Parey: Berlin,) 103 Lamer-Zarawaska, E. (1975). Biflavonoids in Juniperus L. sp. (Cupressaceae). Pol. J. Pharmacol. Pharm. 27, 81-7. Laubenfels, D. J. de (1953). The external morphology of coniferous leaves. Phytomorphology 3, 1-20. Laubenfels, D. J. de (1965). The relationships of Fitzroya cupressoides (Molina) Johnston and Diselma archeri J. D. Hooker based on morphological considerations. Phytomorphology 15, 414-19. Lebreton, P. (1982). Les cupressales: une definition chimiosystematique. Candollea 37, 243-56. Lebreton, P., Jay, M., and Voirin, B. (1967). Sur l'analyse qualitative et quantitative des flavonoids. Chimie analytique 49, 375-83. Lebreton, P., Boutard, B., and Sartre, J. (1978). Biochimie flavonique comparee de coniferes du Maroc et de France. Bulletin de l'Institut Scientifique, Rabot 1978, 155-68. Lemoine-Sebastian, C. (1969). Vascularization du complexe bractee ecaille dans le cone femelle des Cupressacees. Botanies Rhedonica A 7, 3-27. Lemoine-Sebastian, C. (1971). The vascularization of the bract-scale complex in the female cone of the Cupressaceae. New Observations. Botanica Rhedonica A 11, 177-87. Lemoine-Sebastian, C. (1972). Etude comparative de la vascularization et 104 du complexe seminal chez lee Cupressacees. Phytomorphology 22, 246-60. Li, H. (1953). A reclassification of Libocedrus and Cupressaceae. J. Arn. Arb. 34, 17-35. Li, S. (1972). The female reproductive organs of Chamaecyparis. Taiwania 17, 27-39. Liese, W. (1965). The warty layer. In 'Cellular Ultrastructure of Woody Plants', ed. Cote, W. A., pp. 251-69. (Syracuse University Press: Syracuse.) Lin, Y., and Chen, F. (1974a). Agathisflavone from the drupes of Rhus succedanea. Phytochemistry 13, 657-8. Lin, Y., and Chen, F. (1974b). Robustaflavone from the seed kernels of Rhus succedanea. Phytochemistry 13, 1617-19. Lin, Y. M., and Chen, F. C. (1975). TLC Rf values of biflavanoids and their methylated derivatives. J. Chromatogr. 103, D33-4. Lowry, J. B., Lee, D. W., and Hebant, C. (1980). The origin of land plants: a new look at an old problem. Taxon 29, 183-97. Lowry, J. B., Lee, D. W., and Hebant, C. (1983). The origin of land plants: a reply to Swain. Taxon 32, 101-3. 105 Mabry, T. J., Markham, K. R., and Thomas, M. B. (1970). 'The Systematic Identification of Flavonoids.' (Springer-Verlag: Berlin) Markham, K. R. (1982). 'Techniques of Flavonoid Identification'. (Academic Press: London.) Martin, J. T., and Juniper, B. E. (1970). 'The Cuticles of Plants.' (Edward Arnold: London.) Meylan, B. A., and Butterfield, B. G. (1980). 'Three-dimensional Structure of Wood. An Ultrastructural Approach.' 2nd edition. (Chapman and Hall: London.) McClure, J. W. (1975). Physiology and functions of flavonoids. In 'The Flavonoids', eds J. B. Harborne, T. J. Mabry and M. Mabry, pp. 970-1055. (Chapman and Hall: London.) Miura, H., and Kawano, N. (1968). On the distribution of bisflavones in the leaves of Taxodiaceae and Cupressaceae plants. J. Pharm. Soc. Japan 88, 1459-62. Moseley, M. F. Jr (1943). Contributions to the life history, morphology and phylogeny of Widdringtonia cupressoides. Lloydia 6, 109-32. Murti, V. V. S., Raman, P. V., and Seshadri, T. R. (1967). Cupressuflavone, a new biflavonyl pigment. Tetrahedron 23, 397-404. Namboodiri, K. K., and Beck, C. B. (1968). A comparative study of the 106 primary vascular system of conifers. II. Genera with opposite and whorled phyllotaxis. Am. J. Bot. 55. 458-63. Napp-Zinn, K. (1966). 'Anatomie des Blattes. I. Blattanatomie der Gymnospermen.' (Gebruder Borntraeger: Berlin.) Natarjan, S., Murti, V. V. S., and Seshadri, T. R. (1970). Biflavones of some Cupressaceae plants. Phytochemistry 9, 575-9. Neger, F. W. (1907). Die nadelholzer (koniferen) und ubrigen gymnospermen. Sammlung Goschen 355. Neimann, G. J. (1979). Some aspects of the chemistry of Pinaceae needles. Acta. Bot. Neerl. 28, 73-88. O'Brien, T. P., and McCully, M. E. (1981). 'The Study of Plant Structure. Principles and Selected Methods.' (Termarcarphi Pty Ltd: Melbourne.) Ohtani, J., Meylan, B. A., and Butterfield, B. G. (1983). Occurrence of warts in vessel elements and fibres of New Zealand. N.Z. J. Bot. 21, 359-72. Ohtani, J., Meylan, B. A., and Butterfield, B. G. (1984). Vestures or warts - proposed terminology. IAWA Bull. 5, 3-8. Oladele, F. A. (1982). Development of the crystalliferous cuticle of Chamaecyparis lawsoniana (A. Murr.) Parl. (Cupressaceae). Bot. J. Linn. Soc. 84, 273-88. 107 Oladele, F. A. (1983a). Inner surface sculpture patterns of cuticles in the Cupressaceae. Can. J. Bot. 61, 1222-31. Oladele, F. A. (1983b). Scanning electron microscope study of stomatal-complex configuration in Cupressaceae. Can. J. Bot. 61, 1232-40. O'Neill, T. M., and Mansfield, J. W. (1982). Antifungal activity of hydroxyflavans and other flavonoids. Trans. Br. Mycol. Soc. 79, 229-37. Owens, J. N., and Molder, M. (1974). Cone initiation and development before dormancy in yellow cedar (Chamaecyparis nootkatensis). Can. J. Bot. 52, 2075-84. Owens, J. N., and Molder, M. (1975). Pollination, female gametophyte, and embryo and seed development in yellow cedar ( Chamaecyparis nootkatensis). Can. J. Bot. 53, 186-99. Owens, J. N., and Molder, M. (1980). Sexual reproduction in western red cedar (Thuja plicata). Can. J. Bot. 58, 1376-93 •• Parker, W. H., Maze, J., and McLachlan, D. G. (1979). Flavonoids of Abies amabilis needles. Phytochemistry 18, 508-10. Pascual Teresa, J. de, Barrero, A. F., Muriel, L., San Feliciano, A., and Grande, M. (1980). New diterpene acids from Juniperus communis. Phtyochemistry 19, 1153-6. 108 Patel, R. N. (1968). Wood anatomy of Cupressaceae and Araucariaceae indigenous to New Zealand. N.Z. J. Bot. 6, 9-18. Pearse, A. G. E. (1968). 'Histochemistry, Theoretical and Applied.' Vol. 1, 3rd edn. (Churchill: London.) Pelter, A., Warren, R., Hameed, N., Khan, N. U., Ilyas, M., and Rahman, W. (1970). Biflavonyl pigments from Thuja orientalis (Cupressaceae). Phytochemistry 9, 1897-8. Pelter, A., Warren, R., Hammeed, N., Ilyas, M., and Rahman, W. (1971). Biflavonyls pigments from Juniperus chinensis (Cupressaceae). J. Indian Chem. Soc. 48, 204-5. Phillips, E. W. J. (1941). The identification of coniferous woods by microscopic structure. J. Linn. Soc. Bot. 52, 259-320. Peirce, A. S. (1937). Systematic anatomy of the woods of the Cupressaceae. Tropical Woods 49, 5-21. Pilger, R. von (1926). Gymnosperms. In 'Die Naturlichen Pflanzenfamilien', eds A. Engler and K. Prantl, pp. 361-403. (Duncker and Humblot: Berlin.) Pillman, A. (1978). A study of the primary vascular system and evolution in the family Cupressaceae. M.Sc. Thesis, University of Adelaide. Pocknall, D. T. (1981a). Pollen morphology of the New Zealand species of 109 Dacrydium Solander, Podocarpus L'Heritier, and Dacrycarpus Endlicher (Podocarpaceae). N.Z. J. Bot. 19, 67-95. Pocknall, D. T. (1981b). Pollen morphology of Phyllocladus L. C. et A. Rich. N.Z. J. Bot. 19, 259-66. Pocknall, D. T. (1981c). Pollen morphology of the New Zealand sprecies of Libocedrus Endlicher (Cupressaceae) and Agathis Salisbury (Araucariaceae). N.Z. J. Bot. 19, 267-72. Quinn, C. J. (1982). Taxonomy of Dacrydium Sol. ex Lamb. emend. de Laub. (Podocarpaceae). Aust. J. Bot. 30, 311-20. Quinn, C. J., and Gadek, P. A. (1981). Biflavones of Dacrydium sensu. lato. Phytochemistry 20, 677-82. Rahman, W., Hameed, N., and Ilyas, M. (1972). Biflavonyl pigments from Thuja plicata (Cupressaceae). J. Indian Chem, Soc. 49. Roscher, J, (1974). Nonsaccate gymnosperm pollen: micromorphology of the exine. Am. J. Bot. 61, 49. Abstract only, Rudloff, E. von (1975). Volatile leaf oil analysis in chemosystematic studies of North American conifers. Biochem. Syst. Ecol. 2, 131-67. Saunders, J, A,, and McClure, J. W. (1976a). The distribution og flavonoids in chloroplasts of twenty five species of vascular plants. Phytochemistry 15, 809-10. 110 Saunders, J. A., and McClure, J. W. (1976b). The occurrence and photoregulation of flavonoids in Barley plastids. Phytochemistry 15, 805-7. Saxton, w. T. (1910a). Conrtibutions to the life-history of Widdringtonia cupressoides. Bot. Gaz. 50, 31-48. Saxton, W. T. (1910b). Contributions to the life-history of Callitris. Ann. Bot. 24, 557-69. Saxton, W. T. (1913a). Contributions to the life-history of Actinostrobus pyramidalis Migr. S. Afr. J, Sci. 27, 321-46. Saxton, W. T. (1913b). Contributions to the life-history of Tetraclinis articulata Masters, with some notes on the phylogeny of the Callitroideae. S, Afr. J. Sci. 27, 577-606. Saxton, W. T. (1913c). The classification of conifers. New Phytologist 12, 1-21. Saxton, W. T. (1929). Gymnosperms. In 'The Encyclopaedia Britannica', 14th edition, pp. 22-34. (The Encyclopaedia Britannica:London.) Seigler, D. S. (1981). Secondary metabolites and plant systematics. In 'The Biochemistry of Plants', pp. 139-76. (Academic Press: Ney York.) Singh, H. (1978). Embryology of Gymnosperms. Handbuch de Pflanzenatomie Spezieller Band 10, Tiel 2. (Gebruder Borntraege: Berlin.) 111 Singh, H., and Oberoi, Y. P. (1962). A contribution to the life-history of Biota orientalis Endl. Phytomorphology 12, 373-93. Siva Prasad, J., and Krishnamurty, H. G. (1977). 4-epiisocommunic acid and amentoflavone from Callitris rhombodea. Phytochemistry 16, 801-3. Staflen, F. A, (1981), 'Index Herbariorum', Pt, 1. (Dr. W, Junk: The Hague,) Sterling, C. (1963), Structure of the male gametophyte in gymnosperms. Biol, Rev. 38, 167-203, Stockey, R, A., and Taylor, T. N. (1981). Scanning Electron Microscopy of epidermal patterns and cuticular structure in the genus Agathis, Scanning Electron Micros, 1981, 207-12. Swain, T, (1975). Evolution of flavonoid compounds. In 'The Flavonoids', eds J, B. Harborne, T. J, Mabry and H. Mabry, pp. 1096-1129. (Chapman and Hall: London.) Taufeeq, H. M,, Fatma, w., Ilyas, M,, and Rahman, W. (1978). Biflavones from Cupressus lusitanica var. benthami. Ind. J. Chem. 16B, 655-7. Taufeeq, H. M., Mohd, F., and Ilyas, M. (1979). Biflavones from Cupressus govaniana. Ind. J, Chem. 17B, 535, Tissut, M,, and Rvenel, P, (1980). Repartition des flavonols dans l'epaisseur des feuilles de quelques vegetaux vasculaires. Phytochemistry 19, 2977-81. 112 Ueno, J. (1959). Some palynological observations of Taxaceae, Cupressaceae and Araucasriaceae. J. Institute Polytechnics, Osaka City University 10D, 75-87. Ueno, J. (1974). Some palynological observations on the family tree of gymnosperms. Jap. J. Paly11ology 14, 1-32. Van Campo-Duplan, M. (1953). Recherches sur la phylogenie des Cupressacees d'apres leurs grains de pollen. Trav. du Laboratoire forestier de Touleuse Tome II, Etudes Dendrologiques 4, 1-20. Van Campo, M., and Lugardon, B. (1973). Structure grenue infratectale de l'ectexine des pollens de quelques gymnospermes et angiospermes. Pollen et Spores 15, 171-87. Varshney, A. K., Rahman, W., Okigawa, M., and Kawano, N. (1973). Robustaflavone-the first member of a new series of biflavones. Experientia 29, 784-6. Venning, J. (1979). Character variation in Australian species of Callitris Vent. (Cupressaceae). Ph.D. thesis, University of Adeliade. Vierhapper, F. (1910). Entwurf eines neven systems der coniferen. Abhandl. d. Zool.-Botan. Ges. Wien., Bd. 5, Heft 4. Walker, J. W. (1976). Evolutionary significance of the exine in the pollen of primitive angiosperms. In 'The Evolutionary Significance of the Exine', eds I. K. Ferguson and J. Muller, pp. 251-91. (Academic Press: London.) 113 Wallace, J. W., and Markham, K. R. (1978). Apigenin and amentoflavone glycosides in the Psilotaceae, and their phylogenetic significance. Phytochemistry 17, 1313-17. Wannan, B. s., Waterhouse, J. T., Gadek, P. A., and Quinn, C. J. (1985). Biflavonyls and the affinities of Blepharocarya. Biochem. Syst. Ecol. 13, 105-8. Ward, R. S., and Pelter, A. (1974). The analysis of mixtures of closely related naturally-occurring organic compounds using high performance liquid chromatography. J. Chromatogr. Sc. 12, 570-4. Waterman, P. G., and Crichton, E. G. (1980). Xanthones and biflavonoids from Garcinia desivenia stem bark. Phytochemistry 19, 2723-6. WHO (1979). Ultraviolet radiation. World Health Organization, Geneva, Environmental Health Criteria 14. Wollenweber, E. (1982). Flavonoid aglycones as constituents of epicuticular layers on ferns. In 'The Plant Cuticle', eds D. F. Cutler, K. L. Alvin, and C. E. Price, pp. 215-24. (Academic Press: London.) Wollenweber, E., and Dietz, V. H. (1981). Occurrence and distribution of free flavonoid aglycones in plants. Phytochemistry 20, 869-932. Wollenweber, E., and Kohorst, G. (1981). Epicuticular leaf flavonoids from Eucalyptus species and from Kalmia latifolia. z. Naturforsch. 36c, 913-15. 114 Wollenweber, E., Walter, J., and Schilling, G. (1981). New flavanones and chalcones from the farinose frond exudate of pitryogramma pallida. z. Pflanzenphysiol. 104.S., 161-8. Wulf, L. W., and Nagel, C. W. (1976). Analysis of phenolic acids and flavonoids by high-pressure liquid chromatography. J. Chromatgr. 116, 271-9. Yamazaki, T., and Takeoka, M. (1962). Electron-microscope investigations of the fine details of the pollen grain surface in Japanese gymnosperms. Grana Palynologica 3, 3-11. Zhaobang, S., Falshaw, C. P., Haslam, E., and Begley, M. J. (1985). A novel spiro-biflavonoid from Larve gmelini. J. Chem. Soc., Chem. Commun. 1985, 1135-7. APPENDIX. Location and collecting details of voucher specimens are given below, listed in alphabetical order of genera. Abbreviations of herbaria follow Index Herbariorum (Staflau 1981). Actinostrobus acuminatus Parl. UNSW 11567, J. T. Waterhouse 8.viii.1981, Badgingara National Park, WA. A. pyramidalis Miq. UNSW Symon 1.ii.1982, cult. Waite Arboretum 632, SA; UNSW 17594, Gadek 2.viii.1985, cult. NBG, ACT. Agathis robusta (C. Moore ex F. Muell.) F. M. Bailey UNSW 12887, Gadek 8.ix.82, cult. UNSW, NSW. Araucaria bidwillii Hook. UNSW 6469, Puttock 31.i.1977, Bunya Mts, QLD. A. cunninghamii D. Don UNSW 7751, Waterhouse and Hindmarsh 13.x. 78, Ourimba Creek, NSW. Athrotaxis cupressoides D. Don. UNSW 4312, Quinn 2.i.1975, Pine Lake, TAS. A. selagenoides D. Don. UNSW 16513, Quinn 15.i.84, Cradle Mountain,TAS. Austrocedrus chilensis (D. Don) Florin and Boutelje NSW de Barba 980 23.ii.1946, SAm. Callitris canescens (Parl.) S. T. Blake UNSW 9289, Quinn and Findlay 10.v.1980, Parklea, SA. C. columellaris F. Muell. UNSW 9341, Quinn 2.ii.1980, Parkes, NSW; UNSW 9287, Quinn and McDonald 7.v.1980, Euston, NSW. C. endlicheri (Parl.) F. M. Bailey UNSW 9339, Quinn 2.ii.1980, Molong, NSW; UNSW 9291, Quinn and McDonald 12.v.1981, Goolgowi, NSW. C. macleayana (F. Muell.) F. Muell. UNSW 7485, Quinn and Waterhouse 10.xii.1979, Wilson River, NSW (juvenile); UNSW 9239, Hindmarsh and Waterhouse 11.i.1980, Comboyne, NSW; UNSW 12864, Gadek 12.iv.1982, cult. RBG, NSW. C. muelleri (Parl.) F. Muell. UNSW 14289, Gadek 16.vi.1983, cult. UNSW, NSW. C. neocaledonica Dummer NSW de Laubenfels 8.x.1957, NC. C. oblonga Rich. UNSW Symon 1.ii.1982, cult. Waite Arboretum 1225, SA. C. preissii Miq. subsp. murrayensis J. Garden UNSW 9288, Quinn and McDonald 7.v.1980, Mildura, NSW; UNSW 12890, Gadek 8.ix.1982, cult. NSW. C. preisii subsp. verrucosa (A. Cunn. ex Endl.) F. Muell. UNSW 9290, Quinn and McDonald 12.v.1980, Walpeup, VIC. C. sulcata (Parl.) Schlachter NSW 28864 Hotchkiss 15.iii.1954, NC. Calocedrus decurrens (Torr.) Florin NSW, Parks 24251 viii.1943, Darlington, Del Norte County, USA; UNSW, Symon 1.ii.1982, cult. Waite Arboretum 1242. SA; UNSW 17582, Gadek 10.viii.1985, cult. Cooma, NSW. C. formosana (Florin) Florin UNSW 17552, Quinn 4.vi.85, cult. Kew, UK. C. macrolepis Kurz. UNSW 17555, Quinn 4.vi.1985, cult. Kew, UK. Cedrus deodara (Roxb. ex Lamb.) G. Don UNSW 14292, Gadek 16.vi.1983, cult. UNSW, NSW. Cbamaecyparis nootkatensis (D. Don) Spach NSW Calder, Parmelee &. Taylor 19471 26.vi.1956, Mt. Arrowsmith, Vancouver Isl., Canada. C. formosansis Matsumura NSW Wilson 9764 2.ii.1918 Gisan, Prov. Kagi, Formosa. C. thyoides (L.) Britten, Sterns and Poppenberg NSW, Lawrence and Dress 295 20.v.1948, Penn State Forest, New Jersey, USA. C. lawsoniana (A. Murray) Parl. 'Erecta' UNSW 7164 Quinn 26.ii.1980, cult. Sydney, NSW. C. obtusa (Sieb. et Zucc.) Endl. 1- UNSW 10341 Gadek 13.viii.1980, cult. RBG, NSW. C. pisifera (Sieb. et Zucc.) Endl. 'Squarrosa' UNSW 10342 Gadek 13.viii.1980, cult. RBG, NSW. Cryptomeria japonica (Linn. f.) D. Don UNSW 17539, Quinn 3.vi.1985, cult. Kew, UK. Cunninghamia konishii Hay. UNSW 14294, Gadek 18.x.1983, cult. RBG, NSW. C. lanceolata (Lamb.) Hook. f. UNSW 17541 Quinn 3.vi.1985, cult. Kew, UK. Cupressus sempervirens L. UNSW 10339, Gadek 13.viii.1980, cult. RBG, NSW. C. lusitanica Mill. UNSW 10336, Gadek 13.viii.1980, cult. RBG, NSW. C. arizonica Greene UNSW 10340, Gadek 13.viii.1980, cult. RBG, NSW; UNSW Gadek viii.1985, cult. UNSW, NSW. Cycas revoluta Thunb. UNSW 14284, Gadek 16.vi.1983, cult. UNSW, NSW. Dacrycarpus compactus (Wassch.) de Laub. Ethanolic extract ex. Markham. Decussocarpus falcatus (Thunb.) de Laub. UNSW 12892, Gadek 8.ix.1982, cult. UNSW, NSW. Diselma archeri Hooker fil. UNSW 14288, Gadek 15.ix.1981, cult. UNSW; NBG 701046 Gadek 2.viii.1985, cult. NBG, ACT. Fitzroya cupressoides (Mollina) Johnston NSW de Barba 1045 4.iii.1946, SAm. Fokienia hodginsii (Dunn.) Henry and Thomas NSW, Mcindoe 21.ii.1963, cult. RBG, NSW. Ginkgo biloba L. UNSW 14286, Gadek 16.vi.1983, cult. UNSW, NSW. Gnetum latifolium var. minus (Foxw.) Markgr. UNSW 13687, Gadek 12.iv.1984, cult. RBG, NSW. Halocarpus bidwillii (Hook. f. ex Kirk) Quinn UNSW Rattenbury 3, Rattenbury v.1970, Tongariro National Park, NZ. H. kirkii (F. Muell. ex Parl.) Quinn UNSW, Rattenbury 1972, Auckland, NZ. Hevea brasilensis Muell. Elmer 20022, Elmer 10.xii.1921, Sandakan, North Borneo. Juniperus bermudiana L. UNSW 10338, Gadek 13.vii:r.1980, cult. RBG, NSW. J. californica Carr. NSW, Clokey 7823 27.vii.1938, Charleston Mts., Nevada, USA. J. chinensis L. UNSW, Symon 1.ii.1982, cult. Waite Arboretum 1273A, SA. J. communis L. UNSW 10343, Martin 7.vi.1961, Kamloops, British Colombia, Canada; UNSW, Symon 1.ii.1982, cult. Waite Arboretum 744, SA. J. conferta Parl. UNSW 10335, Gadek 13.viii.1980, cult. RBG, NSW; UNSW 14290, Gadek 16.vi.1983, cult. UNSW, NSW. J. deppeana Steud. UNSW, Symon 1.ii.1982, cult. Waite Arboretum 1273, SA. J. drupacea Labillardiere NSW, Hartfield ii.1899, cult. RBG, NSW. J. excelsa Bieb. UNSW, Symon 1.ii.1982, cult. Waite Arboretum 1272, SA. J. foetidissma Wild. NSW, Waleres i.194 7, Troodes Forest, Cyprus. J. monosperma (Engel.) Sarg. UNSW, Symon 1.ii.1982, cult. Waite Arboretum 1278A, SA. J. oxycedrus L. NSW, Ferguson 2924 ll.ii.1971, Sierra des Mos., Spain; UNSW, Symon 1.ii.1982, cult. Waite Arboretum 740, SA. J. procera Hochst. UNSW, Symon l.ii.82, cult. Waite Arboretum 742, SA. J. virginiana L. UNSW, Symon l.ii.1982, cult. Waite Arboretum 570, SA; UNSW, Symon l.ii.1982, cult. Waite Arboretum 617, SA. Lagarostrobos colensoi (Hook.) Quinn UNSW Rattenbury 4, Rattenbury 1972, Tongariro National Park, NZ. Lepidozamia peroffskyana Regel UNSW 14285, Gadek 16.vi.1983, cult. UNSW, NSW. Libocedrus bidwillii Hook. f. L. plumosa (D. Don) Sargent NSW Petrie vi.1910, NZ. L. yateensis Guill. NSW de Laubenfels 4.xii.1957, NC. Neocallitropsis pancheri (Carriere) de Laubenfels CANB Hartley 15068 23.xi.1979, NC; UNSW McPherson 4601 26.vi.1982, La Madeleine River, NC. Papuacedrus papuana (F. Muell.) Li UNSW 4206 Quinn 24.vi.1974, PNG; UNSW 4213 Quinn 24.vi.1974, PNG (seedling). P. torricellensis (Diels) Li NSW van Royan NGF 18250 6.ix.1963 PNG. Pilgerodendron uviferum (D. Don) Florin NSW Sargent 21.i.1905, SAm. Pinus radiata D. Don UNSW 3106, Gadek cult. UNSW, NSW. Platycladus orientalis (L.) Franco 'flagelliformis' UNSW 10337, Gadek 13.viii.1980, cult. RBG, NSW; NSW 12.i.1922, cult. Beauport Park, England; UNSW, Symon 1.ii.1982, cult. Waite Arboretum 551, SA. Podocarpus elatus R. Br. ex Endl. UNSW 14287, Gadek 16.vi.1983, cult. UNSW, NSW. Psilotum nudum (L.) Griseb. UNSW 14283, Gadek 16.vi.1983, cult. UNSW, NSW. Sciadopitys verticillata (Thunb.) Sieb. et Zucc. UNSW 17540, Quinn 3.vi.1985, cult. Kew, UK. Sequoia sempervirens (Lamb.) Endl. UNSW, Evans, cult. RBG, NSW. Taiwania cryptomerioides Hay. NSW Wilson 9836 4.ii.1918, Arisan, Prov. Kagi, Formosa; UNSW 17553, Quinn 4.vi.1985, cult. Kew, UK. Taxodium distichum (L.) Rich. UNSW 14291, Gadek 16.vi.1983, cult. UNSW, NSW; var. imbricarium UNSW 15994, Gadek 28.v.1985, Orlando, USA. Taxus baccata var. fastigiata (Lindl.) Louden UNSW Bruhl vii.1983, cult. NSW. Tetraclinis articulata (Vahl.) Masters UNSW Gadek 20.v.1981, cult. NSW. Thuja koraiensis Nakai NSW Headfort 15.viii.1940, cult. Kells, Co. Meath, Ireland. T. occidentalis L. 'pyramidalis' UNSW 10334, Gadek 13.viii.1980, cult. RBG, NSW. T. plicata D. Don NSW Metcalf viii.1920, Nelson, British Columbia, Canada; NSW Calder, Parmelee & Taylor 18552 7.vii.1956, Bella Coota, British Columbia, Canada. T. standishii (Gordon) Carr. NSW O'Byrne 6 23.ix.1949, cult. RBG Kew, England. Thujopsis dolobrata (L.f.) Sieb. et Zucc. NSW, Togasi 24.viii.1956, Mt. Zaozan, Echigo, Japan. Widdringtonia cedarbergensis Marsh UNSW Symon 1.ii.1982, cult. Waite Arboretum 1275A, SA; UNSW Symon 1.ii.1982, cult. Waite Arboretum 1231, SA. W. nodiflora (L.) Powrie UNSW Symon 1.ii.1982, cult. Waite Arboretum 1283A, SA; NSW Stapf 989 vi.1920, SAf; NSW Darren-Smith 10.viii.1945, cult. UK. Abbreviations: RBG, Royal Botanic Gardens, Sydney; NBG, National Botanic Gardens; ACT, Australian Capital Territory, Australia; NSW, New South Wales, Australia; VIC, Victoria, Australia; SA, South Australia, Australia; TAS, Tasmania, Australia; WA, Western Australia, Australia; NC, New Caledonia; NZ, New Zealand; PNG, Papua New Guinea; SAm, South America; SAf, South Africa; UK, United Kingdom; USA, United States of America. Australian Journals of Scientific Research These offprints are sent to you with the compliments of the authors and publisher of the Australian Journals of Scientific Research. The Journals are published by the Commonwealth Scientific and Industrial Research Organization with the cooperation of the Australian Academy of Science. Annual Subscription Current Titles Journals Microfiche Australian Journal ofAgricultural Research $90 $30 Australian Journal of Biological Sciences $80 $25 Australian Journal of Botany (with Supplements) $90 $30 (without Supplements) $80 $25 Australian Journal of Chemistry $230 $75 Australian Journal of Marine and Freshwater Research $100 $35 Australian Journal ofPhysics $120 $40 Australian Journal of Plant Physiology $90 $30 Australian Journal of Soil Research $50 $15 Australian Journal ofZoology (with Supplements) $120 $40 (without Supplements) $100 $30 All inquiries should be addressed to The Editor-in-Chief, CSIRO 314 Albert Street East Melbourne, Victoria, Australia 3002 Aust. J. Bot., 1984, 32, 15-31 Localization of the Biflavonoid Fraction in Plant Leaves, with Special Reference to Agathis robusta (C. Moore ex F. Muell.) F. M. Bail. P. A. Gadek, C. J. Quinn and A. E. Ashford School of Botany, University of New South Wales, P.O. Box I, Kensington, N.S.W. 2033. Abstract Aluminium chloride-induced fluorescence was used to localize biflavonoids in fresh leaf sections of Agathis robusta. This method indicates that the biflavonoids are confined to the outer periclinal wall and anticlinal walls of the epidermal cells. This was confirmed by extraction and chromatographic analysis of epidermal peels, cuticular scrapings and middle leaf tissue fractions. A survey of representatives of the Psilotales, and of all orders of the gymnosperms using aluminium chloride-induced fluorescence, indicates that localization ofbiflavones in the cuticle is a general feature ofbiflavonoid-containing plants. Members of the Pinaceae and Gnetales, in which biflavonoids have not been found, show no such fluorescence in the cuticle. The possible functional roles ofbiflavonoid accumulation are discussed. It is postulated that biflavones serve a protective role against invasion of the leaf by microorganisms and/or attack by leaf-eating insects. Introduction Biflavones are dimeric flavonoids that are a characteristic component of the leaves of gymnosperms, with the notable exception of the Pinaceae and Gnetales. They are also known to occur in a wide range of other vascular plants and have been recorded from such diverse parts of the plant as bark, heartwood, roots, stamens, fruits and testa (Geiger and Quinn 1975, 1982; Gadek 1982). The cellular and subcellular localization of biflavonoids has not previously been studied; such information should contribute to an understanding of the functional significance of their accumulation by some plants, which at present remains the subject of pure speculation. It is not surprising that the subcellular localization of biflavones has not previously been described, since most preparative techniques employed in histochemical or microscopic studies involve soaking in aqueous or alcoholic solutions that will leach many phenolics from plant tissue. Recently, techniques employing fluorescence microscopy have successfully been used to localize bound phenolic acids in plant cell walls (Fulcher et al. 1972; Harris and Hartley 1976, 1980; Hartley and Harris 1981 ). Flavonoids have been localized in the u.v. absorbing regions of petals by induced fluorescence, using the intense colour change of certain flavonoids produced by alkaline conditions (Brehm and Krell 1975). One method used by Brehm and Krell (l 975) was to mist thin sections prepared by freeze microtoming with a l % aqueous aluminium chloride solution, and then to observe the induced yellow fluorescence of the flavonoids under u.v. microscopy. Biflavones also form an acid-stable complex with aluminium chloride, inducing a strong deep yellow fluorescence. This paper reports on the application of a development of these techniques for the localization of the biflavonoid fraction of the leaves of a wide range of plants in which they are known to be accumulated. 0067-l 924/84/010015$02.00 Biflavonoids in Agathis robusta Leaves 17 1a 1b Fig. la. Leaf margin of Agathis robusta as seen in transverse section. c, Small mesophyll cells adjacent to vascular bundle (see text); e, epidermis plus cuticle; f, thick-walled non-lignified cells (? fibres); h, hypodermis; p, parenchyma; ph, phloem; pm, palisade mesophyll; r, resin canal; sm, spongy mesophyll; t, transfusion tracheids; x, xylem. Scale: 250 µm. Fig. lb. Diagram illustrating the epidermal and hypodermal wall system of the adaxial surface of the leaf of Agathis robusta, as seen in transverse section. e, Epidermal cell; h, thin-walled hypodermal cell; hf, hypodermal fibre. The histochemically recognizable regions are: 1, epicuticular wax; 2, lightly stained in Sudan black B but unstained with all other stains-interpreted as the cuticle proper; 3, very lightly stained as for lignified walls but is intensely black with Sudan black 8-interpreted as cutinized cell wall; 4, reacts with various stains which normally stain unlignified cell walls but is unstained by Sudan black 8-interpreted as uncutinized cell wall. Scale: 20 µm. 18 P. A. Gadek e1 al. 3 6 7 'I Figs 2-7. Light microscopy of adaxial region of Agathis robusta leaves in transverse section. Scales: 20 µm. Fig. 2. GMA-embedded section of a mature leaf under half-crossed polarizing filters showing crystals as white areas. Fig. 3. Fresh section of a mature leaf stained with PAS reaction. Epicuticular deposits on the surface of the leaf are PAS-positive. Anticlinal flange and cuticle proper (region 2 in Fig. lb) are PAS-negative, while the cutinized wall (region 3) stains light pink. Fig. 4. Wax-embedded section of an immature leaf (see text) stained in safranin/ fast green. The inner region of the outer periclinal wall of the epidermis (region 4 in Fig. l b) is stained dark green. The cuticle proper (region 2 in Fig. lb) is devoid of crystals. Biflavonoids in Agathis robusta Leaves 19 are concentrated will be characterized by an intense yellow-induced fluorescence under the above system of illumination. That this AIC1 3-induced fluorescence is due to the presence of biflavones must, however, be confirmed by extraction of the particular tissue fraction and identification of the compounds present. Extraction of Bif[avones Ethanolic extracts of whole leaves, adaxial epidermal peels, middle-leaf tissue (i.e., leaf with both epidermises removed) and waxy scrapings from the adaxial surface were prepared by soaking in 70% ethanol for 24 h. Eluants were dried, taken up again in a small volume of 70% ethanol and subjected to two-dimensional chromatography on paper using tertiary butyl alcohol : acetic acid: water (3: 1 : I, BAW) followed by 15% acetic acid, and one-dimensional chromatography on aluminium-backed precoated silica gel plates using benzene: pyridine: formic acid ( 100: 20: 7, BPF). Papers were viewed under u.v. with and without ammonia fumes, while plates were viewed under u. v. before and after spraying with 5% AIC1 3 in 95% ethanol. The raw extract of each leaf tissue fraction as well as ethanolic extracts of the major bands from unsprayed plates were permethylated using dimethyl sulfate and then compared chromatographically with standard permethyl ethers (Gadek and Quinn 1983). Analysis of the whole leaf extract was carried out using the methods described by Quinn and Gadek (1981 ). This method of combining chemical extraction and identification together with AlClrinduced fluorescence in fresh sections provides an easy means of localizing biflavones in the leaves of species in which they are accumulated. Results Leaf Anatomy of Agathis robusta The leaves of Agathis robusta are broad, flat and elliptical-lanceolate, with downcurved margins (Fig. la), and have a hard leathery texture. The lamina is devoid of a midrib, being traversed by many longitudinal veins that diverge from the base. The leaf has a dorsiventral anatomy, with a single layer of large palisade mesophyll cells, although many of these are irregularly subdivided by one or more transverse walls. The vascular bundles are embedded in the upper part of the spongy mesophyll and alternate with one, sometimes two, resin canals (Fig. la). Large, heavily lignified astrosclereids with radiating arms, similar to those described by Kausik ( 1976) for Agathis dammara (Lamb.) Rich., occur throughout the spongy mesophyll, with arms sometimes projecting into the palisade mesophyll. An arc of cells (fibres ?) with very thick non-lignified walls lies both above and below the vascular bundle, and these cells retain a prominent protoplast even in leaves in their second year on the tree. Transfusion tissue similar to that described in A. dammara (Kausik 1976) can be discerned lateral to the xylem and phloem. The adaxial epidermis contains a small number of stomata (Hyland 1977) and is separated from the mesophyll by a prominent hypodermis of 1-2 layers of non-lignified fibres, in which the lumen is almost occluded, and thinner-walled living cells (Fig. lb). Fibres predominate over thin-walled hypodermal cells by 3 : 2 in these layers and show a strong tendency to be clustered. The abaxial Fig. 5. Wax-embedded section of an immature leaf showing a stoma stained with safranin/fast green. The raised florin rings are clearly visible (arrows). Small crystals in the cuticle appear to be absent around the stoma. Fig. 6. GMA-embedded section of a mature leaf stained with toluidine blue. The cuticle proper and cutinized wall (regions 2 and 3 respectively in Fig. lb) are unstained. Three pits can be seen in the anticlinal wall of an epidermal cell. Note that the section has separated between the cuticle proper and the epicuticular wax deposits (arrowed). Fig. 7. GMA-embedded section of a mature leaf stained with Sudan black B. The innermost (uncutinized) layer of the outer periclinal and anticlinal walls of the epidermis (arrowed) are unstained (region 4 in Fig. I b), while the cuticle proper and the anticlinal flange (region 2) are less intensely stained than the cutinized cell walls (region 3). 20 P. A. Gadek et al. 10 Figs 8-13. Fluorescence microscopy of fresh, transections or epidermal peels from the adaxial surface of Agathis robusta leaves. Note that these black-and-white photographs of fluorescence tend to over represent some colours (e.g. light blue). Scales: 20 µm. Fig. 8. Transverse section (TS) of mature leaf, mounted in oil. The cuticle, outer periclinal and anticlinal walls of the epidermis are non-fluorescent. Epicuticular wax, inner periclinal wall of the epidermis and the primary walls of the hypodermis fluoresce light blue. Figs 9-13. Material mounted in A1Cl3• Fig. 9. TS of immature leaf. The anticlinal flange appears contiguous with a non-fluorescent surface layer (arrows), the cuticle proper (region 2 in Fig. 1b). Biflavonoids in Agathis robusta Leaves 21 epidermis contains numerous stomata and its associated hypodermis is much less regular, with thin-walled cells outnumbering non-lignified fibres by 2 : I. The downcurved leaf margin contains a mass of hypodermal tissue, with 15-30 fibres (Carr and Carr 1977), many of which are lignified by the second year. As described by Stockey and Taylor ( 1981 ), the stomata are sunken to the hypodermal level, opening into a pit, the sides of which are formed by four or sometimes five subsidiary cells. A protuberance of the subsidiary cells projects upwards around the sides of this pit (Fig. 5) to form a pronounced 'florin ring' (fig. 3 in Stockey and Taylor I 981 ). The stomata! pit is almost occluded by a wax plug (fig. 3 in Stockey and Taylor I 981 ). Examination of the adaxial epidermis with a number of histochemical stains allowed the recognition of four distinct layers in the outer periclinal wall of the epidermal cells. Three of these extended into the anticlinal walls (see Fig. lb). A narrow inner layer of the outer periclinal wall (region 4) was intensely stained (purple) with toluidine blue pH 4 · 4 (Fig. 6), (green) with safranin/fast green (Fig. 4), (red) with the PAS reaction (Fig. 3) and (black) with amido black, but was completely unstained with Sudan black B (Fig. 7). This layer was continuous around each epidermal cell and it is interpreted to be uncutinized cell wall. Abutting this was a broader relatively unstained region (region 3) which extended inwards along the full length of the anticlinal walls (Figs lb, 4 and 6). This region was only faintly stained with the PAS reaction (Fig. 3), amido black and toluidine blue (Fig. 6) but was deeply stained with Sudan black B (Fig. 7). These characteristics of region 3 are interpreted as indicating a cutinized cell wall. In GMA sections the boundary between regions 3 and 4 appeared sharp but with minute convolutions at the interface. Overlying the cutinized wall was a narrow layer of material that was completely unstained with PAS, toluidine blue or amido black, and was only lightly stained with Sudan black B (region 2). This layer extended into the central region of the anticlinal wall as an unstained flange that was quite distinct in sections stained with Sudan black B (Fig. 7). Region 2 was usually more pronounced in leaves in their first year on the tree. It appears to be the 'cuticle proper' of Von Mohl and Roelofsen (see Holloway 1982). The thin surface layer (region I) stained with Sudan black Band was PAS-positive in fresh material but not in GMA sections. It seems to correspond to the epicuticular wax layer of Martin and Juniper (1970). The cutinized wall and layers exterior to it are collectively referred to in this paper as the cuticle. Fine crystals, presumably of calcium oxalate (see Cookson and Duigan 1951), were a prominent feature of the cutinized wall and all walls of the epidermal cells on both surfaces of the leaf (Figs 2 and 4). However, the cuticle proper appears to be devoid of crystals, and in the immediate vicinity of the stomata these crystals were seen to be less frequent and sometimes absent from the epidermis as a whole (Fig. 5). Similar crystals were also prominent in the walls of the astrosclereids, and were sparsely scattered throughout the walls of the mesophyll. Fig. 10. TS of mature leaf. The non-fluorescent anticlinal flange (arrow) in the centre of each anticlinal wall of the epidermis is well shown. The fluorescence in the cuticle appears granular because of the non-fluorescent crystals embedded in it. Fig. 11. TS of mature leaf. Inner periclinal walls fluoresce blue in contrast to the intense deep yellow of the cuticle and anticlinal walls. The anticlinal flange (arrow) is non-fluorescent. Fig. 12. Surface view of the epidermis of a mature leaf at the edge of a near paradermal section through the outer periclinal wall. The cuticular flange is non-fluorescent but is crossed by fluorescent pit areas (arrowed). The outer region of the cutinized cell wall (top of figure) fluoresces deep orange while the deeper layers fluoresce yellow. Fig. 13. TS of stomata) chamber. The cuticle is continuous around the chamber, fluorescing deep yellow, while the guard cell protoplasts show an intense light blue fluorescence. 22 P. A. Gadek et al. Fluorescence Microscopy Autofluorescence Fluorescence microscopy of fresh sections mounted in oil showed intense red-orange fluorescence from the chloroplasts, particularly in the palisade mesophyll, and strong blue white fluorescence uniformly throughout the protoplasts of some smaller cells in the spongy mesophyll located immediately below the palisade layer and adjacent to the vascular bundles (Fig. la). In addition, there were abundant droplets of material which fluoresced light blue throughout the spongy mesophyll cells. The thickened walls of the guard cells, astrosclereids and xylem vessel elements showed strong light blue fluorescence, while the thin walls of the spongy mesophyll and the vascular tissue showed faint light blue fluorescence. The walls of the palisade mesophyll cells were non-fluorescent. In some sections the contents of the resin canals showed strong light blue fluorescence. The primary walls of both the non-lignified fibres and the thin-walled hypodermal cells showed a slightly deeper blue fluorescence, as did the inner periclinal wall of the epidermis (Fig. 8). Similarly, the epicuticular wax layer, when present on the sections, fluoresced light blue (Fig. 8). But both the anticlinal and the outer periclinal wall of the epidermis, including the cuticle, were non-fluorescent. Aluminium chloride-induced fluorescence In sections mounted in AIC1 3, an intense deep yellow fluorescence appeared in the outer periclinal wall (Figs 1b, 9-11 ). The anticlinal walls appeared three-layered in transverse section, the central region, which extended into the cuticle, showing no fluorescence (Figs I 0 and 11 ); it was clearly continuous with a non-fluorescent surface layer (region 2) of the cuticle (Fig. 9) in sections of I-year-old leaves. Region 2 was not clearly distinguishable from region 3 by its fluorescence in sections of older leaves, but the flange remained non fluorescent (Figs I Oand 11 ). The non-fluorescent flange of the anticlinal walls could also be discerned in epidermal peels mounted in AIC1 3, where the fluorescent cell walls of adjacent epidermal cells were Separated by a narrow non-fluorescent band (Fig. 12). Small 'bridges' of fluorescence were visible connecting adjacent cells, no doubt corresponding to the pit areas on the anticlinal walls (Figs 6 and 12). The fluorescence of the cuticle often appeared particulate (Fig. 10). Examination of the same section mounted in AIC1 3, firstly between crossed polarizing filters and then using fluorescence epi-illumination, revealed non-fluorescent crystals in the cuticle that were responsible for the overall particulate appearance of the fluorescence. The cuticle lining the stomata! chambers, although much thinner than elsewhere, still showed marked deep yellow fluorescence (Fig. 13); this fluorescent layer extended through the stomata! pore and faded out on the inner side of the guard cells. The subsidiary cells of the raised florin rings often contained large crystalline bodies of intense deep yellow fluorescence. The wax plugs of the stomata were non-fluorescent. A light yellow fluorescence was also visible in the primary wall of the hypodermal fibres, especially in those parts close to the epidermis. This fluorescence was sharply defined and clearly distinguishable from the deep yellow of the cuticle and the outer periclinal and anticlinal walls of the epidermis. The protoplasts of the small spongy mesophyll cells adjacent to the palisade layer and vascular bundles referred to above fluoresced an intense light straw colour under these conditions. Epidermal peels often tended to 'stain up' from the inner surface, i.e. from the epidermal cells. This probably reflects the rate of penetration of the stain and its hindrance by epicuticular wax on the leaf surface, rather than a concentration gradient of flavonoids in the cuticle, as scratches across the outer surface of the cuticle became fluorescent immediately on staining. There was no marked leaching of the yellow fluorescence into the mounting medium as occurred with substances showing light blue fluorescence in sections of other species (see below). Biflavonoids in Agathis robusta Leaves 23 De-waxed sections of wax-embedded material, when mounted in A1Cl 3, did not show the intense deep yellow fluorescence observed in fresh sections; presumably the flavonoid constituent of the cuticle and epidermal cell walls had been leached out during fixation and dehydration. GMA-embedded thin sections, however, did show considerable residual induced fluorescence. Chromatography of Extracts The results of the two-dimensional separations of the extracts of the various layers of the leaf are given in Table 1. Seven spots were repeatedly identified in the whole leaf extract. Spot 1 represented the chlorophylls, and its absence from the extracts of the adaxial epidermal peels indicates that these were reasonably free of palisade mesophyll tissue. Spot 4 appears to be characteristic of the cuticular scrapings but, although it was identified in the whole leaf extract, it was not detected in the extract of the adaxial epidermal peel. Spot 2 was present in all but the middle leaf extract, and in each case was by far the most prominent spot. It had a high RF in BA W, a low RF in 15% acetic acid, and under u. v. was dark-absorbing both with and without ammonia fumes, but showed strong deep yellow fluorescence after being sprayed with AlCh. These characteristics indicate that this spot contained the biflavone fraction. Table 1. Results obtained from two-dimensional paper chromatography of extracts of Agathis robusta leaves BA W, tertiary butyl alcohol: acetic acid: water; HOAc, 15% acetic acid. + Presence; - absence; ? trace Spot RF in RF in Colour under: Presence in extracts No. BAW HOAc U.V. U.V. U.V. Whole Cuticular Epidermal Middle +NH3 +AIC1 3 leaf scrapings peel leaf 90-100 0-10 Dark Red DarkA + + red 2 90-100 0-20 Dark Dark Intense + + + yellow 3 40-70 60-70 Dark Dull Faint + yellow yellow 4 40-80 70-90 Light Blue- Light + + purple green purple 5 60-80 90 None Bright None + + + purple 6 20 90-100 White Bright White + + + white 7 20-80 90-100 Light Bright Light + + + blue blue blue ASpot I obscured by reaction of spot 2. Analysis of the whole leaf extract revealed a complex mixture of agathisflavone and amentoflavone and their partial methyl ethers, robustaflavone and partial methyl ethers of cupressuflavone. The following partial methyl ethers were identified: 7-monomethyl amentoflavone, 4'-monomethyl amentoflavone, 7,4'-dimethyl amentoflavone, a trimethyl amentoflavone (either 4', 7",4"' or 7,4',4"'), 7-monomethyl agathisflavone, 7, 7"-dimethyl agathisflavone, 7-monomethyl cupressuflavone, 7, 7"-dimethyl cupressuflavone and 7,4',7"-trimethyl cupressuflavone. In addition, there was an undetermined dimethyl cupressuflavone and two partial methyl ethers of robustaflavone. There was also a trace amount of hinokiflavone or its methyl ether, as revealed by the presence of hinokiflavone pentamethyl ether in the permethylated raw extract. 24 P. A. Gadck et al. 19 - I ~-. . -~~'·~ Biflavonoids in Agathis robusta Leaves 25 Comparison of the one-dimensional separations in BPF revealed that both the epidermal and cuticular extracts contained the full range of biflavonoid bands obtained from the whole leaf extract. The results of the permethylations of each of the raw extracts are given in Table 2. Apart from the trace of hinokiflavone pentamethyl ether, all the permethyl ethers obtained from the whole leaf extract were also obtained from both the epidermal and cuticular extracts. Minor traces of two permethyl ethers were also found in the middle leaf extract (Table 2). Table 2. Occurrence of biflavone permethyl ethers in permethylated extracts of Agathis robusta leaves ++ Major band; + minor band; t trace detected; ? trace unable to be positively identified Permethyl ether Presence in extracts after permethylation Whole Cuticular Epidermal Middle leaf scrapings peel leaf Amentoflavone hexamethyl ether ++ + ++ ? Cupressuflavone hexamethyl ether ++ ++ ++ Agathisflavone hexamethyl ether ++ ++ ++ Robustaflavone hexamethyl ether + + + Hinokiflavone pentamethyl ether t ? Survey of Other Taxa A survey of the distribution of AIClrinduced deep yellow fluorescence in the leaves of other taxa yielded the observations set out below. Since histochemical studies were not carried out in all cases, the term 'outer wall' is used to describe the full width of the outer periclinal wall of the epidermis. Figs 14-21. Fluorescence microscopy of fresh sections of other taxa. Scales: 20 µm. Fig. 14. TS of stem of Psilotum nudum, in oil. The epicuticular wax layer fluoresces light blue but the outer wall is non-fluorescent. Fig. 15. As for Fig. 14, in A1Cl 3• The outer periclinal and anticlinal walls of the epidermis fluoresce an intense deep yellow. The inner periclinal walls fluoresce light blue. Note that the reduced exposure does not register the initial light blue autofluorescence because of the much greater intensity of the induced fluorescence. Fig. 16. TS of Callitris muelleri leaf, in A1Cl 3• The outer two-thirds of the 'outer wall' and the outer end of the anticlinal walls of the epidermis fluoresce an intense deep yellow. Fig. 17. TS of Ginkgo biloba leaf, in A1Cl 3• A thin surface layer of the 'outer wall' of the epidermis fluoresces intense yellow. The walls of the tracheids fluoresce light blue. Fig. 18. TS of Pinus radiata leaf, in oil. The primary walls of the epidermis and hypodermis fluoresce light blue, while the cuticle is non-fluorescent. Fig. 19. As for Fig. 18, in A1Cl 3• Epidermal cell walls and the primary walls of the hypodermal cells fluoresce an intense light yellow but the cuticle and anticlinal flanges remain non-fluorescent. Light yellow fluorescence is also visible in the protoplasts of the epidermis. (See note on exposure in legend to Fig. 15.) Fig. 20. TS of Cedrus deodara leaf, in oil. The thin layer of epicuticular wax can be seen to fluoresce light blue on the surface of the non-fluorescent cuticle. Some faint light blue fluorescence can be seen in the primary walls of the epidermal cells. Fig. 21. As for Fig. 20, in A1Cl 3. All walls of the epidermis fluoresce bright yellow and this fluorescence extends into the primary anticlinal walls of the hypodermis. The outer region of the 'outer wall' (? cuticle) remains non-fluorescent. (See note on exposure in legend to Fig. 15.) 26 P. A. Gadek et al. Psi/ota/es. The epidermis of both the scale leaves and stems of Psi/otum nudum (L.) Griseb. (Psilotales) showed deep yellow fluorescence throughout the outer wall and along almost the entire length of the anticlinal walls (cf Figs 14 and 15). The inner periclinal walls fluoresced light blue. Cycada/es. Both surfaces of the strongly dorsiventral pinnae of Cycas revo/uta Thunb. (Cycadaceae) showed bright yellow fluorescence in the inner quarter of the outer wall and deep yellow-brown fluorescence in the outer three-quarters. The pinnae of Lepidozamia peroffskyana Regel (Zamiaceae) showed deep yellow fluorescence in the outer three-quarters of the outer wall. Ginkgoa/es. The deciduous, fan-shaped leaves of Ginkgo bi/oba L. bore a superficial layer of intense yellow fluorescence (Fig. 17) which, while clearly present on both surfaces, was much more pronounced on the adaxial surface. These leaves were notable for a strong blue fluorescence, visible both with and without AICl 3, that appeared to occur widely in the mesophyll and rapidly leached into the mounting medium. It was not possible to localize the source of this material. Conifera/es. The flat leaves of Podocarpus e/atus R. Br. ex End!. and P. fa/catus R. Br. ex Mirb. (Podocarpaceae) showed a similar distribution of deep yellow fluorescence to that described for Agathis robusta, viz. throughout almost the full width of the outer wall and extending some distance in along the anticlinal walls of the epidermal cells. The appressed scale-leaves of Dise/ma archeri Hook. f, Ca/litris mue/leri (Par!.) F. Muell. and C. preissii subsp. murrayensis J. Garden (Cupressaceae) all showed an intense deep yellow fluorescence in the outer two-thirds of the outer wall, with a broad wedge of fluorescence projecting inwards towards each anticlinal wall (Fig. 16). In the Ca//itris species the fluorescence was noticeably less concentrated in the elongated decurrent leaf base than in the scale itself. In Dise/ma archeri, large non-fluorescent inclusions were a feature of the fluorescent layer. The spreading, awl-shaped leaves of Juniperus confer/a Par!. (Cupressaceae) showed a marked differentiation between the two leaf surfaces. The outward-facing abaxial surface, which was devoid of stomata, possessed a superficial layer that fluoresced bright yellow and often tended to separate from the leaf during sectioning. Beneath this were two distinct layers of less intense deep yellow fluorescence, and an inner non-fluorescent layer. All other walls of the epidermal cells were devoid of such fluorescence. The inward-facing adaxial surface, which bore numerous stomata, possessed a single, thin superficial layer of orange fluorescence covering an otherwise non-fluorescent epidermis. The outer wall of the deciduous leaves of Taxodium distichum (L.) Rich. (Taxodiaceae) bore a thin superficial layer of bright yellow fluorescent material. The needles of both Pinus radiata D. Don, and Cedrus deodara (Roxb.) G. Don (Pinaceae) showed strong light yellow fluorescence in the inner half of the outer wall, in the inner periclinal and anticlinal walls of the epidermis and in the primary walls of the hypodermal fibres (cf Figs 18 and 19, 20 and 21). In addition, there was an even light yellow fluorescence in the protoplasts of the epidermal cells. The outer half of the outer wall was in each case totally without fluorescence, and in Pinus radiata a narrow non fluorescent wedge (? cuticular flange) was visible between the anticlinal walls of the epidermal cells (Fig. 19). Taxa/es. The strongly dorsiventral, spreading leaves of Taxus baccata var. fastigiata (Lind!.) Loudon (Taxaceae) showed intense deep yellow fluorescence throughout the full width of the outer wall and also for a short distance in along the anticlinal walls. Gneta/es. The broad flat leaves of Gnetum latifolium var. minus (Foxw.) Markgr. (Gnetales) showed no fluorescence in the epidermal walls, but there was a bright yellow fluorescence within the protoplasts of scattered epidermal cells; this was particularly evident in epidermal peels. Examination of fresh sections of Cycas revoluta, Lepidozamia peroffskyana, Callitris mue/leri, Podocarpus elatus and Taxus baccata mounted in Sudan black B indicated that Biflavonoids in Agathis robusta Leaves 27 the extent of cutinization of the epidermal walls in each case correlated strongly with the distribution of the AlC1 3-induced deep yellow fluorescence. Discussion The outer periclinal wall of the epidermis of Agathis robusta is interpreted as comprising a narrow outer cuticle proper and a broad layer of cutinized epidermal cell wall, overlying a narrow inner layer of non-cutinized epidermal cell wall. This interpretation is consistent with that of a generalized cuticle as described by Holloway ( 1982). The observation of a sharp but minutely convoluted boundary between the uncutinized and cutinized regions of the wall suggests a complex interdigitation between the two layers similar to that postulated for Chamaecyparis lawsoniana by Oladele (1982). This is supported by scanning electron microscope studies of the inner surface of the cuticle of Agathis robusta after digestion of wall material (see figs 8 and 9 in Stockey and Taylor 1981 ). The staining reactions of the central layer of the anticlinal epidermal walls indicate it essentially consists of cuticular waxes. The region is therefore interpreted as the 'cuticular peg' (Holloway 1982) or 'anticlinal flange' (Oladele 1982). Such anticlinal flanges have been reported to extend the full depth of the epidermis in Callitris endlicheri (Parl.) F. M. Bail. (Oladele 1982), and are visible as projections of considerable size on the inner surface of the digested cuticles of Agathis robusta illustrated by Stockey and Taylor (1981). Small crystals are scattered throughout the leaf tissue, with marked concentrations in the cuticle and epidermal cell walls. They are reduced or absent around stomata. Concentrations of crystals, stated to be calcium oxalate, have previously been reported in Agathis cuticles (Cookson and Duigan 1951) and are known to occur in the cuticles of other gymnosperms (Johnsen 1963; Alvin and Boulter 1974; Oladele 1982). The biflavones identified in the leaves of Agathis robusta during this study agree well with those reported from other species in the family (Khan et al. 1971, 1972, Byas et al. 1977, 1978). Overall, the same four parental structures extracted from Araucaria rulei (Byas et al. 1977) are present in Agathis robusta, viz. amentoflavone, cupressuflavone, agathisflavone and robustaflavone, and in addition a minor trace of hinokiflavone was detected. By far the major proportion of the biflavonoid fraction consisted of partial methyl ethers of these parental structures. Observations on the AIC1 3-induced fluorescence in fresh leaf sections showed that the overwhelming proportion of the biflavonoid content is localized in the outer periclinal wall of the epidermis. This was confirmed by the isolation of all the major biflavonoid bands detected in the whole leaf extract from extracts of cuticular scrapings and adaxial epidermal peels, and a failure to extract significant amounts of biflavonoids from middle leaf tissue. The trace amounts of biflavones detected in the permethylated middle leaf extract may be due to the presence of small amounts of these compounds in other regions of the leaf. There is evidence to suggest that synthesis of some flavonoids is closely associated with chloroplast activity (Saunders and McClure 1976a, 1976b) and if this also applies to biflavones their presence in the middle leaf fraction would not be surprising. However, there is evidence to indicate that other flavonoids are synthesized and stored entirely within the epidermis (McClure 1975; Hrazdina et al. 1980). It is also possible that the trace of biflavones detected in the middle leaf extract is simply due to contamination by epidermal tissue. The close association between the epidermis and the bundles of hypodermal fibres made the clean separation of adaxial epidermal peels difficult, and of abaxial peels virtually impossible. Hence the inclusion of small fragments of epidermis in the middle leaf tissue fraction cannot be ruled out. The absence of biflavonoids from the anticlinal flange is surprising, since in previous studies no differences in staining properties between the cuticle proper and the flange have been detected (Holloway 1982). Although the flange is continuous with a cuticle proper that is also non-fluorescent in immature leaves, it does not appear to accumulate 28 P. A. Gadek et al. biflavonoids with age as the cuticle proper appears to do. This suggests that there is a differentiation between the cuticle proper and the flange in Agathis robusta. Oladele ( 1982) concludes that the 'almost ubiquitous occurrence of anticlinal flanges in well developed plant cuticles may indicate that they are a consequence of some physical relationship between neighbouring epidermal cells'. This implies complete equivalence in structure of the two regions. Further, no such differentiation between the flange and the cuticle proper was seen in any of the other species studied. The existence of this differentiation in Agathis robusta must therefore be seen as a specialization: a more detailed investigation of the fine structure of the epidermis and cuticle in this species is needed to clarify the point. The survey of other gymnosperm species known to contain biflavonoids revealed that, while there were individual differences in the extent of the AIC1 3-induced fluorescence in the outer periclinal and anticlinal walls of the epidermis, all were notable for the heavy concentration of this fluorescence in the 'outer wall'. The variation in the distribution of fluorescence is interpreted as simply reflecting the extent of cutinization of the walls. Sudan black B staining in several species indicated that the pattern oflipid deposition was strongly correlated with the distribution of the induced fluorescence. It is concluded, therefore, that the biflavonoid content of gymnosperm leaves is characteristically concentrated in the cuticle. This is also true for Psilotum nudum, a member of the primitive order of rootless plants, the Psilotales, which is also known to contain biflavonoids (Wallace and Markham 1978). In contrast, those species belonging to the Pinaceae and Gnetales showed no A!Clrinduced fluorescence in the cuticle; in all cases it was localized in the walls and protoplasts of the epidermal and hypodermal cells. This difference in distribution of induced fluorescence correlates with the known absence of biflavonoids in these taxa. The flavonoid content of the leaves of the Pinaceae has not been studied in great detail (Niemann 1979; Parker et al. 1979), and no attempt was made in this investigation to extract and identify the compounds responsible for the induced fluorescence. It is not possible to resolve the identity of individual flavonoids present in a complex mixture offlavonoids by using AIC13-induced fluorescence alone. It is perhaps not surprising to find lipophilic biflavonoid aglycones associated with the waxy cuticles of gymnosperm leaves. Free flavonoid aglycones have now been isolated from a wide range of plants (Wollenweber and Dietz 1981) and, whereas glycosidic flavones tend to be water-soluble and thus are commonly located in cell vacuoles (McClure 1975; Hrazdina et al. 1980, 1982; Tissut and Ravanel 1980), flavonoid aglycones are scarcely soluble in water and tend to become more lipophilic with increasing methylation (Wollenweber and Dietz 1981 ). Biflavone glycosides are unknown in the gymnosperms, although they are a minor component of the biflavonoid fraction of both Psilotum nudum and Tmesipteris tannensis (Spreng.) Bemh. (Wallace and Markham 1978). The occurrence of flavonoid aglycones in leaf tissue, therefore, tends to be associated with secretory structures or with the production of other lipophilic plant products: e.g. the secretions from glandular trichomes of some ferns and primroses (Chance and Amott 1981; Wollenweber and Dietz 1981; Wollenweber et al. 1981; Wollenweber 1982), epicuticular leaf waxes of some Eucalyptus species (Wollenweber and Dietz 1981; Wollenweber and Kohorst 1981 ), or in the external phenolic resin of Primula and Ma/us leaves (Baker 1982). The marked accumulation ofbiflavonoids in the leaves ofso many gymnosperms leads us to consider their adaptive significance. The function of the epidermal accumulation of some flavonoids, particularly those confined to vacuoles (e.g. the colour determining chalcones and anthocyanins), is related to the visual perception of flowers, fruits or spores by pollinators or dispersal agents (Harborne 1980). The u. v. absorbing properties of flavonoids are also well known (Mabry et al. 1970), and while these may also contribute to visual perception, Caldwell ( 1968, l 971 ), Lowry et al. ( 1980, 1983) and others have suggested that a significant role of epidermal flavonoids is, or has been, in affording some protection against damaging wavelengths of natural u.v. irradiation. Flavonoids (including Biflavonoids in Agathis robusta Leaves 29 biflavonoids) certainly absorb strongly those wavelengths (260-280 nm) that are most effective in producing nucleotide and protein damage (McClure 1975; WHO 1979). Present intensities of these wavelengths in the natural irradiation are quite low or absent (WHO 1979; Lowry et al. 1980), but they are thought to have been much higher at the time of the evolution of the first land plants and this has been postulated as the reason for the presence of flavonoids in the epidermis of such a broad range of the extant vascular flora (Lowry et al. 1980, 1983). Exposure of the plant to wavelengths in the region of 260-280 nm is also known to be very effective in inducing flavonoid accumulation (Caldwell 1968, 1971; McClure 1975). Further, flavonoids are relatively transparent to the longer wavelengths of light essential for photosynthesis (Lowry et al. 1980). Lowry et al. ( 1980) consider that epidermal flavonoids also served secondary roles in early land plants, particularly as defence mechanisms against predation. It certainly seems difficult to explain the almost universal occurrence oflarge concentrations ofbiflavonoids in the leaves of gymnosperms simply as the retention of an ancestorial u.v. screen that was adaptive to a previous, but no longer existing, environment. The occurrence of complex mixtures of biflavonoids based on a range of different skeletons is also difficult to account for on this basis, since there is virtually no difference in their absorption spectra. Both these features favour a strong adaptive role in modern plants. The reported occurrence of biflavonoids in, for example, the heartwood (Chen et al. 1975; Cotterill et al. 1977), bark (Waterman and Crichton 1980) and fruits (Chen et al. 1974; Lin and Chen 1974a, 1974b) of higher plants also indicates these compounds have now assumed other primary roles. Particular flavonoids are known to have antimicrobial properties (Harbome 1977, 1980), although the basis of this activity is not understood. Swain (cited in Wollenweber and Dietz 1981) has postulated that lipophilic flavonoids, particularly methylated flavonoids, offer protection against microorganisms because of their ability to penetrate membranes. A similar explanation has been postulated by O'Neill and Mansfield ( 1982) for isoflavones with antifungal properties. The latter authors report that complete methylation of the hydroxyl groups removed the antifungal activity; this may account for the rarity ofbiflavone permethyl ethers in nature (Geiger and Quinn 1982). Biflavonoids may also perform a protective role as a deterrent to leaf-eating organisms. There is evidence to support such a role for some flavonoids (Harborne 1980). The ability of living organisms to evolve to overcome such defences might explain the variety of biflavonoids found in many species (Geiger and Quinn 1982). Some such defensive mechanism seems probable as the primary role of the biflavonoids in the gymnosperms on the grounds of their localization in the leaf cuticle, the diversity of their structures and their occurrence in such a wide range of taxa. It is possible that the induced fluorescence in the epidermis and hypodermis of the Pinaceae and Gnetales may be found to be due to compounds performing a similar role to that of biflavonoids. Acknowledgments We acknowledge Dr L. A. S. Johnson, Director, for permission to sample Gnetum latifolium var. minus from the Royal Botanic Gardens, Sydney; and Mr J. Bruh! for material of Taxus baccata var. fastigiata. References Alvin, K. L., and Boulter, M. C. (1974). A controlled method of comparative study for Taxodiaceous leaf cuticles. Bot. J. Linn. Soc. 69, 277-86. Baker, E. A. (1982). Chemistry and morphology of plant epicuticular waxes. In 'The Plant Cuticle', eds D. F. Cutler, K. L. Alvin, and C. E. Price, pp. 139-66. (Academic Press: London.) Brehm, B. G., and Krell, D. (1975). Flavonoid localization in epidermal papillae of flower petals: a specialized adaptation for ultraviolet absorption. Science 190, 1221-3. 30 P. A. Gadek et al. Caldwell, M. M. ( 1968). Solar ultraviolet radiation as an ecological factor for alpine plants. Eco/. Monogr. 38, 259-68. Caldwell, M. M. (1971 ). Solar UV irradiation and the growth and development of higher plants. In 'Photophysiology', ed. A. C. Griese, pp. 131-74. (Academic Press: New York.) Carr, S. G. M., and Carr, D. J. (1977). Diagnostic anatomical characters of the leaves of three species of Agathis. Appendix in Hyland (1977). Brunonia 1, 103-15. Chance, G. D., and Amott, H.J. (1981). SEM of frond farina in Pityrogramma triangu/aris. Scanning Electron Microsc. 1981, 273-8. Chen, F., Lin, Y., and Liang, C. ( 1974). Biflavonyls from drupes of Rhus succedanea (Anacardiaceae). Phytochemistry 13, 276-8. Chen, F., Lin, Y., and Hung, S. (1975). Phenolic compounds from the heartwood of Garcinia multif/ora. Phytochemistry 14, 300-3. Cookson, I. C., and Duigan, S. L. ( 1951 ). Tertiary Araucariaceae from south-eastern Australia, with notes on living species. Aust. J. Sci. Res. (B) 4, 415-49. Cotterill, P. J., Scheinmann, F., and Puranck, G. S. (I 977). Phenolic compounds from the heartwood of Garcinia indica. Phytochemistry 16, 148-9. Feder, N., and O'Brien, T. P. ( 1968). Plant microtechnique: some principles and new methods. Am. J. Bot. 55, 123-42. Fisher, D. B. ( 1968). Protein staining of ribboned epon sections for light microscopy. Histochemie 16, 92-6. Fulcher, R. G., O'Brien, T. P., and Lee, J. W. (1972). Studies on the aleurone layer. I. Conventional and fluorescence microscopy of the cell wall with emphasis on phenol-carbohydrate complexes in wheat. Aust. J. Biol. Sci. 25, 23-34. Gadek, P. A. (1982). Biflavonoids from the seed testa of Cycadales. Phytochemistry 21, 889-90. Gadek, P. A., and Quinn, C. J. (I 983). Biflavones of the subfamily Callitroideae, Cupressaceae. Phytochemistry 22, 969-72. Geiger, H., and Quinn, C. J. (I 975). Biflavonoids. In 'The Flavonoids', eds J. B. Harborne, T. J. Mabry and H. Mabry, pp. 692-742. (Chapman and Hall: London.) Geiger, H., and Quinn, C. J. (1982). Biflavonoids. In 'The Flavonoids: Advances in Research', eds J. B. Harborne and T. S. Mabry, pp. 505-34. (Chapman and Hall: London.) Harborne, J. B. (I 977). Flavonoids and the evolution of the Angiosperms. Biochem. Sys/. Eco/. 5, 7-22. Harborne, J. B. ( 1980). Plant phenolics. In 'Encyclopedia of Plant Physiology. Vol. 8. Secondary Plant Products', eds E. A. Bell and B. V. Charlwood, pp. 329-95. (Springer-Verlag: Berlin.) Harris, P. J., and Hartley, R. D. (1976). Detection of bound ferulic acid in cell walls of the Gramineae by ultraviolet microscopy. Nature (London) 259, 508-10. Harris, P. J., and Hartley, R. D. { 1980). Phenolic constituents of the cell walls of monocotyledons. Biochem. Syst. Ecol. 8, 153-60. Hartley, R. D., and Harris, P. J. (1981). Phenolic constituents of the cell walls of dicotyledons. Biochem. Syst. £col. 9, 189-203. Holloway, P. J. ( 1982). Structure and histochemistry of plant cuticular membranes: an overview. In 'The Plant Cuticle', eds D. F. Cutler, K. L. Alvin and C. E. Price, pp. 1-32. (Academic Press: London.) Hrazdina, G., Alscher-Herman, R., and Kish, V. M. (I 980). Subcellular localization of flavonoid synthesizing enzymes in Pisum, Phaseo/us, Brassica, and Spinacia cultivars. Phytochemistry 19, 1355-9. Hrazdina, G., Marx, G. A., and Hoch, H. C. ( 1982). Distribution of secondary plant metabolites and their biosynthetic enzymes in pea (Pisum sativum L.) leaves. Plant Physiol. 70, 745-8. Hyland, B. P. M. ( 1977). A revision of the genus Agathis (Araucariaceae) in Australia. Brunonia 1, 103-15. Ilyas, M., Seligmann, 0., and Wagner, H. ( 1977). Biflavones from the leaves of Araucaria rulei F. Muell. and a survey of biflavonoids of the Araucaria genus. Z. Naturforsch. 32c, 206-9. Ilyas, N., Ilyas, M., Rahman, W., Okigawa, M., and Kawano, N. (1978). Biflavones from the leaves of Araucaria excelsa. Phytochemistry 17, 987-90. Johnsen, T. N., Jr (1963). Anatomy of scalelike leaves of Arizona junipers. Bot. Gaz. (Chicago) 124, 220-4. Kausik, S. B. (I 976). A contribution to the foliar anatomy of Agathis dammara, with a discussion on the transfusion tissue and stomata! structure. Phytomorpho/ogy 26, 263-73. Biflavonoids in Agathis robusta Leaves 31 Khan, N. U., Ansari, W. H., Usmani, J. N., Ilyas, M., and Rahman, W. (1971). Biflavonyls of the Araucariales. Phytochemistry 10, 2129-31. Khan, N. U., Ilyas, M., Rahman, W., Mashima, T., Okigawa, M., and Kawano, N. (I 972). Biflavones from the leaves of Araucaria bidwillii Hooker and Agathis alba Foxworthy (Araucariaceae). Tetrahedron 28, 5689-95. Lin, Y., and Chen, F. ( 1974a). Agathisflavone from the drupes of Rhus succedanea. Phytochemistry 13, 657-8. Lin, Y., and Chen, F. ( 1974b). Robustaflavone from the seed kernels of Rhus succedanea. Phytochemistry 13, 1617-19. Lowry, J. B., Lee, D. W., and Hebant, C. (I 980). The origin of land plants: a new look at an old problem. Taxon 29, 183-97. Lowry, J. B., Lee, D. W., and Hebant, C. (I 983). The origin of land plants: a reply to Swain. Taxon 32, 101-3. Mabry, T. J., Markham, K. R., and Thomas, M. B. (I 970). 'The Systematic Identification of Flavonoids.' (Springer-Verlag: Berlin.) Martin, J. T., and Juniper, B. E. ( I 970). 'The Cuticles of Plants.' (Edward Arnold: London.) McClure, J. W. (1975). Physiology and functions offlavonoids. In 'The Flavonoids', eds J.B. Harborne, T. J. Mabry and M. Mabry, pp. 970-1055. (Chapman and Hall: London.) Neimann, G. J. (1979). Some aspects of the chemistry of Pinaceae needles. Acta. Bot. Neer!. 28, 73-88. O'Brien, T. P., and McCully, M. E. (1981). 'The Study of Plant Structure. Principles and Selected Methods.' (Termarcarphi Pty Ltd: Melbourne.) Oladele, F. A. (1982). Development of the crystalliferous cuticle of Chamaecyparis lawsoniana (A. Murr.) Parl. (Cupressaceae). Bot. J. Linn. Soc. 84, 273-88, O'Neill, T. M., and Mansfield, J. W. ( 1982). Antifungal activity ofhydroxyflavans and other flavonoids. Trans. Br. Mycol. Soc. 79, 229-37. Parker, W. H., Maze, J., and McLachlan, D. G. (I 979). Flavonoids of Abies amabilis needles. Phytochemistry 18, 508-10. Pearse, A. G. E. ( 1968). 'Histochemistry, Theoretical and Applied.' Vol. I, 3rd edn. (Churchill: London.) Quinn, C. J ., and Gadek, P. A. (1981 ). Biflavones of Dacrydium sensu. Jato. Phytochemistry 20, 677-82. Saunders, J. A., and McClure, J. W. (1976a). The distribution offlavonoids in chloroplasts of twenty five species of vascular plants. Phytochemistry 15, 809-10. Saunders, J. A., and McClure, J. W. (1976b). The occurrence and photoregulation of flavonoids in Barley plastids. Phytochemistry 15, 805-7. Stockey, R. A., and Taylor, T. N. (1981). Scanning Electron Microscopy of epidermal patterns and cuticular structure in the genus Agathis. Scanning Electron Microsc. 1981, 207-12. Tissut, M., and Ravanel, P. ( 1980). Repartition des flavonols dans l'epaisseur des feuilles de quelques vegetaux vasculaires. Phytochemistry 19, 2977-81. Varshney, A. K., Rahman, W., Okigawa, M., and Kawano, N. (1973). Robustaflavone-the first member of a new series of biflavones. Experientia 29, 784-6. Wallace, J. W., and Markham, K. R. ( 1978). Apigenin and amentoflavone glycosides in the Psilotaceae, and their phylogenetic significance. Phytochemistry 11, 1313-17. Waterman, P. G., and Crichton, E. G. (I 980). Xanthones and biflavonoids from Garcinia desivenia stem bark. Phytochemistry 19, 2723-6. WHO ( 1979). Ultraviolet radiation. World Health Organization, Geneva, Environmental Health Criteria 14. Wollenweber, E. (I 982). Flavonoid aglycones as constituents of epicuticular layers on ferns. In 'The Plant Cuticle', eds D. F. Cutler, K. L. Alvin and C. E. Price, pp. 215-30. (Academic Press: London.) Wollenweber, E., and Dietz, V. H. ( 1981 ). Occurrence and distribution of free flavonoid aglycones in plants. Phytochemistry 20, 869-932. Wollenweber, E., and Kohorst, G. (1981). Epicuticular leafflavonoids from Eucalyptus species and from Kalmia latifolia. Z. Naturforsch. 36c, 913-15. Wollenweber, E., Walter, J., and Schilling, G. (I 981). New flavanones and chalcones from the farinose frond exudate of Pityrogramma pallida. Z. Pflanzenphysiol. 104.S., 161-8. Manuscript received 16 June 1983, accepted 22 August 1983 Phytochemistry, Vol. 24, No. 2, pp. 267-272, 1985. 0031 -9422/85 $3.00 + 0.00 Printed in Great Britain. © 1985 Pergamon Press Ltd. BIFLA VONES OF THE SUBFAMILY CUPRESSOIDEAE, CUPRESSACEAE P. A. GADEK and c. J. QUINN School of Botany, University of New South Wales, Kensington, NSW, 2033, Australia (Received 8 May 1984) Key Word Index-Cupressoideae; Cupressaceae; leaves; chemotaxonomy; biflavones; amentoftavone; cupressu ftavone; hinokiflavone; taiwaniaflavone. Abstract-Thirty species, representing all eight genera of the subfamily Cupressoideae, were examined for biflavonoid content of the leafy twigs. The major biflavonoid constituents are based on amentoflavone, cupressuflavone and hinokiflavone. The affinities suggested by biflavonyl distribution do not correlate with the currently recognized tribal groupings. There is evidence of closer links between northern and southern hemisphere genera than would be expected on the basis of the presently recognized subfamilies. INTRODUCTION for the family, although as some authors have pointed out [ 5-7], many early reports on the biflavonyl content of The subfamily Cupressoideae sensu Li [ 1] includes all the conifer leaves were obviously incomplete, as only some of northern genera of the family Cupressaceae. Although the major biflavonyls were reported. Despite the small size Li's basic division of the family into two subfamilies is still of samples used in this ·study and the reliance on TLC generally followed, subsequent authors have revised the techniques, a comparison of our data with those of status and affinities of some species within each of the previous studies on the same species gives no indication of subfamilies [2, 3]. Three tribes are recognized within any consequent lack of sensitivity to minor fractions. the northern subfamily: Cupresseae Neger, including Cupressus torulosa D. Don and C. sempervirens were the Cupressus L. (ea 20 spp.), Chamaeeyparis Spach. (6 spp.) original source from which the 8,8" -linked biflavone, and Fokienia Henry & Thomas (1 sp.); Thujopsideae cupressuflavone, was first isolated [8]. The three species of Endlicher, including Thuja L. (5 spp.), Thujopsis (L.f.) Cupressus surveyed here, including a sample of C. semper Siebold & Zuccarini (1 sp.), Biota Endl. (1 sp.) and virens, contained cupressuflavone and the 3,8" -linked Caloeedrus Kurz. (3 spp.); and Junipereae Neger, contain amentoflavone as the major biflavonyls. A total of seven ing Juniperus L. (ea 60 spp.). species of this genus have now been examined for This paper reports on a survey of representative species biflavonyl content of the leaves. All are reported to of all eight genera in the Cupressoideae, and, together contain cupressuflavone, and all but one amentoflavone with a previous paper [ 4], completes a chemotaxo [7, 9-12]. The report of cupressuflavone alone in the nomic survey of biflavone patterns in the leaves of the leaves of C. arizoniea [12] appears likely to be an Cupressaceae. incomplete report, since these authors also reported cupressuflavone alone in C. goveniana. Two subsequent RESULTS analyses of the latter species [7, 11] have revealed the presence of the amentoflavone and hinokiflavone series in Investigations on many of the species were conducted addition to cupressuflavone. The hinokiflavone series, on small samples (20-50 g dry wt) obtained from her however, is of variable occurrence in Cupressus species, barium specimens, but comparisons with the results of having been reported only from C. goveniana, C. lusi larger analyses where abundant material was available taniea, C. funebris and C. torulosa [7, 9, 11]. Previous indicated that such samples were adequate for the iso reports of its absence from C. sempervirens [7, 10] were lation and characterization of the major biflavonyl con confirmed in this survey; neither was it found in C. glabra. stituents. In fact, permethylation of small-scale crude The previous report of the presence of a minor amount of extracts often allowed the detection of trace amounts of a hinokiflavone in C. lusitaniea [9] was also confirmed by biflavonyl series that could not be detected by TLC of a the detection of a trace of hinokiflavone pentamethyl larger-scale unmethylated extract of the same material. ether in the permethylated extract of this species. Hence, The biflavonyl patterns obtained are given in Table 1, the genus Cupressus is typified by a leafbiflavonyl pattern and the results of analyses of permethylated leaf extracts having major amounts of amentoflavone and cupressu from a broader range of species are given in Table 2. flavone, with hinokiflavone and some minor monomethyl ethers also often present. Chamaeeyparis, on the other hand, is chemically hetero DISCUSSION geneous. All four species analysed contain amentoflavone An examination of Tables 1 and 2 shows that typically and a range of its partial methyl ethers (at least some of the the major biflavonyls are derived from three parental latter as major constituents), and also hinokiflavone. C. structures: viz. amentoflavone, cupressuflavone and nootkatensis alone contains cupressuflavone. This species hinokiflavone. This is in agreement with previous reports has been reported to be phytochemically atypical of the 267 Biflavones of the subfamily Cupressoideae 269 Table 2. Permethyl ethers detected in permethylated leaf extract of Cupressaceae Tribe Taxon Am Cu Hi Tw Ro Ag C Cupressus sempervirens + + C C. lusitanica + + C C. glabra + + J Juniperus drupacea + + + J J. communis + + + J J. oxycedrus + + m J J. virginiana + + + J J. excelsa + + m J J. procera + + + J J. bermudiana + + + J J. conferta + + + J J. deppeana + + J J. monosperma + + + J J. chinensis + + + J J. californica + + + J J. foetidissima + m m T Calocedrus decurrens + + + + T Biota orientalis + m + T Thuja occidentalis + m + C Chamaecyparis nootkatensis + m + m T Thujopsis dolobrata + + m C F okienia hodginsii + + t C Chamaecyparis formosansis + + C C. thyoides + + C C. lawsoniana 'Erecta' + + C C. obtusa + + C C. pisifera 'Squarrosa' + + T Thuja koraiensis + + T T. standishii + + T T. plicata + + Am, Amentoflavone hexamethyl ether; Cu, cupressuflavone hexamethyl ether; Hi, hinokiflavone pentamethyl ether; Tw, taiwaniaflavone hexamethyl ether; Ro, robustaflavone hexamethyl ether; Ag, agathisflavone hexamethyl ether; +, major band; m, minor band; t, trace detected by TLC only; C, Cupresseae; J, Junipereae; T, Thujopsideae. [13, 15]. However, we have now determined that the species was included in our survey, and was found to unidentified permethyl ether (U2) previously reported in contain major amounts of both cupressuflavone and Neocal/itropsis pancheri (Carr.) de Laub. [ 4] was also amentoflavone. Since the data for the other three species taiwaniaflavone hexamethyl ether. The occurrence of this are drawn from the same report [21], it seems probable rare series of biflavonyls in three such widely placed that a careful re-examination of these species would also species poses a problem of interpretation. reveal cupressuflavone as a major biflavonyl. Typically, All the species of J uniperus surveyed contained a major cupressuflavone is present unmethylated, and there is only amount of amentoflavone as well as some cupressu one report, for J. recurva, ofa partial methyl ether (7,7" flavone; hinokiflavone derivatives were detected, at least in dimethylcupressuflavone) constituting the major the permethylated extract, in all but one. Few partial cupressuflavone component [23]. Interestingly, this di methyl ethers were detected. Twelve species of Juniperus methyl ether also constitutes the major cupressuflavone have previously been examined for leaf biflavones, in component in several callitroid genera [ 4]. several cases by more than one worker [10, 21-26]; J. drupacea is morphologically separated from other apparent contradictions in these reports appear to be Juniperus species by its broader leaves and larger cones mainly due to incomplete analyses. Thus, for J. communis, [2, 3], and has sometimes been separated from the Lamer-Zarawska [21] records cupressuflavone, amento remaining species at either the generic [1] or subgeneric flavone and 4' -monomethylamentoflavone, while Pascual [3] level. Its biflavonyl pattern is, however, indistinguish Teresa et al. [26] record cupressuflavone and hinoki able from those of the other J uniperus species surveyed, flavone. Our own study of this species reveals the presence which favours its placement with those species. of all three series of biflavones. Amentoflavone has been The presence of traces of other biflavonyls, particularly reported in all 12 species, and cupressuflavone in all but robustaflavone, was detectable in some species of all the four of them: viz. J. sabina L., J. squamata Buch. Ham., cupressoid genera except C upressus; a wider survey of that J. occidentalis Hook. f. and J. virginiana [21]. The last genus would be needed before any significance can be 270 P. A. GADEK and C. J. QUINN attached to this distribution. The occurrence of a trace of robustaflavone and its monomethyl ether. F okienia the agathisflavone series in three species of Juniperus also hodginsii and Chamaecyparis formosansis share the appears to us to be of little systematic significance. 4',4"'-dimethylamentoflavone (as also does C. nootkat The data in Tables 1 and 2 have been used to group the ensis) in addition to the more usual 7,4' -dimethyl ether. taxa as follows: The biflavonyl data presented in this and the previous 1. Cupressus and Juniperus-The biflavonyl pattern of survey [ 4] of the Cupressaceae show that the chemical these genera comprises major amounts of amento discontinuities do not correlate closely with the existing flavone and cupressuflavone; the more highly meth taxonomy, particularly at the tribal and subfamilial levels. ylated biflavones are typically absent. There is a marked similarity between some of the 2. Calocedrus-Assuming that the single species exam biflavonyl patterns in both subfamilies, suggesting closer ined is typical of the other two, this genus is distin affinities than is indicated by the present division into guished by the presence of taiwaniaflavone and 7",4"' northern and southern genera. A summary of groupings dimethylamentoflavone, both unique amongst the cup suggested by biflavonyl pattern is set out in Table 3. ressoid genera, as well as a major amount of The presence of the group 1 biflavonyl pattern in both cupressuflavone. Cupressus and Juniperus suggests an affinity between 3. Biota orientalis, Thuja occidentalis and Chamaecyparis these genera that is not indicated by the present tribal nootkatensis-These are characterized by the presence groupings. Tetraclinus articulata displays an identical of minor amounts of cupressuflavone in combination pattern to that of the group 1 genera, having major with hinokiflavone and amentoflavone and its partial amounts of cupressuflavone and amentoflavone, together methyl ethers. with a minor amount of 4"' -monomethylamentoflavone 4. The remaining taxa constitute a rather heterogeneous and a trace of a hinokiflavone derivative [ 4]. grot•p that is distinguished by the absence of cupressu The pattern in Fitzroya, Diselma and Widdringtonia is flavone. All are characterized by the presence of distinct from the above group, cupressuflavone being amentoflavone and varying numbers of its partial replaced by its dimethyl ether, and the latter two genera methyl ethers, while hinokiflavone or its partial methyl being typified by major amounts ofhinokiflavone and/or ethers are detectable at least in the permethylated amentoflavone partial methyl ethers. Austrocedrus, extracts. Thujopsis dolobrata and Chamaecyparis Libocedrus, Papuacedrus and Pilgerodendron show a close thyoides are characterized within the group by the similarity to the group 4 pattern, having a major amount presence of the 4' -monomethylamentoflavone in con of amentoflavone and trace amounts of hinokiflavone trast to the more usually occurring 4"' -monomethyl derivatives, but being devoid of cupressuflavone. ether, and the former species is further characterized by The presence of taiwaniaflavone and its partial methyl Table 3. Taxa grouped by biflavonyl pattern Group Tribe Taxa Pattern C Cupressus Major amounts of amentoflavone J Juniperus and cupressuflavone; methylated Te Tetrac/inus biflavonyls mostly absent 2 A Fitzroya Partially methylated biflavones con L Diselma spicuous; major amounts of amen L »iddringtonia toflavone and 7, 7"-dimethyl cupressuflavone 3 Tj Biota Major amounts of amentoflavone Tj Thuja occidenta/is and hinokiflavone, plus a minor C Chamaecyparis nootkatensis amount of cupressuflavone 4 Tj Thujopsis and remaining spp. of Amentoflavone partial methyl Thuja ethers conspicuous; variable C F okienia and remaining spp. of amounts of hinokiflavone; cup Chamaec yparis ressuflavone absent L Libocedrus L Papuacedrus L Austrocedrus L Pilgerodendron 5 A Actinostrobus Amentoflavone only, with cup A Callitris ressuflavone and hinokiflavone undetectable 6 Tj Calocedrus Characterized by the presence of L N eocallitropsis taiwaniaflavone A, Actinostrobeae; C, Cupresseae; J, Junipereae; L, Libocedreae; Te, Tetraclineae; Tj, Thujopsideae. Biflavones of the subfamily Cupressoideae 271 Table 4. Chromatographic and spectral data of permethylated biflavones R1 s• UV BPF BPEFD Fluoresc.t spectra (.l.~:f"):j: Emission (nm)§ Hexa-O-methylamentoflavone 0.37 0.40 Yellow 266,328 460 Hexa-O-methylcupressuflavone 0.41 0.45 Orange 268,324 470 Hexa-O-methyltaiwaniaflavone 0.42 0.58 Lt. blue 264, 324 427 Hexa-O-methylagathisflavone 0.46 0.51 Yellow 266, 322 Hexa-O-methylrobustaflavone 0.50 0.68 Lt. blue 263, 324 Penta-O-methylhinokiflavone 0.53 0.75 Lt. blue 265, 323 437 • R1 values are variable, but the relative positions of the per,nethyl ethers in each solvent are characteristic. t As observed on plates run in BPF and dried in a hood for ea 0.5 hr. The residual formic acid which remains on the plate will partially affect the fluorescence colour (see §). :j:Compounds purified using an RP C-18 HPLC column (updating the previously reported data [4]). §Emission maximum at 320 nm (uncorrected), in MeOH, characterizing the fluorescence of the free compounds (as opposed to the fluorescence of the partially protonated form illustrated by t). ethers as minor fractions in N eoca/litropsis suggests an Acknowledgements-We thank Mr. D. Symon of the Waite affinity with Calocedrus, though the absence of even trace Institute, University of Adelaide, South Australia for assistance amounts of cupressuflavone from the former clearly in obtaining plant material; Dr. L. A. S. Johnson, Director, for distinguishes it. Indeed, the atypical phyllotaxis and leaf permission to sample specimens held by the National Herbarium morphology of N eocal/itropsis leads one to question ofN.S.W. and the Royal Botanic Gardens; Dr. G. D. McPherson, whether it is correctly placed in the Cupressaceae. There is Herbarium, Missouri Botanical Gardens, for specimens of certainly nothing in the biflavonyl patterns to support the Neocallitropsis pancheri; and Dr. I. McFarlaile, School of contention of de Laubenfels that this monotypic genus is Biochemistry, University of New South Wales, for assistance closely related to Callitris neoca/edonica [27]. The fact with HPLC equipment. that taiwaniaflavone also occurs in the Taxodiaceae raises the suggestion of a close relationship between the REFERENCES Cupressaceae and Taxodiaceae that has been commented I. Li, H. (1953) J. Arnold Arbor. Harv. Univ. 34, 17. on by several authors [28-30]. This compound is not, 2. Dallimore, W. and Jackson, A. B. (1966) A Handbook of however, a common constituent in either family. We have Coniferae and Ginkgoaceae (Revised by Harrison, S. G.) 4th been unable to detect it in leaf extracts of Athrotaxis edn Edward Arnold, London. selaginoides, Sequoia sempervirens or Cunninghamia 3. Gaussen, H. (1968) Trav. Lab. F oresl. Toulouse Pt 2, F asc. I 0, lanceolata. Nor has cupressuflavone, which occurs in a 13, I. number of cupressaceous genera, been detected or re 4. Gadek, P.A. and Quinn, C. J. (1983) Phytochemistry 22,969. ported in any taxodiaceous species [5, 6]. Hence the 5. Geiger, H. and Quinn, C. J. (1975)The Flavonoids (Harborne, taxonomic significance of the distribution of taiwania J. B., Mabry T. J., and Mabry, H., eds.), pp. 692-742. flavone in these three diverse genera will only be revealed Chapman & Hall, London. by a detailed analysis of other data sources in order to 6. Geiger, H. and Quinn, C. J. (1982) The Flavonoids: Advances reassess their affinities properly. in Research (Harborne, J. B. and Mabry, T. J., eds.), pp. It is apparent that the affinities of species and genera 505-534. Chapman & Hall, London. suggested by their biflavonyl patterns highlight many 7. Natarajan, S., Murti, V. V. S. and Seshadri, T. R. (1970) inconsistencies in the present tribal and possibly familial Phytochemistry 9, 575. groupings, and underlie a need noted by other authors 8. Murti, V. V. S., Raman, P. V. and Seshadri, T. R. (1964) [2, 3] for a critical reappraisal of the taxonomy of the Tetrahedron Lellers 2995. family. A re-definition of the taxa, based on a reassessment 9. Taufeeq, H. M., Fatma, W., Ilyas, M., Rahman, W. and of a broad range of character-states, is currently under Kawano, N. (1978) Indian J. Chem. Sect. B 16, 655. way in this laboratory and will be reported elsewhere. 10. Lebreton, P., Boutard, B. and Sartre, J. (1978) Bull. Inst. Sci. 155. 11. Taufeeq, H. M., Mohd, F. and Ilyas, M. ( 1979) Indian J. Chem. Seel. B 17, 535. EXPERIMENT AL 12. Miura, H. and Kawano, N. (1968) J. Pharm. Soc. Jpn. 88, Details of voucher specimens are given in the Appendix. 1459. Extraction and identification of biflavonyls were carried out by 13. Lebreton, P. (1982) Candol/ea 37, 243. the methods described previously [ 4, 31]. Standards of agathis 14. Er 19. Kami!, M., Ilyas, M., Rahman, W., Hasaka, N., Okigawa, M. Symon 1. ii. 82, cult. Waite Arboretum 742, SA. J. bermudiana L. and Kawano, N. (1977) Chem. Ind. 160. UNSW 10338, Gadek 13. viii.1980, cult. RBG, NSW. J. conferta 20. Kami!, M., Ilyas, M., Rahman, W., Hasaka, N., Okigawa, M. Par!. UNSW 10335, Gadek 13. viii.1980, cult. RBG, NSW. J. and Kawano, N. (1981) J. Chem. Soc. Perkin Trans. 1, 553. deppeana Steud. UNSW, Symon I.ii. 1982, cult. Waite 21. Lamer-Zarawska, E. (1975) Pol. J. Pharmacol. Pharm. 27, 81. Arboretum 1273, SA. J. monosperma (Engel.) Sarg. UNSW, 22. Fatma, W., Taufeeq, H. M., Shaida, W. A. and Rahman, W. Symon I. ii.1982, cult. Waite Arboretum 1278A, SA. J. chinensis (1979) Indian J. Chem. Sect. B 17, 193. L. UNSW, Symon 1. ii.1982,cult. Waite Arboretum 1273A, SA. J. 23. Hameed, N., Ilyas, M., Rahman, W., Okigawa, M. and californica Carr. NSW, Clokey 7823 27. vii.1938, Charleston Kawano, N. (1973) Phytochemistry 12, 1494. Mts., Nevada, U.S.A. J.foetidissima Wild. NSW, Waleres i. 1947, 24. Ilyas, M., Ilyas, N. and Wagner, H. (1977) Phytochemistry 16, Troodes Forest, Cyprus. Calocedrus decurrens (Torr.) Florin 1456. NSW, Parks 24251 viii. 1943, Darlington, Del Norte County, 25. Pelter, A., Warren, R., Hameed, N., Ilyas, M. and Rahman, W. U.S.A. C. decurrens (Torr.) Florin UNSW, Symon 1. ii. 1982, cult. (1971) J. Indian Chem. Soc. 48, 204. Waite Arboretum 1242. SA. Biota orientalis (L.) End!. UNSW 26. Pascual Teresa, J. de, Barrero, A. F., Muriel, L., San Feliciano, 10337, Gadek 13. viii.1980, cult. RBG, NSW. B. orientalis (L.) A. and Grande, M. (1980) Phytochemistry 19, 1153. End!. NSW 12.i.1922, cult. Beauport Park, U.K. Thuja oc 27. de Laubenfels, D. J. (1972) in Flora de le Nouvelle Caledonie cidentalis L. UNSW 10334, Gadek 13. viii. 1980, cult. RBG, NSW. et Dependences, (Aubreville, A. and Leroy, J., eds.), Vol. 4, T. occidentalis L. UNSW, Symon 1. ii. I 982, cult. Waite pp. 144-164. Museum National d'Histoire Naturelle, Paris. Arboretum 551, SA. T. koraiensis Nakai NSW Headfort 28. Keng, H. (1975) Taxon 24, 289. 15. viii. 1940, cult. Kells. Co. Meath, Ireland. T. standishii 29. Eckenwalder, J. E. (1976) Madrano 23, 237. (Gordon) Carr. NSW O'Bytne 6 23. ix.1949, cult. RBG Kew, 30. Eckenwalder, J. E. (1976) Taxon 25, 337. U.K. T. plicata D. Don NSW Metcalfviii.1920, Nelson, British 31. Gadek, P.A. and Quinn, C. J. (1982) Phytochemistry 21,248. Columbia, Canada. T. plicata D. Don NSW Calder, Parmelee & 32. Wannan, B., Waterhouse, J. T., Gadek, P.A. and Quinn, C. J., Taylor 18552 7. vii.1956, Bella Coota, British Columbia, Canada. Biochem. Syst. Ecol. (in press). Thujopsis dolobrata (L.f.) Sieb. et Zucc. NSW, Togasi 33. Stafleu, F. A. (1981) Index Herbariorum Part 1. Dr. W. Junk, 24. viii. 1956, Mt. Zaoz.an, Echigo, Japan. Fokienia hodginsii The Hague. (Dunn.) Henry and Thomas NSW, Mcindoe 21.ii.1963, cult. RBG, NSW. Chamaecyparis nootkatensis (D. Don) Spach NSW Calder, Parmelee & Taylor 19471 26. vi. 1956, Mt. Arrowsmith, APPENDIX Vancouver ls!., Canada. C.formosansis Matsumura NSW Wilson Location and collecting details of voucher specimens are given 9764 2. ii.1918 Gisan, Prov. Kagi, Formosa. C. thyoides (L.) below. Abbreviations of herbaria follow Index Herbariorum Britten, Sterns and Poppenberg NSW, Lawrence and Dress 295 [33]. Cupressus sempervirens L. UNSW 10339, Gadek 20. v.1948, Penn State Forest, New Jersey, U.S.A. C. lawsoniana 13. viii. 1980, cult. RBG, NSW. C. lusitanica Mill. UNSW 10336, (A. Murray) Par!. 'Erecta' UNSW 7164 Quinn 26. ii.1980, cult. Gadek 13. viii.1980, cult. RBG, NSW. C. glabra Sudworth Sydney, NSW. C. obtusa (Sieb. et Zucc.) End!. UNSW !0341 UNSW 10340, Gadek 13. viii.1980, cult. RBG, NSW. Juniperus Gadek 13. viii.1980, cult. RBG, NSW. C. pisifera (Sieb. et Zucc.) drupacea Labillardiere NSW, Hartfield ii. 1899, cult. RBG, NSW. End!. 'Squarrosa' UNSW 10342 Gadek 13. viii.1980, cult. RBG, J. communis L. UNSW 10343, Martin 7. vi.1961, Kamloops, NSW. Taiwania cryptomerioides Hay. NSW Wilson 9836 British Colombia, Canada. J. communis L, UNSW, Symon 4. ii. 1918, Arisan, Prov. Kagi, Formosa. Athrotaxis cupressoides 1. ii. 1982, cult. Waite Arboretum 744, SA. J. oxycedrus L. NSW, D. Don. UNSW 4312, Quinn 2.i.1975, Pine Lake, Tas. Sequoia Ferguson 292411. ii.1971, Sierra des Mos., Spain. J. oxycedrus L. sempervirens (Lamb.) End!. UNSW, Evans, cult. RBG, NSW. UNSW, Symon l.ii.1982, cult. Waite Arboretum 740, SA. J. Cunninghamia konishii Hay. UNSW 14294, Gadek 18.x.1983, virginiana L. UNSW, Symon 1. ii. 1982, cult. Waite Arboretum cult. RBG, NSW. Abbreviations: RBG, Royal Botanic Gardens; 570, SA. J. virginiana L. UNSW, Symon 1. ii.1982, cult. Waite NSW, New South Wales, Australia; SA, South Australia, Arboretum 617, SA. J. excelsa Bieb. UNSW, Symon l.ii.1982, Australia; Tas, Tasmania, Australia; U.S.A., United States of cult. Waite Arboretum 1272, SA. J. procera Hochst. UNSW, America. Phytochem,istry, Vol. 22, No. 4, pp. 969-972, 1983. 003 I -9422/83/040969-04$03.00/0 Printed in Great Britain. © 1983 Pergamon Press Ltd. BIFLA VONES OF THE SUBFAMILY CALLITROIDEAE, CUPRESSACEAE P. A. GADEK and c. J. QUINN School of Botany, University of N.S.W., Kensington, N.S.W., 2033, Australia (Received 9 July 1982) Key Word Index-Callitroideae; Cupressaceae; leafy twigs; chemotaxonomy; biflavones; amentoflavone; cupressu flavone; hinokiflavone. Abstract-Twenty species, including representatives of all 11 genera of the Callitroideae, were examined for biflavonoid content of the leafy twigs. The major biflavonoids are based on amentoflavone, cupressuflavone and hinokiflavone. Their uneven distribution amongst the genera allows the distinction offive groups. These do not correlate strongly with currently recognized tribal groupings. The affinities of these genera are discussed. INTRODUCTION derivatives of which could not be isolated in the normal way. Most BPF bands were permethylated to determine The subfamily Callitroideae, according to Li [1 ], includes the parent biflavones present; often these bands contained all the southern genera in the family, as well as the derivatives of more than one parent biflavone. All bands monotypic northern genus Tetrac/inus Masters. Although were further refined on cellulose plates in BN, and the de Laubenfels [2] has expressed doubt as to the basis of fractions identified using co-chromatography with auth some recently defined genera, 10 southern genera are entic markers; a final permethylation was performed usually recognized [3-5]. Li divides them into two tribes: where sufficient material was available as confirmation of viz. Actinostrobeae containing Actinostrobus Miq., the parental biflavonyl structure. Permethyl ethers were Callitris Vent. and Fitzroya Hook.; and Libocedreae, identified by chromatographic comparisons with per which contains Neocallitropsis Florin, Widdringtonia methyl ethers obtained from authentic samples of bifla End!., Dise/ma Hook. f., Papuacedrus Li, Pilgerodendron vones, as well as by comparisons of UV fluorescence and Florin, Austrocedrus Florin and Boutelje, and Libocedrus UV spectra. End!. emend. Florin and Boutelje. The biflavonyl patterns obtained for the Callitroid Previous studies on the occurrence of biflavones from genera are given in Table 1. the subfamily Callitroideae have been confined to a single genus, Callitris [6-9]. These indicate that Callitris con tains a simple pattern of derivatives based on amento DISCUSSION flavone. The occurrence of hinokiflavone as a minor fraction has been reported in some studies [7, 9], but this C allitris, with 15 species, is the largest Callitroid genus. could not be confirmed in others [6, 8]. Biflavones have Nine species have now been examined for biflavone previously been shown to be taxonomically useful, being content and only one species does not conform to the generally uniform within species, but often highly variable pattern reported by Gadek and Quinn [6]; viz. amento between species or genera [ 6, 10-12]. It has been sug flavone as the major band with possible minor bands of gested that the affinities of the genera of this family, the 4"'- and 4"',7"-methyl ethers. C. neocaledonica is particularly between northern and southern Hemisphere exceptional in having major bands of all three biflavones, genera, may be indicated by biflavonyl distribution as well as the 4' -monomethyl ether. It is interesting to note [10,6]. that this species is also atypical of the genus in cone As part of a comprehensive study of the Cupressaceae, morphology. de Laubenfels [2] considers it a natural but representatives of the Callitroideae were sampled to specialized member of the genus, showing closest affinity obtain the pattern of occurrence of biflavones amongst to C. su/cata, the other New Caledonian species and the the genera. chemical evidence does not disagree with this view. Dise/ma contains the richest array of biflavones in the Callitroideae, including trace amounts of a fourth par RESULTS ental biflavone detectable only by permethylation of the Investigations on many of the species were made on crude extract and tentatively identified as robustaflavone. relatively small samples (20-50 g dry wt), but these proved This biflavone has previously been isolated from adequate for the isolation of the major biflavones and Juniperus phoenica, a member of the northern subfamily, their characterization by TLC and permethylation. Cupressoideae [13]. Permethylations were carried out at three stages to check Four of the five species of Widdringtonia have been initial identifications against authentic markers, and to examined. A chemical discontinuity has been detected, indicate the identity of parental biflavones, in various with W dracomantana and W juniperoides containing 7" samples. The crude extracts were permethylated to ident monomethylhinokiflavone but not 4"',7"-dimethyl ify the parent biflavonyls present. This allowed the amentoflavone, while in W whytei and W cupressoides the recognition of trace amounts of parent compounds, the situation is reversed. This difference coincides with a 969 970 P. A. GADEK and c. J. QUINN Table I. Occurrence and distribution of biflavones is the leaves of Callitroid species Biflavones Extracts Permethyl extract Tribe/Genus/Species 2 3 4 5 6 7 8 HAm HCu PH PR U! U2 Act. Actinostrobus acuminatus + + Act. A. pyramidalis + + m Act. Callitris macleayana + t + Act. C. oblonga + + m + Act. C. sulcata + m + Act. C. neocaledonica + + + + + m Lib. N eocallitropsis pancheri + + m m + + m Lib. Papuacedrus papuana + + m + + Lib. P. torricellensis + + + + Lib. Pilgerodendron uniferum + + + + Lib. Libocedrus yateensis + m + m Lib. L. p/umosa + m + m Lib. Austrocedrus chi/ensis + + + + Act. Fitzroya cupressoides + m + + + m Lib. H-iddringtonia dracomantana + + + m + m + Lib. W juniperoides + + + + + + + Lib. W whytei + + m m + + + + Lib. W cupressoides + m + + + + + + Lib. Diselma archerii + + + + + + + + Tet. Tetraclinus articulata + m + + + Key: Act.= Actinostrobeae; Lib.= Libocedreae; Tet. = Tetraclineae. I, Amentoflavone; 2, 4"'-monomethylamentoflavone; 3, 4'- monomethylamentoflavone; 4,4"', 7"-dimethylamentoflavone; 5, hinokiflavone; 6, 7"-monomethylhinokiflavone; 7, cupressuflavone; 8, 7,7"-dimethylcupressuflavone. HAm = Hexamethylamentoflavone; HCu = hexamethylcupressuflavone; PH= pentamethylhinoki- flavone; PR= putative hexamethylrobustaflavone, Ul = unknown methyl ether I; U2 = unknown methyl ether 2. + = Major band; m = minor band; t = trace detected TLC only. morphological distinction noted in refs. [ 14.- 15]. The first (I) C allitris and Actinostrobus two species are characterized by 3-4 ovules per cone scale Characterized by the presence of amentoflavone de (cf. 6-10). up to 12 seeds per cone (cf. up to 30) and pollen rivatives and the absence of detectable hinokiflavone sacs that are concealed within the male cone (cf. pollen derivatives. The only recent report of the occurrence of sacs protruding from the cone). It will be interesting to see hinokiflavone in Callitris is for C. glauca R.Br [9], where it if this agreement between th·e chemical and morphological was detected as a minor component. Our own exam discontinuities is maintained when material of W schwar ination of Australian material of this species revealed no zii, which belongs to the first group on morphological trace of hinokiflavone derivatives, or of its permethyl criteria, is available for chemical analysis. ether in the permethylated crude extract. This is in agree While the separation of Pilgerodendron from ment with Prasad and Krishnamurti's observations on C. Libocedrus sensu la10 by Florin [ 16] has received general rhomboidea [8], as well as our own studies on seven other acceptance. de Laubenfels [2] has questioned Li's recog species. It can be concluded, therefore, that the occurrence nition of Austrocedrus and Papuacedrus as distinct of hinokiflavone in concentrations that can be detected by genera. The only difference in biflavonyl pattern between our survey methods is not a feature of C allitris species. these genera is the absence ofhinokiflavone as a detectable This is in marked contrast to its reliable detection in all band in both species of Libocedrus examined (Table 1), other Callitroid genera, except Actinostrobus, and in but minor amounts of hinokiflavone pentamethyl ether dicates a clear genetic difference between these two groups were detected in the crude extracts on permethylation. of genera. This homogeneity of the biflavonyl patterns certainly supports the recognition of these genera as constituting a closely allied group. (2) N eocallitropsis An examination of Table 1 shows that the major biflavonyls of the Callitroideae are derived from the same Characterized by major bands of amentoflavone and a three parental biflavones as have been reported from monomethyl amentoflavone, and minor bands of hinoki members of the northern subfamily, Cupressoideae: viz. flavone, a monomethyl hinokiflavone and a dimethyl amentoflavone. hinokiflavone and cupressuflavone. As in amentoflavone. de Laubenfels [2] concluded that the northern genera, however, they are not uniformly N eocallitropsis has an affinity with C allitris, more particu distributed. The following five generic groups can be larly with C. neocaledonica, on the basis of similarities in recognized on the biflavonyl pattern. cone structure and intermediate leaves. Chemically, how- Biflavones of the subfamily Callitroideae, Cupressaceae 971 ever, the two are clearly distinguished by the absence of The last two groups are clearly related by the presence detectable amounts of hinokiflavone derivatives in all the of cupressuflavone or its derivatives, which must be Callitris species examined, and the presence of an as yet regarded as a chemical specialization within the family undetermined compound as a minor component of the [10, 22]. amentoflavone band of Neocallitropsis in BPF. A comparison of the above generic grouping with those of Li [ 1J shows marked inconsistencies. Cupressuflavones are found in members of all three tribes of the (3) Lihocedrus, Pap11acedrns. A11strocedrns and Pilyrro Callitroideae, as well as in members of the other sub dendron family. Only the treatment of Tetrac/inis as a highly Chemically a highly uniform group characterized by distinctive genus assigned to its own tribe appears to be major amounts of amentoflavone and its 4'"-monomethyl supported by the chemical data, though its placement with ether, as well as trace amounts of hinokiflavonc or its the southern genera in the Callitroideae is suspect. derivatives detectable by permethylation of the extract. Since Li's classification is based primarily on the analysis of cone-scale characters, it is possible that the taxonomic importance of this character has been over (4) Fitzroya, Widdringtonia and Dise/ma rated. Indeed, recent work by de Laubenfels [2, 17] has Characterized by the presence of 7, 7"-dimethylcup questioned the validity and importance of this character in ressuflavone, as well as major amounts of amentoflavone assessing taxonomic relationships. A redefimtion of the and its partial methyl ethers. This is a rather diverse group, supragcneric taxa within the family as a whole must await both chemically and morphologically. The monotypic a full reassessment of the morphological data. Fitzroya is distinguished by the absence of hinokiflavone derivatives (detectable only in the permethylated crude extract) and dimethyl ethers of amentoflavone, and has EXPERIMENT AL somewhat smaller amounts of the monomethylamento Details of collections and voucher specimens are given in the flavone. de Laubenfels [ 17] has recently proposed a close Appendix. For each sample, the leaves and small branchlets were relationship between Fitzroya and Dise/ma on morpho dried, crushed, and extracted in 70 ~ .. Et OH for 48 hr, filtered, logical grounds. Boutelje [ 18] and Moseley [ 19] consider washed with petrol ( bp 60- 80' ) if needed, and the residue dried. Fitzroya a distinctive genus, with wood and cone scale The residue was re-extracted in EtOH and the resulting extract characters indicating a transitional position between was chromatographed on both thick and pre-coated Si gel plates northern and southern genera in the family. Widdringtonia developed in C 0 H 0 pyridine-HCO2 H (BPF) (100:20:7). has previously been allied with C allitris and Actinostrobus Biflavones appeared as a number of dark, UV-absorbing bands, [2, 19, 20], but there is no evidence of such affinity on the fluorescing yellow or orange on addition of AICl 3 . Each band was basis of biflavonyl pattern. Heartwood chemistry [21 J extracted individually and a further separation carried out on indicates Widdringtonia to be a distinctive genus having pre-coated cellulose plates developed in freshly prepared 11- affinities with both northern and southern genera. BuOH- 2 N NH 4 OH, [I: I (upper layer)] (BN). Initial identifi cations were made in comparison with authentic markers in both solvents. Authentic markers of amentoflavone, 4"'-monomethyl (5) Tetraclinus amentoflavone, 4"', 7"-dimethylamentoflavone, hinokiflavone This genus has amentoflavone and cupressuflavone as and 7"-monomethylhinokiflavone were obtained from H. Geiger, its major biflavonyls. It is distinguished from the previous whilst cupressuflavone, 4'-monomethylamentoflavone and 7,7" group by the absence of dimethyl ethers. Hinokiflavone dimethylcupressuflavone were isolated from Araucaria bidwillii derivatives are present in the trace amounts detectable and A. cunninghamii (23-26]. It was observed that bands only by permethylation of the raw extract. Long consid composed of amentoflavone based derivatives fluoresced yellow ered distinct amongst the Cupressaceous genera [20], its when reacted with AICl 3 , hinokiflavone derivatives dark to bright chemical alliances would appear to be with the northern yellow and cupressuflavone derivatives orange. genera, where similar patterns have been recorded in Permethylations of crude extracts, bands and purified com Cupressus [ 1OJ. Heartwood chemistry also casts doubt on pounds were carried out using Me 2 SO 4 in dry boiling Me 2 CO its placement in the Callitroideae and indicates affinities and fused K 2 C03 . Permethyl ethers of amentoflavone, cup with northern genera [21]. ressuflavone and hinokiflavone were identified by permethylating Table 2. Chromatographic and spectral data of permethylated parental biflavones R 1 s• Permethylated ethers BPF BPEFD UV fluorescencet UV spectra ().~~~H) Amentoflavone 0.37 0.40 yellow 267, 328 Cupressuflavone 0.41 0.45 orange 268,332 Compound 2 0.42 0.58 light blue 265,323 Compound I 0.44 0.54 white/yellow 258,335 ?Robustaflavone (trace) 0.50 0.68 light blue Hinokiflavone 0.53 0.75 blue 265,323 • R I values are variable, but the relative positions of the permethyl ethers in each solvent are characteristic. t As observed on plates run in BPF and dried in a hood for ea 0.5 hr. 972 P.A. GADEK and c. J. QUINN authentic markers and comparing them in BPF and 19. Moseley, M. F., Jr. (1943) Lloydia 6, 109. C 6 H 6 -pyridine-ethyl formate-dioxan (BPEFD). as well as UV 20. Baird, A. M. (1953) Phytomorphology 3, 258. fluorescence and UV spectral analysis. as documented in the lit. 21. Erdtman, H. and Norin, T. (1966) Fortschr. Chem. Org. [26-29). Naturst. 24, 206. Two as yet unidentified compounds were extracted in associ 22. Gadek, P.A. (1982) Phytochemistry 21, 889. ation with other biftavonoid bands. Compound I was extracted 23. Khan, N. U., Ansari, W. H., Usmani, J. N., Ilyas, M. and from Actinostrobus pyramidalis and C allitris neocaledonica, com Rahman, W. (1971) Phytochemistry 10, 2129. pound 2 from Neocallitropsis. Both compounds were extracted in 24. Khan, N. U., Ilyas, M., Rahman, W., Mashima, T., Okigawa, association with amentoftavone or its derivatives in BPF and BN M. and Kawano, N. ( 1972) Tetrahedron 28, 5689. and when permethylated gave the same fluorescent products 25. Ilyas, M., Seligmann, 0. and Wagner, H. (1977) Z. identified in the permethylated crude extracts. Initial UV spectral N aturforsch. Tei/ C 32, 206. analysis on these permethylated ethers, together with UV 26. Ilyas, N., Ilyas, M., Rahman, W., Okigawa, M. and Kawano, fluorescence and chromatographic analysis (see Table 2), in N. (1978) Phytochemistry 17, 987. dicates that these compounds do not align with any reported data 27. Chexal, K. K., Handa, B. K. and Rahman, W. (1970) J. for biftavones. Further work is proceeding. Chromatogr. 48, 484. 28. Lin, Y. M. and Chen, F. C. (1975) J. Chromatogr. 103, D33. Acknowledgements-We thank Mr. D. Symon, of the Waite 29. Dossaji, S. F. and Mabry, T. J. (1975) Rev. Latinoam. Quim 6, Institute, University of Adelaide, South Australia, for assistance 37. in obtaining plant material; Mr. J. T. Waterhouse for collect 30. Staffen, F. A. (1981) Index Herbariorum Part I. Dr. W. Junk, ing material of Actinostrobus acuminatus and the use of The Hague. the Herbarium of the School of Botany, University of N.S.W.; Dr. L. A. S. Johnson, Director, for permission to sample APPENDIX specimens held by the National Herbarium of N.S.W.; Mr. L. Craven for facilitating access to specimens of Neocallitropsis held Location and collecting details of voucher specimens are given by the Herbarium Australiense, Canberra, A.C.T.; and Dr. P. G. below. Abbreviations ofherbaria follow Index Herbariorum [30). Wilson for helpful advice. Actinostrobus acuminatus. Part. UNSW 11567, J. T. Waterhouse 8. viii. 1981, WA. A.pyramidalis Miq. UNSW Symon I.ii. I 982, cult. Waite Arboretum 632, SA. Callitris macleayana (F. REFERENCES Muell.) F. Muell. UNSW 12864, Gadek 12.iv.1982,cult. NSW. C. I. Li, H. (1953) J. Arn. Arb. 34, 17. oblonga Rich. UNSW Symon l.ii.1982, cult. Waite Arboretum 2. de Laubenfels, D. J. ( 1972) in F Iara de la N ouvel/e C aledonie 1225, SA. C. neocaledonica Dummer NSW de Laubenfels et Dependences 4. 8.x.1957, NC. C. sulcata (Parl.) Schlechter NSW 28864 Hotchkiss 3. Dallimore, W. and Jackson, A. B. (1966) A Handbook of 15.iii.1954, NC. Neocallitropsis pancheri (Carriere)de Laubenfels Conifera and Ginkgoaceae 4th edn (revised by S. G. CANB. Hartley 15068 23.xi.1979, NC. Papuacedrus papuana (F. Harrison). Arnold, London. Muell.)Li UNSW 4206 Quinn 24.vi.1974, PNG. P. papuana (F. 4. Florin, R. and Boutelje, J. (1954) Acta Hort. Bergiani 17, 7. Muell.)Li UNSW 4213 Quinn 24.vi.1974, PNG. P. torricellensis 5. Willis, J. C. (I 966) A Dictionary of the Flowering Plants and (Diels)Li NSW van Royen NGF 18250 6.ix.1963 PNG. Ferns 7th edn (revised by H. K. Airy Shaw). Cambridge Tetrac/inus articulata (Vahl.)Masters UNSW Gadek 20.5.1981, University Press, London. cult. NSW. Dise/ma archeri Hooker fil. UNSW Gadek 15.9.1981, 6. Gadek, P.A. and Quinn, C. J. (1982) Phytochemistry 21,248. cult. UNSW. Austrocedrus chilensis (D. Don)Florin & Boutelje 7. Sawada, T. (1958) J. Pharm. Soc. Jpn. 78, 1023. NSW de Barba 980 23.ii.1946, SAm. Pilgerodendron uviferum (D. 8. Siva Prassad, J. and Krishnamurty, H. G. (1977) Don)Florin NSW Sargent 21.i.1905, SAm. Fitzroya cupressoides Phytochemistry 16, 801. 0 (Mollina)Johnston NSW de Barba 1045 4.iii.1946, SAm. 9. Ansari, F. R., Ansari, W. J., Rahman, W., Okigawa, M. and Libocedrus yateensis Guill. NSW de Laubenfels 4.xii.1957, NC. L. Kawano, N. (1981) Indian J. Chem. 208, 724. plumosa (D. Don)Sargent NSW Petrie vi.1910, NZ. 10. Geiger, H. and Quinn, C. J. ( 1982) The F/avonoids: Advances ffiddringtonia dracomamana Stapf NSW 989 vi.1920, SAf. W in Research (Harborne, J. B. and Mabry, T. J., eds.) pp. whytei Rendle NSW Darren-Smith 10.viii.1945, cult. W juni 505-534. Chapman & Hall, London. perioides (L.)Endl. UNSW Symon 1.ii.1982, clllt. Waite 11. Dossaji, S. F., Mabry, T. J. and Bell, E. A. (1975) Biochem. Arboretum 1275A, SA. Wjuniperioides (L.)Endl. UNSW Symon Syst. Ecol. 2, 171. l.ii.1982, cult. Waite Arboretum 1231, SA. W cupressoides 12. Quinn, C. J. and Gadek, P.A. (1981) Phytochemistry 20,677. (L.)Endl. UNSW Symon l.ii.1982, cult. Waite Arboretum 1283A, 13. Fatma, W., Taufeeq, H. M., Shaida, W. A. and Rahman, W. SA. (1979) Indian J. Chem. 178, 193. 14. Stapf, 0. (1933) in Flora Capensis (A. W., Hill, ed.) Vol. 5, 2, suppl. Reeve, London. 15. Chapman, J. D. (1960) Kirkia I, 138. Abbreviations: WA, Western Australia; SA, South Australia; 16. Florin, R. (1930) Svensk. Bot. Tidskr. 24, 132. NSW, New South Wales; NC, New Caledonia; PNG, Papua New 17. de Laubenfels, D. J. (1965) Phytomorpho/ogy IS, 414. Guinea; SAm, South America; SAf, South Africa. 18. Boutelje, J. B. ( 1955) Acta Hart. Bergiana I 7, 177.