Christoph Oberprieler The Systematics of Anthemis L. - Herbmedit.org

Christoph Oberprieler 

The Systematics of Anthemis L. (Compositae, Anthemideae) in W and 

C North Africa 

Abstract 

Oberprieler, Ch.: The Systematics of Anthemis L. (Compositae, Anthemideae) in W and C 

North Africa. - Bocconea 9: 1-328.1998. - ISSN 1120-4060. 

Anthemis L., a mainly Mediterranean and SW Asian genus of the Compositae-Anthemideae, 

is revised far the W and C North African part (Marocco, Algeria, Tunisia, Libya) 

of its distribution range. On the basis of detailed morphological studies ali species are fully described 

and illustrations of habit, leaves, involucral bracts, and pales are presented, along with 

photographs of mature achenes under SEM and of transverse sections of achenes under LM. 

Keys far the deterrnination of ali species, subspecies, and varieties are provided. Complete 

synonymies are given, inc1uding the typification of the names as far as possible. Oistribution 

maps of ali taxa studied are presented. Chromosome numbers and karyotypes far most taxa are 

provided and discussed. 

In the area covered, Anthemis is represented by 25 species belonging to two subgenera, 

four sections, and four series. Three taxa are described as new to science: A. maritima subsp. 

pseudopunctata Oberprieler, A. stiparum subsp. intermedia Oberprieler, and A. zaianiea 

Oberprieler. The following ten new combinations are established: A. abylaea (Font Quer & 

Maire) Oberprieler, A. maroeeana subsp. aguilarii (Maire & Sennen) Oberprieler, 

A. mauritiana subsp. faurei (Maire) Oberprieler, A. peduneulata subsp. atlantica (Pomel) 

Oberprieler, A. pedunculata subsp. clausonis (Pomel) Oberprieler, A. peduneulata var. discoidea 

(Boiss.) Oberprieler, A. peduneulata subsp. turolensis (Pau ex Caballero) Oberprieler, 

A. punctata subsp. kabylica (Battand.) Oberprieler, A. stiparum subsp. sabulieola (Pomel) 

Oberprieler, A. tenuiseeta subsp. jahandiezii (Maire) Oberprieler. 

Multivariate statistical methods (principal component analysis) and/or analyses of random 

amplified polymorphic ONAs (RAPO) were used to assess morphological and genetic variation 

and to address problems of species delimitation in the Anthemis boveana group and the 

A. pedunculata - A. punetata complex. 

Contents 

1. Introduction .......... ......... .. ......... .. .. .. .. .. .. ........... ........................... ......... ...... .. ....... .. ...... 6 

2. Material and methods ........................... .. ....... .. .................................... .. ... .......... ........ 7 

3. Taxonomic history ............. .. ...... .. ........... .. .... ......... .. ........... .. ............ ........ .. .. ..... .. ..... . 9 

4. Generic relationships .... ..... .. ........ .. ...... .. ....... .. .. .. ............. ....... .. ..... ....... .. . .. ........... .... lO 

5. Supraspecific taxonomy ............ .. ..... :................................. .. .. ... .. ..... ....... .. .. .. ........... 12 

6. Phytogeography .... .. ... .................................... .. ..... ....... .. ....................... .. .. ............. .. 14 

7. Modes of speciation within Anthemis in N Africa ................ .............. ..................... 16 

8. Delimitation oftaxa - Concepts of species, subspecies, and varieties .... .. .... ...... .. .... 18 

9. Morphological features ............................................................................................ 20 

lO. Chromosomes ....... ....... ............................... ............. .. ...................................... ... .. ... 35 

Il. Pollen morphology ...... ......... .. ....... .. .......... .... .. .. ....... .......... ..... ............... .. .. ............ .. 66

6 Oberprieler: Anthemis in N Africa 

12. Species delimitation in the Anthemis boveana group .............................................. 76 

13. Species delimitation in the Anthemis pedunculata - Anthemis punctata complex ... 85 

14. Systematic descriptions and keys .. ......................................................................... 107 

Key to the N African species .... .................. ................... ......... ..... .. ................ 109 

Clé des espèces de l'Afrique du Nord .......... ...... ............. ....... .................... ... 111 

. Anthemis subg. Anthemis ............... .......... ...................... .................. .............. 114 

Anthemis sect. Anthemis ..................................... .. .................. ...... ... 114 

Anthemis sect. Hiorthia ..................................... ....................... ....... 227 

Anthemis sect. Maruta ..... ................................. .......... ............... .... .. 297 

Anthemis subg. Cota .. .............. ......................... ............................................. 307 

Anthemis sect. Cota ....................................... .................................. 307 

15. Acknowledgements .... .. ........................................................ .................................. 313 

16. References ........... .................... ........................................................... ................... 314 

17. Index to scientific names ........... ........ .. .. ................................................. .. .... ......... 324 

1. Introduction 

The genus Anthemis belongs to the family Compositae Giseke, alternatively named Asteraceae 

Dumort. It is the name-giving genus of the Anthemideae Cass., a tribe which 

presently consists of 109 genera (Bremer & Humphries 1993; adding Castrilanthemum 

recently described by Vogt & Oberprieler 1996) and shows an extratropical, mainly Old 

World-centred distribution. According to the most recent phylogenetic reconstruction of 

the family (Bremer 1996), the Anthemideae are a member of the subfamily Asteroideae 

and form the sister group to the Astereae. While Heywood & Humphries (1977) indicate 

around 130 species in the genus, more recent estimates give a total of c. 210 annual, 

biennial, or perennial species (Bremer & Humphries 1993). 

Since Candolle's account of Anthemis in his Prodromus (Candolle 1838), the genus has 

never again been subject to a complete revisionary treatment. However, quite comprehensive 

regional treatments like Boissier's (1875) Flora orientalis account and Eig's (1938) 

studies on the orientai species of Anthemis, along with revisions made in the course of the 

large flora projects of the last 40 years, have resulted in our up-to-date understanding of 

the genus over a large part of its distributional range, viz. Europe (Fernandes 1975a, 

1975b, 1976, 1983), E Europe and W Asia (Fedorov 1961), Turkey (Grierson & Yavin 

1975), the Near East (Feinbrun-Dothan 1978), the Flora iranica area (Iranshahr 1986), 

and Saudi Arabia (Ghafoor & AI-Turki 1997). Additional information was contributed by 

revisionary treatments of defined infrageneric groups (A. ammanthus group, Greuter 1968; 

A. sect. Maruta, Yavin 1970; A. sect. Anthemis, Yavin 1972; A. tomentosa group, Georgiou 

1990) and by geographically limited biosystematic revisions (Bulgaria: Kuzmanov & 

al. 1981 and Thin 1983; Spain: Benedf i Gonzalez 1987). 

In N Africa, as compared to other regions within the distributional range of Anthemis, 

the taxonomy of the genus is poorly understood, and a criticaI revision is lacking. While 

there are flora treatments available for Algeria (Quézel & Santa 1963), Tunisia (Pottier 

Alapetite 1981), Libya (Alavi 1986), and Egypt (Tackholm 1974), for Morocco there is 

only a checklist (Jahandiez & Maire 1934). The main aim of the present study was to fill

Bocconea 9 - 1998 7 

this gap by elaborating a complete taxonomic treatment of the genus for the area covered 

by Maire's Flore de l'Afrique du Nord (Maire & al. 1952-1987), i.e. the territories of 

Morocco, Algeria, Tunisia, and Libya. 

2. Material and methods 

This revision is based partly on plant material collected during field trips to Morocco 

and Tunisia in 1992-1995, which is deposited at the herbarium of the Botanic Garden & 

Botanical Museum Berlin-Dahlem (B). In June 1992, the study and collection of Anthemis 

in N Morocco was enabled by the participation in the "Iter Mediterraneum V" sponsored 

by the Organization for the Phyto-Taxonomic Investigation of the Mediterranean Area 

(OPTIMA). Further collections in Morocco (April-May 1993, May 1995) and Tunisia 

(May 1994) were made during field trips organised by the Botanic Garden and Botanical 

Museum Berlin-Dahlem. 

Additionally, the study is based on herbarium material from the following public herbaria, 

abbreviated as in Holmgren & al. (1990) and Holmgren & Holmgren (1996), and 

private collections: 

B 

BC 

BC-Sennen 

BM-Clifford 

C-Vahl 

FI 

G 

GOET 

K 

LINN 

MA 

MPU-AtN 

MPU-Braun-Blanquet 

MPU-Dubuis 

MPU-Weiller 

MPU-Sauvage 

MSB 

p 

Botanischer Garten & Botanisches Museum Berlin-Dahlem 

Institut Botanic de Barcelona 

Institut Botanic de Barcelona (Herbarium F. Sennen) 

The Natural History Museum, London (Herbarium G. Clifford; 

microfiche) 

Botanical Museum and Library, University ofCopenhagen 

(Herbarium M. Vahl; microfiche) 

Museo di Storia Naturale, Museo Botanico, Firenze 

Conservatoire et Jardin botaniques de la ville de Genève 

Systematisch-Geobotanisches Institut der Universitlit Gèittingen 

Royal Botanic Gardens, Kew (London) 

The Linnean Society of London (Herbarium C. Linné; microfiche) 

Rea! Jardfn Botanico, Madrid 

Institut de Botanique Montpellier (Herbier de l' Afrique du 

Nord) 

Institut de Botanique Montpellier (Herbarium J. Braun 

Blanquet) 

Institut de Botanique Montpellier (Herbarium A. Dubuis) 

Institut de Botanique Montpellier (Herbarium M. Weiller) 

Institut de Botanique Montpellier (Herbarium Ch. Sauvage) 

Institut filr Systematische Botanik der Ludwig-Maximilians 

Universitlit, Miinchen 

Muséum National d'Histoire Naturelle - Laboratoire de 

Phanérogamie Paris

8 

Oberprieler: Anthemis in N Africa 

P-DESF 

P-LA 

RNG 

SEV 

UPS-Burser 

Herb. Bayreuth 

Herb. Oberp~ieler 

Herb. Vogt 


Phanérogamie Paris (Herbarium R. L. Des fontaines) 


Phanérogamie Paris (Herbarium J. B. P. A. de Monet de 

Lamarck) 

University of Reading, School of Plant Sciences 

Departamento de Biologfa Vegetai y Ecologfa, Universidad de 

Sevilla 

Botanical Museum at Uppsala University (Herbarium 1. Burser; 

microfiche) 

Universitiit Bayreuth, Lehrstuhl fiir Ptlanzengeographie 

private collection of Ch. Oberprieler (Jena) 

private collection of R. Vogt (Berlin) 

Specimens were exarnined with a stereo-microscope (Wild M5A), measurements being 

performed using the camera-lucida device to slot in a millimetric scale in the corresponding 

magnification. 

The description of taxa is based on measurements made on dried plants. The variation 

ranges cited cover the total observed variation exhibited by a particular taxon. Extreme 

values have been placed in parentheses, while mai n range values relate to the interval 

covered by the mean ± l standard deviation of a particular character distribution. 

A list of specimens seen is presented under each species. Localities are given countrywise 

from W to E in the following order: Spain (N African part), Marocco, Algeria, Tunisia, 

and Libya. Within each country, administrative districts (Morocco: Province; Algeria: 

Wilaya; Tunisia: Gouvernorat; Libya: Baladiyah) are listed in alphabetic order. Within 

each district, specimens are listed from N to S. Most of the specimens are given with longitude 

and latitude of the collecting locality, written as stated by the collector(s), or taken 

from various maps [when they are cited in square brackets]. Distribution maps were drawn 

using the public domain map projection program MICROCAM by S. A. Loomer (ftp://ftp. 

usma.edulg&ene/gsl/microcarn/). 

For comparative carpological studies, achenes were taken from herbarium specimens 

and soaked for two days in a mixture (6 : l) of lO % aqueous solution of dioctyl sodium 

sulfosuccinate and 96 % ethanol (Peterson & al. 1978). Following dehydration through a 

l: 1,3: 7, l : 9 incrementai series of sulfosuccinate/ethanol, with changes every 24 hours, 

and a final step in pure 96 % ethanol for 45 min, achenes were subsequently soaked for 4- 

6 weeks in a pretreatment solution of Kulzer Technovit 7100® and solidifier no. 1 at 4°C. 

After final embedding in a l : 15 mixture of the pretreatment solution and Kulzer solidifier 

no. 2, median-transverse and median-Iongitudinal sections of achenes were made using a 

rotary microtome. 3-5 flm thick sections from the middle half of the achenes were stained 

with toluidine blue and mounted in Vitroclud (Langenbrinck). 

For scanning electron microscopy (SEM) achenes were taken from herbarium specimens, 

mounted on preparation stubs, and coated with a gold-palladium layer 2 nm thick in 

a Polaron sputter-coater. The observations and photography were performed with a Philips 

SEM 515 at the Botanical Museum Berlin-Dahlem.

Bocconea 9 - 1998 9 

3. Taxonomic history 

Though mentioned already in Genera pIanta rum (Linnaeus 1737a) and Hortus Cliffortianus 

(Linnaeus 1737b), the generi c name Anthemis was validated by Linnaeus (1753) in 

the first edition of Species plantarum. By doing so, Linnaeus preferred the generi c name 

already used by Micheli (1729) over the name Chamaemelum which had been used by 

Bauhin (1623) and Tournefort (1700) for most of the Linnean Anthemis species. 

In contrast to definitions given in Generaplantarum (Linnaeus 1737a: 256; 1754: 381) 

where Anthemis and Anacyclus are described to differ in the shape of achenes which are 

stated to be compressed in Anacyclus and terete in Anthemis, Linneaus (1753) arranged 

the species of these two genera according to the rather artificial character of presence vs. 

absence of ray florets in Species plantarum. This led to decades of uncertainty about the 

correct demarcation of the two genera which was still apparent in Willdenow's fourth 

edition of Species plantarum (Willdenow 1803) and Persoon's Synopsis plantarum 

(Persoon 1807) where both mentioned authors stressed the essential fruit characters as 

diagnostic for Anacyclus and Anthemis but ignored them when assigning the species to 

genera. 

Shortly after Linnaeus' Species plantarum (Linnaeus 1753), Miller (1754) validated the 

generic name Chamaemelum (Druce 1914) to compri se some Anthemis species cultivated 

in English gardens. In Miller (1768), however, Anthemis is adopted and Chamaemelum 

treated as a synanym. A few decades later, Necker (1790), Gaertner (1791), and Moench 

(1794) revived the name Chamaemelum to accommodate some former Anthemis species 

with apically rounded achenes in a separate genus. Though initially refraining from the 

recognition of these two enti ti es on a level higher than subgenus (Cassini 1817: 83), 

Cassini (1823) contributed to the excessive dismemberment of Anthemis into a number of 

genera by accepting Anthemis and its meanwhile described segregates Chamaemelum, 

Maruta, Ormenis, Cladanthus, and Lepidophorum as members of his "Anthémidées-Prototypes 

vraies", later (Cassini 1825) adding the monotypic genus Lyonnetia for Anacyclus 

creticus L. (L. pusilla Cass.; currently recognized as Anthemis rigida Sm.). 

Lessing and Candolle were considerably influenced by the work of Cassini when they 

provided their synthetic classifications for the genera of Compositae. Lessing (1832) 

placed Anthemis species, which, on the basis of discoid capitula, had been assigned to 

Necker's Hiorthia (H. orientalis = Anacyclus orientalis L.) or Cassini's Lyonnetia in the 

subtribe Artemisiinae, far apart from the radiate representatives in subtribe Anthemidinae. 

This extremely unnatural dismembering of Anthemis species was overcome by Candolle 

(1838) who included Lyonnetia, along with Anthemis, Ma ruta, Anacyclus, Ormenis, 

Cladanthus, and Lepidophorum in his "Euanthemideae " . 

Moris (1840-1843) was the first to note that the basally saccate disc florets observed by 

Cassini in his Ormenis mixta are also found in Maruta fuscata and Anthemis nobilis. He 

united them ali under the generic name Maruta, foreshadowing to some extent the present 

circumscription of the genus Chamaemelum, except by also including Anthemis cotula, the 

type of Maruta. 

A further step toward a natural arrangement of Anthemis species was made by Gay, the 

designated Anthemis monographer of his period, when he contributed an account of the 

genus in the appendix to the second volume of Gussone's Synopsis (Gussone 1844: 866- 

872). Here, Gay created the new genus Cota to accommodate four Anthemis species with

lO 


dorso-ventralIy compressed and indistinc1y ribbed achenes, while leaving the species with 

terete achenes and IO distinct ribs in Anthemis. He also inc1uded A. fuseata and A. eotula 

in Anthemis, while noting their similarity with A. mixta which he left in Ormenis. 

Problems relating to the correct demarcation of Ormenis against Anthemis owe their 

solution to Schultz's revised taxonomy of the Anthemideae (in Schnitzlein 1854: 69-70; 

Schultz 1860). Schultz proposed a subdivision of the tribe into six subtribes, mainly on 

carpological grounds, with members of Anthemis assigned to Cotinae and Anthemidinae 

and members of Ormenis (O. fuseata, O. mixta, O. nobilis), along with the unispecific 

Cladanthus forming subtribe Ormenidae. 

Bentham & Hooker (1873) and Hoffmann (1891-1892) reacted to the excessive splitting 

of Anthemis by more or less re-establishing the Linnean circumscription of the genus, 

in which they merged such rather unrelated genera as Ormenis, Lepidophorum, or the N 

African Rhetinolepis described by Cosson (1856), while keeping separate other segregates 

of Anthemis, e.g. Ammanthus (described in Boissier 1849) and Hiorthia (containing the 

poorly understood Anaeyclus orientalis L.). By that time, however, Boissier (1875) had 

already worked out a taxonomy of Anthemis that corresponds almost completely with our 

present circumscription and supraspecific taxonomy of the genus. Only Ammanthus was 

stilI to face a long period of generic independence, or rnisplacement in Chrysanthemum, 

until Greuter (1968) proposed its inc1usion in Anthemis, thus completing OUT present notion 

of the genus. 

4. Generic relationships 

Anthemis provides the type of the name Anthemideae, used by Cassini (Ì819: 192) to 

designate his eleventh tribe of Compositae. Cassini (1823) divided the tribe into two major 

groups: "Anthérnidées-Chrysanthémées" and "Anthémidées-Prototypes", based on the 

absence vs. presence of paleso These two subtribes, later-on validly named Chrysantheminae 

Less. and Anthemidinae Dumort., were used by most systematists concerned with the 

infratribal taxonomy in the folIowing decades (e.g. Candolle 1838, Boissier 1875, Hoffmann 

1891-1892, Hegi 1928). The artificiality of this subdivision c1early stated by Merxmiiller 

(1954) and Wagenitz (1964) and demonstrated by Greuter (1968) when he found 

that in Ammanthus, up till then c1assified with Chrysantheminae, the presence or absence 

of scales "does not even necessarily suffice to distinguish species", which led him to trans~ 

fer its species to Anthemis. Hybridisation experiments among members of Anthemideae 

made by Mitsuoka & Ehrendorfer (1972) have shown that the inheritance of pales is 

probably under simple oligogenic control. Further evidence carne from the ambiguous 

position of the unispecific genus Lepidophorum, usualIy placed in the vicinity of Anthemis 

due to the presence of pales, but in morphological, phytochernical (Bohlmann & al. 1973, 

Harborne & al. 1976), and embryological features (Harling 1960) c10sely resembling palelacking 

members of the tribe. Observations of sporadic paleate capitula in Chrysanthemum 

(Napp-Zinn & Eble 1978) and inc1usion of paleate and epaleate species into Athanasia by 

KiilIersj6 (1991) point in the same direction. 

Owing to the artificiality of a c1assification based on the presence vs. absence of pales, 

numerous attempts have been made in the last forty years to elaborate a more satisfactory

Bocconea 9 - 1998 11 

taxonomy of the tribe. Thorough embryological studies of representatives of Anthemideae 

by Harling (1950, 1951, 1960) contributed to a natural delimitation and grouping of genera. 

Features of embryo sac formation can be used to set off Anthemis against other 

genera that have, at one time or another, been treated as synonymous or closely related. 

While ali members of Anthemis were found to have a tetrasporic embryo sac development, 

members of the paleate genera Chamaemelum (sub Ormenis), Lepidophorum, Anacyclus, 

and Cladanthus showed monosporic embryo sac formation (Harling 1950, 1960). Further 

findings (Harling 1951) suggest close relationships of Anthemis with genera that had been 

considered to be unrelated due to their epaleate receptacles: A tetrasporic embryo sac 

development war found in representatives of Tanacetum (sub Chrysanthemum parthenium, 

C. millefolium), Tripleurospermum (sub Matricaria maritima, M. oreades), Heteranthemis 

(sub C. viscidehirtum), and, in an older embryological contribution by Martinoli 

(1940), also in Nananthea. 

Cytological (Uitz 1970) and carpological surveys (Reitbrecht 1974) yielded a subdivision 

of Anthemideae into seven provisional groups. Anthemis was considered to hold a 

'somewhat centrai position in the so-called Matricaria group, also including Chamaemelum, 

Cladanthus, Tripleurospermum, Matricaria, Anacyclus, Otospermum, Daveaua, and 

the S African Pentzia. In contrast to embryological findings mentioned above, Reitbrecht 

(1974) thought of the generic relationships of Anthemis to lie in two directions: with Anacyclus, 

due to the dorso-ventrally flattened achenes in Anthemis subg. Cota, and with 

Chamaemelum and Tripleurospermum. Outside the Matricaria group, additional relationships 

of Anthemis were suggested to exist with Tanacetum which was treated as the centrai 

member of the so-called Chrysanthemum complex. 

In Heywood & Humphries's (1977) account of Anthemideae the so-called Anthemis assemblage 

was but a slight modification of Reitbrecht's (1974) Matricaria group, with the 

exclusion of Pentzia and related S hemisphere genera. The close relationship of Anthemis 

and Chamaemelum that had been suggested by crossing data (Mitsuoka & Ehrendorfer 

1972) was considered by Heywood & Humphries (1977) to be obscured by differences in 

fruit and corolla structure (Briquet 1916), embryo sac development (Harling (1960), and 

phytochemical data (Bohlmann & al. 1965, Bohlmann & Zdero 1966). Further achene 

morphological and anatomical studies by Hl1mphries (1977, 1979) revealed the suggested 

close relationship between Anthemis subg. Cota and Anacyclus to be rather superficial: 

While achenes of Anacyclus species were found to have only two vascular bundles in the 

pericarp wall, those of Anthemis species, though having ten or even more epicarpic ribs, 

invariably had five. Cyanogenic glycoside and flavonoid profiles (Greger 1977, 1978) also 

suggest a closer relationship of Anacyclus with Achillea than with Anthemis. 

Bremer & Humphries (1993) elaborated a subtribal classification of Anthemideae using 

ali morphological, embryological, and phytochemical information available at that ti me in 

a cladistic ana1ysis. They classified Anthemis together with the unispecific Nananthea as 

the only members of the subtribe Anthemidinae. The subtribe Chrysantheminae with four 

genera (Argyranthemum, "Chrysanthemum", Ismelia, and Heteranthemis) was sister to 

the Anthemidinae. However, Bremer & Humphries (1993: 131) state that Nananthea was 

included only provisionally due to its tetrasporic embryo sac development and that "the 

immediate relatives of [Anthemis] are unknown". AIso, the sister group relationship of 

Anthemidinae and Chrysantheminae remains rather uncertain since the analysis of the 

interrelationships of the 12 proposed subtribes yielded a number of equally parsimonious


solutions and the strict consensus tree of these solutions collapsed completely (Bremer & 

Humphries 1993: 90). 

Chloroplast restriction site studies in Anthemideae by Francisco-Ortega & al. (1995) 

and Watson (1996) contributed little to the solution of the problem of the position of Anthemis 

within the tribe, due to the restricted sampling of genera or the absence of Anthemis 

representatives in the data set, respectively. Francisco-Ortega & al. (1997), using sequence 

data of the internai transcribed spacer (ITS) regions of the nuc1ear ribosomal repeat to 

assess the intergeneric relationships of Argyranthemum, provide molecular information on 

32 genera of eight subtribes of Anthemideae. Phylogenetic analyses of ITS data showed 

that most of the subtribes suggested by Bremer & Humphries (1993) are not resolved as 

monophyletic c1ades. Anthemis was found to be sister to Tanacetum vulgare within a 

rather basaI c1ade grouping genera with predorninantly Eurasian and Mediterranean distribution. 

However, low bootstrap values for many of the nodes of the c1adogram suggest 

that caution is needed when interpreting the tree presented. 

As Bremer & Humphries (1993) pointed out, our further understanding of the phylogeny 

of the tribe depends on a better resolution of non-monophyletic taxa, e.g. 

Tanacetum and Tanacetinae. Solutions for these problem taxa may also have positive 

bearings on positioning Anthemis within the tribe. At present, information on additional 

interesting features (e.g. ca1cium-oxalate crystals in epicarp cells) are too sparse to allow 

an improved c1adistic analysis of the tribe. It is nevertheless prornising to compare the 

results of the c1adistic analysis of morphological data given by Bremer & Humphries 

(1993) with cytological data given by Uitz (1970) and Mitsuoka & Ehrendorfer (1972). As 

discussed later (see chapter lO), karyotype similarities appear to indicate relationships 

between the subtribes Anthemidinae, Tanacetinae, Chrysantheminae, Achilleinae, and 

Matricariinae through their predorninantly perenni al and (morphologically) basaI genera 

Anthemis, Tanacetum, Argyranthemum, Achillea, and Tripleurospermum. In such a 

scheme, Anthemis would hold a rather centrai (basaI) position within the tribe, with most 

of the mentioned subtribes showing sister-group relationships to it. 

5. Supraspecific taxonomy 

In the presently accepted circumscription, Anthemis comprises c. 211 species (Bremer 

& Humphries 1993). Traditionally, and mainly on carpological grounds, two main subgroups 

are distinguished within the genus (Wagenitz 1968): A. subg. Anthemis is characterised 

by actinomorphic achenes which are round or tetragonal in cross-section, usually 

have lO longitudinal epicarpic ribs and epicarpic cells filled with sand of ca1cium-oxalate 

crystals; A. subg. Cota exhibits disymmetrical achenes which are dorso-ventrally flattened, 

usually have 8-22 ribs and epicarpic cells that either lack crystals or contain single, large 

crystals (Fig. 3). Since Gay's treatment in Gussone (1844), the latter subgenus was sometimes 

considered as an independent genus on morphological and phytochemical 

(Reitbrecht 1974) or cytogenetical grounds (Mitsuoka & Ehrendorfer 1972). However, as 

discussed in the previous chapter, there is hardly any evidence for a para- or polyphyletic 

nature of Anthemis that would necessitate the acknowledgement of the two entities in a 

rank higher than subgenus.

Bocconea 9 - 1998 13 

The species of the former genus Ammanthus, which were transferred to Anthemis by 

Greuter (1968), were (at least partly) considered to be sufficiently deviating by Fernandes 

(1975b, 1976) to merit recognition as a third subgenus, A. subg. Ammanthus. However, 

acknowledgement of this subgenus will likely cause A. subg. Anthemis to become paraphyletic. 

Yavin' s (1972) treatment of this group as a seri es within A. sect. Anthemis is 

therefore preferred. 

The classification of Anthemis subg. Anthemis traditionally follows the common pattern, 

with recognition of a number of apomorphic sections in addition to more plesiomorphic 

and presumably paraphyletic ones. Representatives of A. sect. Hiorthia (including 

A. sect. Rumata), which are characterised by suffruticose or pleiocormous perennial habit, 

were often considered to form the most basai group of the subgenus (Meusel & Jager ' 

1992). Here, in contrast to the other sections of Anthemis, polyploidy has played a significant 

role in the evolution of species and species groups. Mitsuoka & Ehrendorfer (1972) 

and Kuzmanov & al. (1981) suggested that perennial habit may enable plants to overcome 

the setbacks of reduced fertility caused by polyploidisation. Infrasectional taxonomies 

proposed by Fedorov (1961) and Thin (1983) have to be considered preliminary since they 

concern restricted geographical areas and are based on rather arbitrary selections of representative 

species. 

In contrast to Anthemis sect. Hiorthia, A. sect. Anthemis has received a comprehensive 

assessment of its infrasectional taxonomy by Yavin (1972). In addition to two series accepted 

by Fedorov.(1961), Yavin (1972) defined 14 new. series, .each with.one tO .l2 species, 

segregated 12 species with extremely slender achenes into an independent section 

A. sect. Rascheyanae, and . the fruit-anatomically deviating A. chia into the unispecific 

A. sect. Chiae. However, since besides the annual habit no other synapomorphy for members 

of A. sect. Anthemis is presently known and sister-group relationships of A. sect. Rascheyanae 

and sect. Chiae with species or species groups of A. sect. Anthemis are rather 

probable, the acknowledgement of these two annual sections would cause A. sect. Anthemis 

to become paraphyletic. In the case of A. sect. Rascheyanae, sister-group relationships 

may exist with A. ser. Scariosae, ser. Melampodinae, ser. Haussknechtianae, or ser. 

Cornucopiae due to shared apomorphies (hairy tubes of ray florets and hairy disc florets) 

in some of their members. 

Anthemis sect. Maruta is characterised by the apomorphic feature of subulate pales. 

This morphologically rather distinct section is presumably nested within A. sect. Anthemis 

since it shows some similarities with A. scariosa, a species grouped within A. sect. Anthemis 

by Yavin (1972) but eonsidered to hold a transitional position between the two 

seetions by Grierson & Yavin (1975). A. sect. Maruta was subdivided into three series by 

Fedorov (1961): ser. Cotulae, ser. Microcephalae, and ser. Odontostephanae. When presenting 

a revision of the seetion, Yavin (1970) found the latter series distinct enough to 

exclude it from A. seet. Maruta and proposed its aeknowledgement as a separate section, 

or even genus. 

Anthemis subg. Cota is traditionally subdivided into two sections, A. sect. Cota and 

se et. Anthemaria (e.g. Fernandes 1975b, 1976, Benedf i Gonzales 1987). The main difference 

between the two seetions relates to life span: members of A. seet. Anthemaria are 

eharaeterised by a perennial, suffruticose to pleiocormously herbaeeous habit, A. seet. 

Cota consists of annual to biennial herbs. Fedorov (1961) proposed a different classification, 

subdividing the subgenus (his A. sect. Cota) into five series. Pending a comprehen-


sive morphological and c1adistic study, no deliberate choice is made here between the two 

proposed scenarios, but for practical purposes the c1assification proposed by Fernandes 

(1975b, 1976) is provisionally followed. 

6. Phytogeography 

The total geographical range of Anthemis encompasses almost the whole western Eurasia, 

the Mediterranean, and a small part of E Africa. According to Meusel & Higer (1992) 

the distribution range is very similar to those of Daucus, Cichorium, and Filago. While C 

Europe is inhabited by few archaeophytic species only, the mai n centre of diversity is 

found in SW Asia where about 150 of the c. 210 described species occur and all of the 

presently accepted subgenera and sections are found. The W and C North African part of 

the distribution area, which is covered by the present revision, hosts a markedly lower 

number of species and certainly has played a rather peripheral role in the evolution of the 

genus. 

Anthemis sect. Hiorthia, formed by perenni al herbs and subshrubs of predominantly 

mountainous habitats and often considered to be the most primitive section of Anthemis, 

shows a marked centre of diversity in SW Asia (Turkey and the adjacent Caucasus countries) 

and the mediterranean and submediterranean parts of SE Europe. According to 

Kuzmanov & al. (1981), Asia Minor is considered the primary evolutionary centre of the 

genus in generai and of A. sect. Hiorthia in particular, since diploid representatives of this 

section, in which polyploidy occurs, are concentrated here. While most of the SW Asian 

and SE European species of A. sect. Hiorthia are restricted to small areas, the highly 

polymorphic A. eretica (inc1uding A. carpatica and A. saxatilis) which is represented in 

Turkey by not less than 12 subspecies (Grierson & Yavin 1975) extends to C and even W 

southern Europe, and A. eretica subsp. columnae, a subspecies which is centred on the 

Balkan Peninsula and Italy, even occurs in Algeria in an extremely restricted area. The 

distribution range of three other N African members of A. sect. Hiorthia, A. maritima, 

A. pedunculata, and A. abylaea, contrasts markedly with the pattern just described for 

A. eretica. A. maritima is restricted to the W Mediterranean area and, due to its deviating 

ecology, presumably represents a evolutionallineage quite independent from the remainder 

of the section. Additionally, diploid plants are found scattered throughout the distribution 

range of the species (Menorca, Benedi' i Gonzales 1987; Algeria, A. maritima subsp. 

bolosii, present paper). Similarly, the Ibero-Maghrebine A. pedunculata is also restricted 

to the W Mediterranean region and occurs at the diploid and tetraploid level. The deviating 

fruit morphology, with achenes circular instead of rhombic in cross-section, again 

suggests that A. pedunculata and its c10se tetraploid relative A. abylaea evolved independently 

from the remainder of the section. An obvious pattern of vicariance within A. sect. 

Hiorthia thus emerges, with a main centre of diversity in SW Asia and a minor centre 

formed by the lineages of A. maritima, A. pedunculata, and the Spanish diploid 

A. alpestris in the W Mediterranean area. The tetraploid NE Algerian and N Tunisian 

A. punctata, the fifth N African species of A. sect. Hiorthia, geographically links 

A. pedunculata with the A. eretica group (the Sicilian endemic species A. cupaniana and 

A. eretica subsp. columnae), being restricted to the area where the two lineages overlap.

Bocconea 9 - 1998 15 

Additionally, it shows a elinal morphological variation mediating between two extremes: 

strongly tuberculate achenes which are round in cross-section (as in A. pedunculata) in 

plants from the western part of the species' distribution area and weakly tuberculate achenes 

tending to be rhombic in cross-section (as in A. eretica) in its eastern part. 

Anthemis sect. Anthemis which according to the preliminary Iist given by Yavin (1972) 

comprises c. 60 annual species, shows an originally Mediterranean-Oriental-Submediterranean-Pontic 

distribution area that was enlarged palaeosynanthropically in an Atlantic-W 

Sarmatic-Scandinavian direction (Meusel & Jiiger 1992). The main concentration of species 

is in the E Mediterranean and SW Asian region, where Il of Yavin's (1972) 14 seri es 

are found . The individuaI species of the section are often restricted to rather small areas, 

and even the series, usually, grow only in a small portion ofthe section's total range. In the 

area covered by the present revision four seri es occur: A. ser. Anthemis, ser. Bourgaeinianae, 

ser. Chrysanthae, and ser. Secundirameae, whose members show distinct vicariant 

distribution patterns. 

In Anthemis ser. Anthemis, the elosely related species A. arvensis and A. ruthenica provide 

a good example of vicariance. A. arvensis is considered an originally iberic-centromediterranean-hellenic 

species (A. arvensis subsp. incrassata) which enlarged its areaI 

synanthropically into the Oceanic-Suboceanic parts of W and C Europe (subsp. arvensis); 

A. ruthenica is restricted to the adjacent, dryer, Pontic-Balkanic-Pannonic area (Meusel & 

Jiiger 1992); and the third species of this series, A. auriculata, occupies a comparably 

small area in the S parts of the Balkan Peninsula and the N Aegean region. N Africa is 

reached only by the Mediterranean-Submediterranean A. arvensis subsp. incrassata, restricted 

here to the rather humid regions of the Tangier peninsula and the adjacent W and 

C parts of the Rif mountains. 

Anthemis ser. Bourgaeinianae consists of five allopatric species growing in the SW 

Mediterranean and N-Saharic region. A. bourgaei is known only from a small area in SW 

Spain, and A. mauritiana, A. zaiànica, and A. monilicostata occur locally in Morocco and 

Algeria. The polymorphic A. stiparum is more widespread, being distributed throughout 

the arid areas of W Morocco and Algeria. A. stipa rum subsp. stipa rum and subsp. intermedia 

grow among steppe vegetation on the High Plains between Tell Atlas and Sahara 

Atlas. A. stipa rum subsp. sabulicola in the even more xeric, semi-desert habitats of the 

northern fringe of the Algerian Sahara. 

Vicariant distribution pattérns are also observed in the two other species groups, which 

are largely centred on N Africa: Anthemis ser. Chrysanthae and ser. Secundirameae. The 

five species of the Moroccan and NW Algerian A. ser. Chrysanthae show allopatric areas 

of distribution arranged in a well marked more or less linear sequence spanning NW Algeria 

(A . boveana, A. chrysantha), N Morocco (A. maroccana, A. gharbensis), and SW Morocco 

(A. tenuisecta). Like its mirror image, the N African members of A. ser. Secundiramea 

are distributed allopatrically between N Tunisia and N Egypt. The northernmost 

member of the series, A. secundiramea, covers a larger geographic area than the other 

species and shows a C Mediterranean distribution with some outposts in S Europe 

(Menorca, S France, Corsica, Sardinia). The elosely related, discoid species A. muricata is 

endemic to Sicily and, in contrast to A. secundiramea, is only found in inland habitats. The 

sequence of allopatric N African endemics of the seri es starts with A. ,ubensis in NE Algeria 

and N Tunisia, followed by A. confusa in C and S Tunisia, A. glareosa in W Libya,


A. kruegeriana, A. cyrenaica, and A. taubertii in E Libya, and A. retusa in N Egypt, while 

A. rigida is found in the Aegean archipelago. 

Anthemis sect. Maruta consists of 14 species and, according to Meusel & Jager (1992), 

inhabits an E Mediterranean-SW Iranian area with one species (A. cotula, incl. 

A. lithuanica) being archeophytically distributed throughout Europe and SW Asia. Two 

representatives of the section, A. cotula and A. pseudocotula, reach N Africa, the former is 

presumably introduced by Man into N Morocco (Tangier peninsula, Rif mountains, 

Melilla peninsula) and the vicinity of cities like Oran (Algeria), Sfax, and Gabès (Tunisia), 

the latter reaching the SW Iimit of its natural distribution in Cyrenaica. 

Anthemis subg. Cota, like sect. Anthemis and sect. Hiorthia, shows a concentration of 

its species in the SE European and SW Asian part of its distributional range. Most of its 

species are restricted to small areas in the submediterranean parts of Turkey and the adjacent 

Caucasus countries. However, some perennial members of A. sect. Anthemaria (e.g. 

A. tinctoria, A. triumfetti), along with monocarpic representatives of A. sect. Cota 

(A. altissima, A. austriaca), penetrate C and SW Europe. Since A. austriaca is known to 

occur synanthropically in ruderal plant cornrnunities of C Europe, its occurrence in N 

Africa, which is demonstrated in the present paper for the first time, is Iikely due to human 

activity. 

7. Modes of speciation within Anthemis in N Africa 

The N African representatives of the genus Anthemis offer an interesting insight into the 

evolutionary patterns realised in this genus. Most of the observed mechanisms also occur 

in groups of related taxa in other parts of the generic range, and in other genera of Anthemideae. 

As Ehrendorfer (1970) has pointed out, main changes of evolutionary patterns 

and strategies in this tribe are observed in connection with the switch from perennial to 

annuallife formo 

New data on N African taxa, along with published records by Uitz (1970), Ehrendorfer 

(1970), Mitsuoka & Ehrendorfer (1972), and Kuzmanov & al. (1981), show that Anthemis 

sect. Hiorthia is the only section of the genus where polyploidy has played a significant 

evolutionary role. Ali of the N African representatives of this section were found to exist 

at the tetraploid level. While A. abylaea, A. eretica subsp. columnae, and A. punctata are 

restricted to this level, in the widespread A. pedunculata, and according to pollen dimensions 

in some N African populations of A. maritima (in A. maritima subsp. bolosii) as 

well, diploids are found. These populations, along with the discoid A. alpestris which is 

endemic to the Iberian Peninsula, are the only diploid representatives of the section in the 

W Mediterranean area. 

While in Anthemis maritima morphological differences between (presumably) diploid 

and tetraploid representatives were found to be marked enough to recognise the two entities 

as separate subspecies, there seems to exist but little correlation between ploidy level 

and morphology in A. pedunculata. Multivariate statistical analyses of morphological 

variation along with results of molecular studies (see chapter 13) indicate that most of the 

tetraploids have to be considered to be of autotetraploidorigin. Though the difference in 

ploidy leve I may act as a rather effective crossing barrier, most of the tetraploid popula-

Bocconea 9 - 1998 17 

tions have not developed conspicuously deviating morphological characters yet that would 

permit their acknowledgement at any taxonomic level. As alone exception, tetraploid 

populations of the limestone mountains in the E parts of the Tangiers peninsula were 

found to deviate morphologically from other diploid and tetraploid representatives of 

A. pedunculata and are considered to represent an independent species, A. abylaea. 

As multivariate statistical analyses in the A. pedunculata - A. punctata complex 

(chapter 13) and in the A. boveana group (chapter 12) indicate, the geographical pattern of 

morphological variation in perennial and annual representatives of Anthemis in N Africa 

shows some marked differences. Variation in perennials occurs on a larger geographical 

scale, morphologically divergent forms are geographically isolated but connected by morphological 

intermediates, resulting in a c1inal pattern of morphological variation. In contrast, 

annual taxa show more restricted distribution areas, morphological variation within 

taxa is rather small as compared to these between taxa. Evolutionary divergence appears to 

be more rapid in the annual representatives of the genus than in the perenni al ones. As 

Wright (1940) and Stebbins (1952) pointed out, the population structure most favourable 

for rapid evolution is that of a large or medium-sized population divided into many small 

subunits which are largely isolated from each other, but can occasionally interchange 

genes. In such a set-up, the small subunits permit new genotypes to become stabilised 

through natural selection or genetic drift, while migration between colonies prevents the 

subunit's genetic stagnation and ensure genetic coherence of the large population for a 

certain period of time. Since such a population structure occurs more frequently in arid or 

semiarid regions than in areas with abundant moisture, Stebbins (1952) considers aridity to 

be a stimulus for plant evolution. 

A further acceleration of genetic and morphological divergence may be caused by the 

switch from self-incompatibility and allogamy in perennials towards self-compatibility and 

autogamy in annuals. The combination of annuallife form and autogamy enables plants to 

rapidly occupy unstable or pioneer habitats and to survive unfavourable seasons as diaspores. 

Populations of annuals have to be built up every year from a seed-bank and, as a 

consequence, are affected by rapid natural selection or gene tic drift through a so-called 

bottleneck effect (Runemark 1970, Nei & al. 1975), since only part of the gene pool of the 

parent population will be maintained in the next one when conditions are suboptimal. The 

repeated contraction of populations narrows down genetic and morphological variability 

within populations and increases discontinuities between them, because inbreeding and 

genetic drift may lead to the fixation of alleles different from those found in the parental 

populations, or at least to deviating allele frequencies. 

In contrast to the allopolyploid mode of speciation, the frequency of hybrid speciation 

on the diploid level, as hypothesised for Anthemis ubensis on morphological grounds in 

the present paper, is less well known (Rieseberg & Brouillet 1994) and its occurrence is 

often questioned (Grant 1981), although there is positive evidence for it in some plant 

groups (e.g. Helianthus, Rieseberg 1991; Iris, Arnold & al. 1990, 1991; Stephanomeria, 

Gallez & Gottlieb 1982). Since even intergeneric crosses between Anthemis species and 

representatives of Chamaemelum, Matricaria, and Tripleurospermum may produce fertile 

or semifertile offspring (Mitsuoka & Ehrendorfer 1972), the existence of fertile or semifertile 

hybrids involving two Anthemis species is plausible, even if they belong to different 

sections, as in the case of A. pedunculata and A. secundiramea, the presumed parent species 

of A. ubensis. As Grant (1981) set out, the centrai problems with diploid hybrid spe-


ciation are genotype segregation in the hybrid progeny as well as backcrossing, introgression, 

and consequent resorption of the hybrid genotype. Grant (1981) lists seven strategies 

that may overcome segregation of a hybrid genome and enhance its stability. Only two 

such strategies, however, are connected with sexual propagation and the homoploid nature 

of hybrids: speciation by recombination and isolation of the hybrids from the parent species 

by external factors. Speciation by recombination may occur in the progeny of a sterile 

or semi sterile hybrid formed by hybridisation of two parent species which differ by two or 

more chromosomal rearrangements, when recombination types evolve that are homozygous 

for these rearrangements; the resulting hybrid descendants are fertile and diploid, but 

genetically isolated from the parent species due to a chromosomal sterility barrier. Isolation 

may occur when new combinations of characters lead to a mechanical, ethological, or 

ecological crossing barrier between descendants and their parent species. In the case of 

A. ubensis, both possibilities exist, but evidence, which would require thorough molecular 

and cytological studies, is lacking. 

8. Delimitation of taxa - Concepts of species, subspecies, and varieties 

In the present revision of Anthemis in W and C North Africa the taxonomic categories 

of species, subspc;:cies, and variety are used to classify the encountered patterns of morphological 

variation at and below the species level. 

As noted by Jansen (1985), "much debate and little consensus exists among biologists 

concerning the definition of species." A number of different species concepts exist (see 

Rieseberg & Brouillet 1994: 24 for a comprehensive list), based on biological (e.g. the 

biological, better termed reproductive, species concept of Mayr 1969, 1970), phenetic 

(Cronquist 1988), ecological (Van Valen 1976), and phyIogenetic criteria (the phylogenetic 

species concept of Cracraft 1989 and the autapomorphic species concept of Mishler 

& Brandon 1987). 

A primarily morphoIogical or phenetic species concept has been used in the present 

study. As discussed in the previous chapter, morphological discontinuities are assumed to 

be due to genetic differences and, together with cytoiogical and molecular information in 

some taxa groups, to a certain degree are taken to reflect speciation processes and phylogenetic 

relationships. Since patterns of morphological variation show marked differences 

between different subgroups of the genus, the resulting species may not be equivalent. As 

noted by Lewis (1955, cited in Stuessy 1990:179), "the pattern ofmorphological differentiation 

may differ from one group of pIants to another and is a reflection of the diversity of 

evolutionary processes. Consequently, species and subspecies are not necessariIy equivalent 

in different genera or different sections of the same genus." As discussed above, the 

prevailing modes of speciation in perenniaIs of Anthemis sect. Hiorthia are hybridisation 

and polypIoidy, while in annuals of A. sect. Anthemis allopatric speciation and Iocal 

(founder effect) speciation (Levin 1993) is usually encountered. Therefore, species acknowledged 

in the former section tend to be wide-spread, poIymorphic, and often exhibit 

clinal variation of morphoIogical characters; while those of the latter section tend to be 

more local and Iess poIymorphic.

Bocconea 9 - 1998 19 

Both subspecies and variety were used as infraspecific categories to describe patterns of 

infraspecific morphological variation. Since little is known on genetic divergence and 

reproductive behaviour, only geographical information was used to decide on the rank of 

morphologically defined infraspecific taxa. Following suggestions made by Stuessy 

(1990), subspecies rank is used for a cohesive series of morphologically similar individuals 

forming allopatric or parapatric neighbour populations to other cohesive series of 

morphologically deviating populations, when morphologically intermediate individuals or 

populations occur along the contact zones. This subspecies concept, based on phenetics 

and geography, dates back to Wettstein (1898) who found that similar and closely related 

young taxa (subspecies) show allopatric distribution patterns, since their ecological demands 

are too similar to allow their sympatric co-existence. Only after a further divergence 

involving ecological demands (and often correlated with further genetic and morphological 

divergence) taxa (species) will be able to sympatrically co-exist. Subspecies, 

therefore, are considered by Wettstein (1898) to represent intermediate stages in an allopatric 

speciation process, which modern evolutionary biologists (e.g. Stebbins 1950, Grant 

1981, Ehrendorfer 1984) still consider to be the prevalent mode of speciation in plants. 

According to Rieseberg & Brouillet (1994), allopatric speciation by subdivision (or vicariance) 

rather results in monophyletic (sub)species pairs, while allopatric founder effect 

speciation may lead to a monophyletic derivative (sub)species and leave a paraphyletic or 

metaphyletic progenitor behind. As Crisp & Chandler (1996) have demonstrated, the 

avoidance of paraphyletic or metaphyletic taxa at and below the species level is neither 

realistic nor reasonable and may be theoreticallY impossible. Over and above, the occurrence 

of reticulate evolution in Anthemis due to polyploidization and homoploid hybrid 

speciation also argues against a strict cladistic species concept. 

While subspecies have often been regarded as regional facies of a species, varletIes 

were considered to represent a local facies (Du Rietz 1930, Stuessy 1990). In the present 

revision the rank of variety is used for morphologically deviating populations or population 

groups when all individuals are uniform but do not form a geographically cohesive 

population or population series. For example, varietal status was used for discoid representatives 

of Anthemis pedunculatll subsp. pedunculata which occur nearly throughout the 

range of the subspecies but show no geographical integrity because they are usually neighboured 

by uniformly radiate representatives. Since morphological distinctions are few and 

presumably under a simple genetic control, one may assume that discoid plants have 

evolved in different populations in parallel. The resulting variety, A. pedunculata var. 

discoidea, is therefore quite likely of polyphyletic origino 

The rank of forma, perhaps bes t applied to morphologically deviating individuals appearing 

sporadically within normal populations of a taxon (Stuessy 1990), has not been 

used in the present work. Some do formallY name such minor differences to draw attention 

to populational variation (Valenti ne 1975), others (Burtt 1970, Jansen 1985) consider this 

unnecessary nomenclatural clutter. I agree with the latter authors and will refer to minor 

morphological variants in the discussion of taxa while denying them formaI acknowledgement.


9. MorphologicaI features 

Duration and life form 

The majority of N African representatives of Anthemis are annuals belonging to A. sect. 

Anthemis, sect. Maruta, and sect. Cota. The annual habit may be interpreted as an adaptation 

to the dry semi-desert and steppe habitats of the adjacent areas of the Sahara, and to 

disturbed habitats or arable fields. A. stipa rum, A. confusa, or A. glareosa are prostrate 

annuals with reduced main axes and procumbent stems with smallleaves, found in natural 

stony and sandy habitats of the deserts. The vigorous growth of A. gharbensis and 

A. cotula, found as weeds in arable fields and along field margins in more hurnid areas, 

forms the opposite extreme habit of therophytic Anthemis in N Africa. 

Anthemis sect. Hiorthia contains short- to long-lived perenni al herbs which are restricted 

in N Africa to the hurnid areas of the coasts (A. maritima) or to the mountains of 

the Atlas ranges of Morocco, Algeria, and Tunisia (A. abylaea, A. pedunculata, 

A. punctata, A. eretica subsp. columnae). At least A. pedunculata is capable to flower and 

set fruit already in the first year, and especially diploid plants are often found growing as 

weeds in arable fields, so that the differences between annual and perenniallife form seem 

not to be too c1ear-cut in this species. However, in better growing conditions 

A. pedunculata is c1early a short-lived perennial, with annual shoots arising from a woody 

stock (pleiocorm) formed by the bases of the primary and subsequent annual shoots. 

Tetraploid plants of A. pedunculata, along with representatives of A. abylaea, A. punctata, 

and A. eretica subsp. columnae, also grow as pleiocormous hemicryptophytes. Observations 

in the field and herbarium suggest that most representatives of these species have 

life spans of three to five years. 

Roots and rhizomes 

The root system type is c10sely correlated with habit and duration. In ali annuals of 

Anthemis sect. Anthemis, sect. Maruta, and sect. Cota, taproots are found. The perennials 

in A. sect. Hiorthia are characterized by a woody caudex formed by the lignified bases of 

previous years' shoots fused with the basai taproot. In younger individuals the caudex may 

be unbranched and only partly subterranean, in older ones branched and completely subterranean 

caudices (pleiocorms) may be found. In the perennial species, especially in 

A. maritima, accessory roots bome at the basai stem nodes are frequently observed which 

are missing in annuals, even if procumbent. 

Growth habit, stems and capitulescence 

The most simple condition in the organisation of stem and disposition of capitula found 

in Anthemis is shown in Fig. lA: the mai n axis is erect and ends in a cymose capitulescence 

characterised by a terminai capitulum which flowers first. This type of capitulescence 

organisation is found in most N African annuals, e.g. A. austriaca, A. ubensis, or 

A. monilicostata. Unsuitable growing conditions like drought or malnutrition may considerably 

reduce the number of developing branches and lead to poor capitulescences of few 

or a single capitulum. 

A more complicated condition of stem and capitulescence organisation results when side 

axes are added beneath the main capitulescence (Fig. lB). Troll (1964) and Panero (1992) 

use the term parac1adia to designate these accessory flowering branches, each of which

Bocconea 9 - 1998 21 

resembles the main structural axis of the capitulescence. The number of parac1adia 

depends on growth conditions. In very robust annual species it is not uncommon that the 

main axis produces a paracladium from every leafaxil of the main axis, down to the basaI 

nodes (Fig. le; e.g. A. cotula). In some species, e.g. Anthemis gharbensis, a more or less 

B 

G 

H 

Fig. l . Growth forrns and capitulescence types in N African representatives of Anthemis (see text 

for explanation).


conspicuous inhibition zone is found beneath the capitulescence of the main axis, and side 

axes will arise from the basaI nodes of the main axis after its capitulescence has flowered 

(Fig. ID). For convenience, such side axes are termed stems in the present revisiono 

Some annual N African representatives of Anthemis (e.g. A. tenuisecta, A. confusa, or 

A. glareosa) are characterised by a prostrate habit. Their main axis is reduced and its basaI 

intemodes are extremely short (Fig. lE). Reduction of the main axis is often so complete 

that branching occurs directly from the taproot (Fig. lF). The larger and usually petiolate 

lower cauline leaves that normally make up the basaI rosette are almost totally missing; 

most leaves are sessile and resemble upper cauline leaves of plants with a normally developed 

main axis. This character may be used to distinguish between N Tunisian A. ubensis 

and the closely related C Tunisian A. confusa: While A. ubensis has an erect habit with a 

main axis and leaves that successively decrease in size and dissection from base to top, 

A. confusa is characterised by prostrate habit and poorly dissected leaves. 

In perennial N African representatives of the genus, the trend towards reduction of the 

main axis is well marked, but realised in a different way. The current year's shoots arise 

from the lignified bases of primary and subsequent shoots of the last years (Fig. lG). 

Sometimes the main axis is basally more or less completely fused with the taproot to form 

a ± subterranean rootstock (caudex) that often bears both non-flowering and flowering 

shoots (Fig. lH). The capitulescences are often reduced to a single capitulum, and rarely 

more than three (up to lO) capitula. 

The stems are usually herbaceous in their distaI part but conspicuously lignified basalIy, 

particularly in perennials. Stems formed by the main axis are usualIy straight basally, 

while those corresponding to side branches or bome on a rootstock are basalIy bent upwards. 

In species with a procumbent habit, only the capitulescence or single peduncles are 

lifted above the ground. In alI cases, the stems are round in cross-section and somewhat 

sulcate. They are usually green, often tinged with redin the lower half. 

In Anthemis maritima in particular, and to a lesser extent in the other perennial representatives 

of A. sect. Hiorthia, the lower stem nodes may be beset with accessory roots. 

Stem length varies considerably with growth conditions: in the perennials of Anthemis 

sect. Hiorthia it is 20-40 cm, in representatives of A. sect. Anthemis and sect. Maruta 

usualIy up to 25 cm, but under favourable conditions sometimes much more (e.g. in 

A. gharbensis 60 cm, A. cotula 90 cm). Accordingly, the diameter of stem bases may vary 

between less than 1 mm in poorly developed plants up to 3 or occasionalIy 6 mm in vigorously 

growing annuals and some perennials. 

Indumentum 

The indumentum in most N African representatives of Anthemis consists of biseriate 

glands and medifixed covering hairs, as already found by Napp-Zinn & Eble (1980). 

Medifixed hairs are usually bome on a stalk formed by 2-4 (sometimes even up to 7) 

more or less isodiametic celIs and a large, strongly elongated apical celI that is symmetrically 

or asymmetricalIy suspended at the tip of the stalk. While the stalks may reach a 

length of c. 50-80 !lm, the length of the apical celI is usualIy found to be 400-1000 !lm. In 

some species, extremely asymmetrically medifixed apical cells are found. Sometimes, 

especialIy on peduncles and involucral bracts, asymmetry is so marked that the apical celI 

is no longer medifixed and an uniseriate flagellar covering hair is formed. These hairs are

Bocconea 9 - 1998 23 

referred to as basifixed hairs in the descriptions. However, their presence seems to have 

little taxonomic relevance. 

GIands are usualIy found on alI overground parts of plants, even on pales, ray florets, 

disc florets and achenes. Their structure and shape is very similar in alI species studied. 

They are formed by two parallel series of 3-4 isodiametrical celIs and have a knob-like 

appearance. They are found in alI N African representatives of Anthemis and do not have 

taxonomic significance. 

Covering hairs are usualIy found on stems, Ieaves, peduncles, involucres, and sometimes 

on the peripheral pales (Anthemis mauritiana, A. monilicostata). Covering hairs on 

disc florets and tubular parts of ray florets occur in some species of Anthemis, but none 

were found N African representatives of the genus. 

In most N African Anthemis species, the stems are sparseIy to denseIy appressed-hairy, 

with medifixed hairs and interspersed glands. However, in A. maritima, A. pedunculata, 

A. punctata subsp. kabylica, A. secundiramea, and A. cotula completely glabrous stems 

were encountered. 

In leaves, biseriate glands are sunk in pits on the upper and lower surface. The density 

of the non-glandular leaf indumentum is extremely variable and has littIe taxonomic significance. 

CompleteIy glabrous leaves were encountered in Anthemis maritima, 

A. pedunculata, A. punctata subsp. kabylica, A. confusa, and A. secundiramea. AlI other 

species have sparsely to rarely densely hairy leaves. A. eretica subsp. columnae is characterized 

by an extremely dense, sericeous leaf indumentum. 

Peduncles are aIways furnished with at Ieast some covering hairs, except in Anthemis 

maritima subsp. bolosii where even peduncles are devoid of covering hairs. 

The involucral bracts (at least the outer and middle ones) of most species studied are 

covered with long hairs, but are completeIy glabrous in Anthemis maritima subsp. bolosii 

and subsp. pseudopunctata (those of A. maritima subsp. maritima are usualIy sparsely to 

densely hairy) , and, contrary to the other members of A. ser. Bourgaeinianae, also in 

A. monilicostata. Interspersed biseriate gIands are present in alI taxa. Inside, the of invoIucral 

bracts are usualIy glabrous, except in some plants of A. ubensis where some hairs 

were found. 

Leaves 

The Ieaves . are consistently alternate, though sometimes forming basaI rosettes. The 

shape, size and dissection of the leaves may vary considerably depending on their position. 

Rosette Ieaves'and Ieaves from the basaI parts of erect stems are usualIy narrowly elIipticaI 

to narrowly obovate in outline, while Ieaves from the upper parts of erect stems and Ieaves 

of prostrate sterns tend to be more ovate in outline. This is correlated with the tendency of 

basaI and Iower cauline Ieaves to be petiolate, while the upper cauline Ieaves are usualIy 

sessile. 

In most of the species studied, the base of the petiole bears severai pairs of Iaterai teeth. 

In basaI Ieaves primary Iobes differ markedly from the basaI teeth in size and dissection: 

basaI teeth are rather smalI, entire or pinnatisect; primary Ieaf Iobes are Iarger, 1-3-pinnatisect 

or -pinnatipartite. AIso, basaI teeth and Ieaf Iobes are welI spaced and separated by 

the petiole. In the upper half of the stem Ieaf petioles become more and more reduced, 

basaI teeth and Ieaf Iobes become closer to each other, and as basaI teeth are Iarger and 

Ieaf-Iobes Iess dissected, upper Ieaves are sessile in appearance. While teeth at the leaf


base are found in all perennial and most annual species, they tend to be absent in Anthemis 

boveana, A. chrysantha, and A. gharbesis of A. ser. Chrysanthae, and in most of the 

members of A. ser. Bourgaeinianae. 

Leave texture is usually herbaceous, but succulent leaves are common in species from 

maritime habitats (e.g. Anthemis maritima, A. secundiramea, A. chrysantha, A. mauritiana). 

Dimensions of leaves are variable and depend on growth conditions. However, in the 

Anthemis pedunculata - A. punctata group of species they may help in the determination 

of taxa: basalleaves in A. pedunculata are usually up to 45(-75) mm long and up to 20(-30) 

mm wide, those of A. punctata are conspicuously longer (up to 115 mm) and wider (up to 

53 mm). A. abylaea is also characterized by rather long (up to 80 mm) and wide (up to 35 

mm) basalleaves. 

The dissection of leaf blades proves to be taxonornically relevant, its degree depending 

on the position of the leaf on the stem. As a rule, the most basaI leaves are less strongly 

dissected than the lower cauline ones, and further upward leaf dissection is again gradually 

reduced. Therefore, differences in Ieaf dissection are particularly important in the lower 

cauline leaves. In most N African Anthemis taxa the lower leaves are pinnatisect with 

primary lobes cut right to the axis, but in A. maritima subsp. bolosii and some plants of 

A. punctata subsp. kabylica they are pinnatifid or pinnatipartite. 

2- to 3-pinnatisect or -pinnatipartite leaves are found in most species. The shape of ultimate 

leaf segments may vary within taxa from broadly triangular to narrowly elliptical or 

even linear, and their dimensions also varies. No clear-cut discontinuities between taxa 

could be observed, but some tendencies are sufficiently marked to be used as distinguishing 

features. For example, Anthemis pedunculata usually has 1.0-5.5 mm long and 0.3-1.8 

mm wide ultimate leaf segments, but those of A. punctata (1.4-9.0 mm x 0.6-2.3 mm) and 

A. eretica varo columnae (1.7-20.0 mm x 0.6-2.0 mm) are considerably larger. In most 

species, ultimate leaf segments end in a mucro of 0.1-0.3 mm length. No mucronate leaf 

segments were observed in some populations of A. secundiramea. 

Peduncles 

In the present study, the peduncle is defined as the distaI portion of the stem subtending 

a capitulum, from the most distaI leaf on this axis to the base of the involucre. While 

peduncle length was found to be very variable in all species examined and therefore unhelpful 

for the characterisation of taxa, the inflation of peduncles at maturity proved to be 

relevant. In the N Mrican representatives of Anthemis subg. Cota (A. austriaca) and of 

A. sect. Hiorthia the peduncles usually remain slender or become only slightly inflated at 

maturity. In A. sect. Anthemis and sect. Maruta, strongly inflated peduncles are found in 

A. arvensis subsp. incrassata of A. ser. Anthemis and in A. chrysantha of A. ser. Chrysanthae. 

In A. secundiramea of A. ser. Secundirameae, this character is used to distinguish the 

two N African varieties of that species, varo secundiramea with and varo cossyrensis without 

conspicuously inflated peduncles. A further representative of this series, A. ubensis, 

and some populations of A. glareosa al so show incrassate peduncles. In contrast, all representatives 

of A. ser. Bourgaeinianae have peduncles remaining slender at maturity. In 

A. sect. Maruta, this character can be used to distinguish A. cotula (slender peduncles) 

from A: pseudocotula (inflated peduncles).

Bocconea 9 - 1998 25 

Capitula 

Most species are gynomonoecious, with a single row of female ray florets and a centrai 

mass of perfect, hermaphrodite disc florets. Monoecious plants with ali florets tubular and 

hermaphrodite are found in Anthemis pedunculata var. discoidea, A. cyrenaica, and 

A. kruegeriana. In the latter two taxa, however, radiate capitula are also found, sometimes 

on the same plant as discoid ones. In these cases, ray florets are minute (up to c. 5 mm 

long) but still fertile, and have sometimes been called hemiligules (e.g. Benedf i Gonzalez 

1987). 

The diameter of capitula (including ray florets) is a useful criterion in the Anthemis 

pedunculata - A. punctata group of species where in radiate representatives of A. pedunculata 

it is 6-37 mm, while A. abylaea (30-50 mm) and A. punctata (30-55 mm) have larger and 

more showy capitula. 

The number of ray florets is correlated with the dimensions of capitula. In small-headed 

species it often scatters around the Fibonacci numbers 8 and 13, while in other species 

even numbers between 13 and 21 are preferred. In A. abylaea and A. punctata capitula 

with up to 23 ray florets were observed. 

Involucre and involueral bracts 

The involucres of N African representatives of Anthemis usually have the shape of 

hemispherical cups. Only in A. austriaca and A. secundiramea a tendency towards a more 

conical shapes was observed. The dimensions of involucres are usually correlated with the 

diameter of capitula and are very variable within taxa. Taxonomically relevant differences 

in the diameter of involucres may be observed in the A. pedunculata - A. punctata complex 

(A. sect. Hiorthia): A. pedunculata is characterised by smallish involucres with diameters 

of 6-14 mm, the closely related A. abylaea and A. punctata have involucres with 

13-21 mm and (10-)14-22 mm in diameter, respectively. In ali other N African representatives 

of the genus the dimension of involucres fall within the range indicated for A. pedunculata. 

In most species studied, involucres, most conspicuously in Anthemis punctata, 

A. chrysantha, A. cyrenaica, and A. kruegeriana, become umbonate at maturity. However, 

taxa with conspicuously inflated peduncles andJor conical involucres like A. arvensis 

subsp. incrassata, A. secundiramea varo secundiramea, or A. austriaca usually have attenuate 

or only very slightly umbonate involucres. 

The involucre consists of free involucral bracts (sometimes called phyllaries). In ali 

species, the involucral bracts are unequal in size and shape and arranged in 3-4 imbricate , 

layers and in a series of spirai parastiches (here referred to as "rows':). In Anthemis eretica 

subsp. columnae and A. cyrenaica up to 5 layers of involucral bracts may be observed. For 

convenience, involucral bracts of different positions in the parastiches will be referred to 

as "outer", "middle" and "inner" involucral bracts. 

The involucral bracts are variable in texture, but there is always a centraI, conspicuously 

thickened portion that is several celi layers thick and a marginaI, membranous portion 

of only one celi Iayer, referred to as "membranous margin". The centraI part may be 

rather thin and almost membranous in texture in Anthemis cotula, thickish and somewhat 

fleshy as in A. maritima, or hard and tough as in A. austriaca, and usually consists of a 

green longitudinal strip surrounding the midvein of the bract and laterally adjacent stramineous 

portions. At maturity, at least the base of this centrai part of the .Ìnvolucral bract


becomes indurated, and sometimes, as in A. austriaca, A. maritima, or A. chrysantha, the 

whole bract is sclerified. 

While the outer involucral bracts are invariably triangular or ovate in outline, the middle 

and inner ones show differences in shape between the different taxa, ranging from 

narrowly elliptical to narrowly obovate. Such differences are particularly useful to discriminate 

between the subspecies of Anthemis pedunculata and A. punctata: A. pedunculata 

subsp. pedunculata and subsp. clausonis, along with A. punctata subsp. kabylica 

have narrowly elliptical inner involucral bracts, those of A. peducnulata subsp. atlantica 

and A. punctata subsp. punctata are conspicuously broader and narrowly obovate in outline. 

This is mainly due to the membranous margins being narrow in narrowly elliptical but 

wider in narrowly obovate involucral bracts. 

Although variable in some taxa (e.g. Anthemis confusa), the colour of the membranous 

margins of involucral bracts is a go od criterion in others. In A. maritima subsp. maritima 

and subsp. bolosii the membranous margin is usually pale and only at the tip it is sometimes 

tinged with brown or black, while in A. maritima subsp. pseudopunctata it is brown 

or black throughout. In A. pedunculata subsp. pedunculata involucral bracts usually have 

brown and in A. pedunculata subsp. clausonis invariably pale membranous margins, in 

A. pedunculata subsp. atlantica the margins are brown or black in the outer but pale in the 

inner Ìnvolucral bracts, causing a somewhat two-coloured appearance of the involucre. 

A. ubensis is easily distinguished from the rather similar A. secundiramea varo cossyrensis 

by its involucral bracts with brown, as opposed to pale, membranous margins. 

As a matter of course, dimensions of involucral bracts correlate with those of the involucre 

and capitulum. The longest middle and inner involucral bracts are therefore found 

in Anthemis abylaea (to 8.0 mm) and A. punctata (to 8.2 mm) and may be used to distinguish 

these species from the related A. pedunculata. In the A. boveana group of species, 

bract size was also found to have some taxonomic importance (see chapter 12). 

Receptacle 

The receptacles of N African Anthemis taxa are usually slightly convex to hemispherical 

in flower, but considerable variation between taxa is found in mature capitula. In 

A. austriaca, the sole N African member of A. subg. Cota, the receptacle remains slightly 

convex during maturation, whereas in most members of A. subg. Anthemis receptacles 

elongate, the degree of elongation characterising the different sections. The perennials of 

A. sect. Hiorthia usually have hemispherical to ovate or conical receptacles, in A. sect. 

Anthemis receptacles elongate more strongly at maturity and usually become conical or 

even, in some species, narrowly conical to narrowly cylindrical in shape (e.g. in 

A. confusa, A. secundiramea, and A. glareosa of A. sect. Secundirameae). In A. arvensis 

subsp. incrassata, A. secundiramea var. secundiramea, or A. ubensis, where peduncles 

become strongly inflated at maturity, this also affects the shape of the receptacles which 

become basally broadened and somewhat pyriform in outline. The two studied 

representatives of A. sect. Maruta, A. cotula, and A. pseudocotula, are also characterised 

by very slender and elongated, conical to narrowly conical or even cylindrical-fusiform 

receptacles. 

Receptacles are paleate throughout in ali species of the present ,revision, i.e. each disc 

floret is subtended by a bract (pale, receptacular scale), with the Ione exception of An-

Bocconea 9 - 1998 27 

themis cotula where pales are restricted to the apical half of the receptacIe, while the peripheral 

disc florets are devo id of subtending bracts. 

In ali representatives of Anthemis in N Africa, receptacIes were found to be filled with 

spongy tissue with large intercellulars; hollow receptacIes as found in Matricaria are absent. 

Pales 

The pales or receptacular scales offer a number of characters that are diagnostic of single 

taxa, species groups, series, or sections. In ali N African taxa, pales were found to have 

a single vascular bundle. The shape of pales varies conspicuously between taxa: Anthemis 

cotula and A pseudocotula of A. sect. Maruta, like ali other members of this section, have 

very narrow, subulate pales (up to 0.2-0.3 mm wide) comprising little more than the vascular 

bundle and the adjacent scIerenchymatous tissue. In ali other species, the midvein of 

the pale is bordered by a flimsy and scarious membrane formed by strongly elongate cells. 

Usually, this membrane is only one cell-Iayer thick and stramineous (red-tinged in some 

populations of A confusa, A. glareosa, and A. kruegeriana). In A. pedunculata and 

A. punctata the shape of the pales is diagnostic of subspecies: pales are narrowly linear to 

almost subulate in A pedunculata subsp. pedunculata but narrowly elliptical to narrowly 

obovate in A pedunculata subsp. atlantica; in A punctata subsp. punctata pales are narrowly 

obovate, and in A punctata subsp. kabylica narrowly linear. The width of pales also 

contributed to the circumscription of subspecies in A stiparum, where subsp. intermedia 

and subsp. sabulicola have rather narrow (0.2-0.6 mm) and subsp. stipa rum broader (0.4- 

1.2 mm) pales. The broadest pales were observed in A. maritima subsp. maritima (1.2-1.7 

mm), A chrysantha (1.1-1.8 mm), A. punctata subsp. punctata (0.75-1.5 mm), and 

A. ubensis (0.8-1.6 mm). In these taxa, pales are usually narrowly elliptical to narrowly 

obovate. 

A further valuable character diagnostic of taxa is the shape of the tip of the pales. In 

most species it is formed by the protruding midrib, but in some species of Anthemis ser. 

Chrysanthae and ser. Secundirameae the midrib does not reach the apex, which is usually 

membranous, often tinged with yellow and somewhat hooded owing to the slightly inrolled 

margino In A ser. Chrysanthae this character is useful to discriminate between the Algerian 

taxa (A boveana and A. chrysantha, with a protruding midrib) and the Moroccan ones 

(A . maroccana, A gharbensis, and A tenuisecta, with membranous tips). In A. ser. Secundiramea 

the Libyan A kruegeriana and A glareosa are easily distinguished by the same 

character from their Tunisian relatives A secundiramea, A ubensis, and A confusa. Pales 

with a membranous tip are acuminate to bluntly rounded, in A gharbensis and A glareosa 

they even may be truncate or emarginate. 

There are two types of pales with excurrent midrib among the N African representatives 

of Anthemis sect. Hiorthia: A. abylaea, A. pedunculata subsp. pedunculata and subsp. 

clausonis, A. maritima subsp. pseudopunctata, and A. punctata subsp. kabylica have pales 

which gradually taper into the tip, whereas the pales of A pedunculata subsp. atlantica, 

A. maritima subsp. maritima and subsp. bolosii, and A punctata subsp. punctata taper 

rather abruptly into a tip that may be tricuspidate due to the apical projecting lateral margins. 

In some cases the lateral tips are even overtopping the centraI tipo In Acretica subsp. 

columnae, A. arvensis subsp. incrassata, and A. cyrenaica the tips are often abruptly 

acuminate and erose.


A further character of some taxonomic importance is the colour of the pale tips. In taxa 

of A. sect. Hiorthia they are often tinged with dark brown or even black (e.g. 

A. pedunculata, A. eretica subsp. columnae, A. punctata), but such tinge is lacking in 

populations of A. maritima and some plants of A. punctata subsp. punctata. Pales of 

A. monilicostata and some populations of A. ubensis also end in blackish tips. 

The strength of attachment of pales on the receptacles is also a useful character. In most 

of the species studied, at least the peripheral pales persist on the receptacles even after the 

dissemination of achenes. In A. sect. Bourgaeinianae, however, pales fall off very easily, 

so that after dissemination receptacles are left completely naked. 

The hypothesis of Stuessy & Spooner (1988), that one of the functions of pales may be 

the protection of flowers, ovaries, and achenes from apical and lateral insect attack, is 

plausible though the adaptive significance may be less obvious in Anthemis than in the 

Heliantheae they had studied. 

Ray florets 

Ray florets of N African Anthemis have few taxonomically relevant characters. Corollas 

are divided into a lower tubular part and an upper, flat corolla limb. The tubular part is 

usually flattened dorso-ventrally but may get somewhat inflated and winged at maturity. In 

most species the tubular part and the abaxial face of the limb bear biseriate glands. Four 

vascular bundles are found in the tubular part of the ray florets. In the proximal region of 

the limb, several minor traces that end blindly split off from these main bundles which 

continue up to the 3-lobed apex of the limb where they anastomose. Baag0e (1977) studied 

four representatives of the genus and found that the adaxial epidermis of their ray florets 

corresponds to her "helianthoid type", consisting of more or less isodiamterical and papillose 

cells without thickened walls. Contrary to her indications, the corolla glands are not 

one-celled but correspond to the typical biseriate glands formed by usually eight cells that 

are found all over the vegetati ve and fiorai parts of all Anthemis species. 

The length and the colour of ray floret limbs may be taxonomically relevant. Most radiate 

species are characterised by white ray florets, but those of Anthemis ser. Chrysanthae 

are yellow, ranging from a saturated canary yellow (9A in Anon. 1966) in A. boveana, 

A. chrysantha, A. maroccana, and A. tenuisecta to a pale yellow (3A in Anon. 1966) in 

A. gharbensis. White ray florets are usually monochromous during flowering, but tend to a 

brownish tinge at the base of the limb when withered. Limb length is diagnostic for some 

closely related taxa, for instance A. maritima subsp. maritima (7.5-12.0 mm) and subsp. 

bolosii (3.8-8.0 mm); A. boveana (7.0-14.0 mm) and A. chrysantha (2.1-7.0 mm); and 

A. stipa rum subsp. stiparum (7.2-11.5 mm) or subsp. intermedia (7 .0-10.5 mm) and subsp. 

sabulicola (4.5-8.5 mm). The width of ray floret limbs being fairly constant, variation in 

length usually also affects the shape of the limb, which ranges from nearly circular to narrowly 

elliptical or narrowly oblong. 

Usually the florets are female and furnished with a style. In contrast to disc florets, the 

style branches are often laterally enlarged and devoid of sweeping hairs, having lost their 

function for pollen presentation. In Anthemis cotula, however, some plants with sterile ray 

florets lacking a style were observed, along with others in which they are pistillate and 

even fertile. Occasionally, in most species anther rudiments can be found in the tubular 

part of the ray florets.


Fig. 2. Micrographs of details from Anthemis flowers. - A: A. tenuisecta subsp. tenuisecta, longisection 

through the basaI part of a disc floret showing long and interwoven celi rows causing the 

inflated and spongy appearance at maturity (Podlech 45076; cult. in HB Berlin No. 049-26-93-10). 

B: A. gharbensis, vascularisation ofbasal part of disc floret: A = anther trace, C = corolla trace, S = 

style trace (Vogt 10161 & Oberprieler 4609). C: A. gharbensis, apical anther appendage (Vogt 

10161 & Oberprieler 4609). D: A. gharbensis, endothecial tissue; an asterisk denotes a celi with 

evenly arranged wall thickenings (Vogt 10161 & Oberprieler 4609). E: A. gharbensis, filament 

collar (Vogt 10161 & Oberprieler 4609). F: A. gharbensis, vascularisation of an immature achene; 

asterisks denote the five vascular bundles (Vogt 10161 & Oberprieler 4609). - Scale bars = 0.1 mm 

(SEM, top Ieft, and LM).

Bocconea 9 - 1998 31 

The styles are also very uniform throughout the genus. The base of the style becomes 

usually swollen and hardened due to enlarged and sclerified cells. Two vascular bundles 

run through the cylindrical and glabrous style shaft and end in the stigmatic branches, 

which are truncate-penicillate, bearing two separate lines of stigmatic papillae on the inside 

and obtuse sweeping hairs at their apex. Usually the stigmatic branches contain elongate 

resin ducts with a brown or orange content. 

Achenes 

Like the vast majority of Compositae, Anthemis has dry, monosperrnous fruits developed 

from a bicarpellate coenocarpic inferior ovary, with the seed testa adhering to but not 

coalescent with the pericarp. Although the ovary of Compositae is inferior, the fruit wall is 

generally terrned pericarp, and its structures usually been described in the terrns of epi-, 

meso- and endocarp, by analogy to the fruit wall in a superior ovary. Following Wagenitz 

(1976) and Roth (1977), the terrn "achene" as defined by Candolle (1813) is used in the 

present work for this type of fruits. The term "cypsela" was introduced by Mirbel (1815) 

and is frequently used in Anglo-Saxon and Nordic synantherologicalliterature (e.g. Humphries 

1979, Kiillersj6 1991, Anderberg 1991a, b, c, Bremer & Humphries 1993, Bremer 

1994). However, exceptions are found (e.g. Hansen 1991), and in Anglo-American literature 

the term "achene" still prevails (e.g. Stuessy 1968, Funk 1982, Jansen 1985, Panero 

1992). 

Dating back to the thorough studies of Briquet (1916), the paramount importance of 

achene morphology and anatomy for the taxonomy of Anthemideae and for the delirnitation 

of genera in this tribe is well known. In Anthemis the achenes provide some very 

important diagnostic features for the subdivision of the genus and the circumscription of 

species groups. 

Achenes of ray florets were found to be rather similar in most studied representatives of 

Anthemis subg. Anthemis. They are usually subcylindrical or fusiforrn in shape and slightly 

bent inward. Being tightly confined to the trigonal space between the achenes of peripheral 

disc florets and their pales and the flat inner surface of the inner involucral bracts, they are 

roundish to triangular in cross section, with a flattish abaxial and a rounded or edged adaxial 

face. The achenes of ray florets usually have (8-)9( -lO) ribs which are rather indistinct 

on the adaxial but more prorninent and in some species even tuberculate on the 

abaxial face. Like the achenes of disc florets, they are usually beset with mucilage cells on 

their ribs and by glands in the intercostal furrows. The apical plate is usually bordered with 

a ridge that in some species is adaxially protracted into a membranous auricle up to 0.3 

mm long. The achenes of ray florets in A. austriaca (A. subg. Cota), unlike those in 

A. subg. Anthemis, strongly resemble achenes of disc florets in being conspicuously flattened 

dorso-ventrally. 

The achenes of disc florets provide a greater number of characters of use for species 

delirnitation and supraspecific classification than the achenes of ray florets. While essentially 

homomorphic, the achenes'of disc florets tend towards heteromorphy in most of the 

species due to overall structural reduction from the larger peripheral to the smaller centraI 

achenes. In some species, the differences between achenes of marginaI and centraI disc 

florets are particularly well marked: In Anthemis mauritiana and A. arvensis the achenes 

of peripheral disc florets are extremely stout and subcylindrical or obconical, while the 

centraI achenes are much more slender and obconical-obovoidal. In most N African An-


themis taxa the peripheral disc achenes persist on the receptac1es at maturity, while the 

centrai ones fall off readily. In A. pseudocotula even the centrai achenes persist, and the 

whole capitulum forms a compact, cone-like dispersal unit (synaptospermy). 

Achenes of Anthemis austriaca differ markedly from those of all other representatives 

of the genus in N Africa. As in other species of A. subg. Cota, they are dorso-ventrally 

flattened and rhombic in cross section. While the lateral ribs are very pronounced and 

acutely angled, the 3-5 ribs on the adaxial and abaxial faces are inconspicuous. In A. subg. 

Anthemis disc achenes are never dorso-ventrally flattened but more or less terete and provided 

with (9-)10(-11) conspicuous, longitudinal ribs (except in A. maritima, A. eretica 

subsp. columnae, and A. punctata subsp. punctata where they are usually tetragonal in 

cross section and inconspicuously ribbed). The ribs may be sharply angled as in 

A. gharbensis or rounded and shallow as in A. cyrenaica and A. taubertii. 

The sculpturing of the achene surface is a very important character taxonomically. Most 

of the N African representatives of Anthemis have tuberculate achenes, and the degree of 

tuberculation may vary considerably within most taxa. Nevertheless, this character may be 

of use in several species groups. Achenes of A. maritima subsp. maritima, A. punctata 

subsp. punctata, and A. eretica subsp. columnae are usually smooth, those of A. maritima 

subsp. bolosii and A. punctata subsp. kabylica, as in the other N African representatives of 

A. sect. Hiorthia, are tuberculate. In A. ser. Chrysanthae, A. gharbensis, A. tenuisecta 

subsp. jahandiezii, and some populations of A. tenuisecta subsp. tenuisecta have rather 

smooth achenes, in all other taxa of this series achenes are strongly tuberculate. 

A. taubertii and A. cyrenaica of A. ser. Secundiramea have smooth to feebly tuberculate 

achenes, those of the superficially very similar A. glareosa and A. kruegeriana are prominently 

tuberculate. A. cotula of A. sect. Maruta has strongly tuberculate achenes, those of 

A. pseudocotula are smooth or moderately tuberculate. Extremely tuberculate achenes are 

sometimes found in A. pedunculata, but other populations of this species have moderately 

tuberculate or nearly smooth achenes. 

An important character observed in all species is the shape and size of mucilage celli; of 

the achene surface. These specialised epidermal cells are usually confined to the ribs or 

tuberc1e tips (Fig. 3A). In most species mucilage cells are small, isodiametrical or only 

moderately elongate longitudinally, but in Anthemis monilicostata they are drastically 

enlarged and protrude from the achene surface (Fig. 39). Achenes of A. gharbensis are 

characterised by strongly elongate mucilage cells that form continuous slime ribbons on 

the ridges of the ribs (Fig. 3B). Additionally, in most species with ribbed achenes sessile, 

biseriate glands are found in the intercostal furrows (Fig. 2F, 3B, 3C). 

In most species the apical plate is flat or slightly convex, circular or quadrangular in 

outline, and bears a relatively large, bowl-shaped discus ("coronet" in KynClova 1970, 

"Nektarium" in Vogt 1991, "stylopodium" in Bremer & Humphries 1993 and Kiillersjo 

1991) in its centre. In Anthemis tenuisecta subsp. jahandiezii the apical plate is strongly 

convex, in A. austriaca it is rhombic. The apical plate is bordered by a corona ("border'.' in 

Kynclova 1970), which in some species (e.g. A. arvensis, A. cotula, A. pseudocotula) may 

be an inconspicuous, rounded and crenulate rim formed by the apically protruding ribs of 

the achene, but in others is adaxially protracted into an up to l mm long, membranous 

auric1e.

Bocconea 9 - 1998 33 

Fig. 3. SEM micrographs from Anthemis achenes. - A: A. maroccana subsp. maroccana, epicarp 

with mucilage celi (Jahandiez 159). B: A. gharbensis, epicarp with glands and extremely elongated 

mucilage cells (Podlech 43409, cult. in HB Berlin No. 049-27-93-10). C: Anthemis tenuisecta 

subsp. tenuisecta, longisection of achene wall showing epicarpic cells with a strongly wrinkled 

cuti cui a and a longisectioned gland (Podlech 45076; cult. in HB Berlin No. 049-26-93-10). D: 

A. mauritiana subsp. faurei, epicarpic cell filled with sand of needle-like ca1cium oxalate crystals 

(Vogt 10872 & Oberprieler 5320). E: A. austriaca, epicarpic cells with single, large, octahedral 

crystals (Vogt 13151 & Oberprieler 7456). F: A. altissima, epicarpic cells with single, large, 

prismatic crystals (HB Liège, cult. in HB Berlin No. 194-17-92-10). - Scale bars = 100 flm (A-C, 

E-F) or lO flm (D).


The epicarpic cells (i.e. the epiderrnis of the pericarp) of N African species usually have 

strongly cutinised outer walls and wide celi lurnina. In transverse section the cells are 

isodiametric, while they are usually elongate and rectangular in longisection. In ali cases 

studied, the epicarpic cells remain unsclerified and parenchymatous. Often they were 

found to collapse during maturation. Marked differences in shape and number of ca1cium 

oxalate crystals in the epicarpic cells were observed. In Anthemis austriaca, the single 

studied representative of A. subg. Cata, cells contain single, large, octahedral crystals (Fig. 

3E); in ali species of A. subg. Anthemis, epicarpic cells contain crystal sand formed by 

large numbers of needle-like crystals (Fig. 3D). While in the case of A. austriaca the 

crystals are responsible for the rninutely punctulate macroscopic appearance of achenes, 

the crystal sand in the other subgenus causes an more or less intensive whitish tinge of 

achenes. 

The mesocarp exhibits considerable variation, its thickness and degree of sclerification 

showing specific differences. The sclerenchymatic tissue is formed by two kinds of elements, 

libriform fibre cells and sclereids, which can be best distinguished in longisection. 

Libriform fibre cells are strongly elongate longitudinally, have strongly sclerified cell 

walls with small pits, and extremely restricted cell lurnina with large ca1cium oxalate 

crystals at maturity. Sclereids are usually much larger, more or less isodiametric in cross 

section, moderately elongate longitudinally, and less sclerified. Initially parenchymatous, 

these cells become progressively sclerified during maturation due to continuous thickening 

of their walls, so that sclereids eventually have thick, concentrically multi-layered cell 

walls, but retain large pits and cell lurnina not as restricted as in fibre cells, which only 

occasionally contain ca1cium oxalate crystals. 

Vascularisation of the achene wall is very sirnilar in all species studied. About lO 

(Anthemis subg. Anthemis) or more (A. subg. Cata) sclerenchymatic ribs are formed in the 

mesocarp of the achenes, but only 5 vascular bundles can be found (Fig. 2F), contrary to 

the indications ofWagenitz (1968), Kynclova (1970), and Reitbrecht (1974) who assumed 

that each rib of the achenes has its vascular strand, but in accordance with the findings of 

Rumphries (1979). Especially in young, immature achenes it is easy to observe that only 

every second sclerenchymatic bundle in the pericarp contains one or several vessel elements 

with spirally arranged wall thickenings. 

In most species studied, libriform fibre cells, in cross sections of achenes, form a continuous 

ring at the inner side of the mesocarp. This ring is 1-5 celllayers thick between the 

ribs, but conspicuously thicker in the vascularised and non-vascularised ribs. Between this 

inner layer of sclerenchyma, formed by fibre cells, and the exocarp, the mesocarp usually 

consists of 1-5(-10) cell layers of sclereids. The sèereid layer may either be continuous or 

restricted to the ribs, when in the furrows the epicarp is in direct contact with the mesocarpic 

fibre celllayer. 

In Anthemis austriaca the sclereids are regularly rectangular in transverse section and 

form a continuous, palisade-like layer under the exocarp. In the rather smooth and inconspicuously 

ribbed achenes of A. eretica subsp. calumnae, A. maritima subsp. maritima, 

and A. punctata subsp. punctata the furrows between the ribs formed by the mesocarpic 

fibre cell layer are filled out by the sclereid layer. Rere, sclereids are often found to be 

radially elongate and somewhat pyriform, separated by large intercellulars and propping 

the somewhat lifted-off epicarpo In most other cases, when achenes are conspicuously

Bocconea 9 - 1998 35 

ribbed, the large sc1ereids are found on top of and/or along the sides of the internaI ribs of 

the fibre cell layer, accentuating the external ribs and their sculpturing with tuberc1es. 

However, achenes of A. arvensis and related species of A. ser. Anthemis display a quite 

differently patterned mesocarp: The inner cylinder of libriform fibre cells is conspicuously 

ribbed, as in all other species of A. sect. Anthemis, but the external ribs of the achene are 

placed on top of the internaI furrows. This is caused by a parenchymatous outer mesocarpic 

layer of cells that divide intensively during maturation. Since the parenchyma is thicker 

above the internaI furrows of the sc1erenchymatic cylinder than above its ribs, the latter 

will eventually coincide with the furrows of the achene. 

From peripheral disc achenes to centraI ones, one may note a trend to reduction of 

sc1erenchymatic mesocarp tissue. This trend is particularly obvious in some representatives 

of Anthemis ser. Chrysanthae and ser. Bourgaeinianae. Unlike other taxa in 

A. sect. Chrysanthae, A. tenuisecta subsp. tenuisecta has achenes that usually lack a continuous 

layer of sc1erenchymatic mesocarp tissue. Libriform fibre cells are confined to 

bundles in the ribs, and the outer mesocarpic cell layer remains parenchymatous. Same is 

true for all achenes of A. monilicostata and the centraI achenes of A. stiparum and 

A. mauritiana of A. sect. Bourgaeinianae. 

The cells of the endocarp remai n parenchymatous and soon become compressed during 

achene maturation. In ripe achenes they cannot usually be identified. The testa consists of 

an epiderrnis and 2-3 layers of large, parenchymatous cells that soon disintegrate during 

maturity. Cells of the testa epiderrnis are characterised by curved and conspicuously thickened 

radiai walls. The testa epiderrnis and the endosperm layer surrounding the cotyledons 

are still visible in mature achenes. 

lO. Chromosomes 

Material and methods 

Far the study of pollen mother cell (PMC) meiosis, buds of capitula were fixed in 96 % 

ethanol/glacial acetic acid (3 : l) in the field and stored in a refrigerator after return to 

Berlin. For chromosome staining several buds of disc florets were removed from the capitula, 

coarsely crushed, and squashed in aceto-orcein. 

For the study of rnitosis, root tips were obtained either by growing plants in the Botanic 

Garden Berlin~Dahlem or by gerrninating achenes in petri dishes. In both cases root tip 

meristems were pre-treated with hydroxyquinoline (0.002 molar aqueous solution) for 4 

hours, fixed in 96 % ethanol/glacial acetic acid (3 : l) and refrigerated. Hydrolysis was 

performed with 1-2n hydrochloric acid for lO rninutes at 60°C. For chromosome staining 

root tips were squashed in aceto-orcein. 

Camera lucida drawings of chromosomes were made with a Zeiss Standard 16 rnicroscope 

and measured with an ocular rnicrometer of a Wild M5A stereornicroscope for arm 

lengths. The PC program CHROMEX described by Voss & al. (1994) was used to transfer 

measurements to a karyological database, to sort chromosomes automatically according to 

their length and arm ratios, to find homologous chromosomes interactively, to construct 

idiograms, and to calculate means of multiple probes. For the description of karyotypes the

Bocconea 9 - 1998 35 

ribbed, the large sc1ereids are found on top of and/or along the sides of the internai ribs of 

the fibre cell layer, accentuating the external ribs and their sculpturing with tuberc1es. 

However, achenes of A. arvensis and related species of A. ser. Anthemis display a quite 

differently patterned mesocarp: The inner cylinder of libriform fibre cells is conspicuously 

ribbed, as in all other species of A. sect. Anthemis, but the external ribs of the achene are 

placed on top of the internai fUITows. This is caused by a parenchymatous outer mesocarpie 

layer of cells that divide intensively during maturation. Since the parenchyma is thicker 

above the internai fUITOWS of the sc1erenchymatic cylinder than above its ribs, the latter 

will eventually coincide with the fUITOWS of the achene. 

From peripheral disc achenes to centrai ones, one may note a trend to reduction of 

sc1erenchymatic mesocarp tissue. This trend is particularly obvious in some representatives 

of Anthemis ser. Chrysanthae and ser. Bourgaeinianae. Unlike other taxa in 

A. sect. Chrysanthae, A. tenuisecta subsp. tenuisecta has achenes that usually lack a continuous 

layer of. sc1erenchymatic mesocarp tissue. Libriforrn fibre cells are confined to 

bundles in the ribs, and the outer mesocarpic cell layer remains parenchymatous. Same is 

true for all achenes of A. monilicostata and the centrai achenes of A. stipa rum and 

A. mauritiana of A. sect. Bourgaeinianae. 

The cells of the endocarp remain parenchymatous and soon become compressed during 

achene maturation. In ripe achenes they cannot usually be identified. The testa consists of 

an epiderrnis and 2-3 layers of large, parenchymatous cells that soon disintegrate during 

maturity. Cells of the testa epiderrnis are characterised by curved and conspicuously thickened 

radiai walls. The testa epiderrnis and the endosperm layer sUITounding the cotyledons 

are still visible in mature achenes. 

lO. Cbromosomes 


For the study of pollen mother cell (PMC) meiosis, buds of capitula were fixed in 96 % 

ethanoVglacial acetic acid (3 : 1) in the field and stored in a refrigerator after return to 

Berlin. For chromosome staining several buds of disc florets were removed from the capitula, 

coarsely crushed, and squashed in aceto-orcein. 

For the study of rnitosis, root tips were obtained either by growing plants in the Botanic 

Garden Berlin-Dahlem or by gerrninating achenes in petri dishes. In both cases root tip 

meristems were pre-treated with hydroxyquinoline (0.002 molar aqueous solution) for 4 

hours, fixed in 96 % ethanoVglacial acetic acid (3 : 1) and refrigerated. Hydrolysis was 

performed with 1-2n hydrochloric acid for lO rninutes at 60°C. For chromosome staining 

root tips were squashed in aceto-orcein. 

Carnera lucida drawings of chromosomes were made with a Zeiss Standard 16 rnicroscope 

and measured with an ocular rnicrometer of a Wild M5A stereornicroscope for arm 

lengths. The PC prograrn CHROMEX described by Voss & al. (1994) was used to transfer 

measurements to a karyological database, to sort chromosomes automatically according to 

their length and arrn ratios, to find homologous chromosomes interactively, to construct 

idiograrns, and to ca1culate means of multiple probes. For the description of karyotypes the


Table 1. Summary of karyotype asymmetry analyses. Intrachromosomal and 

interchromosomal asymmetry were calculated as defined by Romero Zarco 

(1986: A, and A 2 

) and Barghi & al. (1989: Asl and R). 

Taxon 

Origin 

Collection 

2n A, A 2 

Asl R 

2 

Anthemis sect. Hiorthia 

A. maritima 

subsp. maritima 

3 A. pedunculata 

varo pedunculata 

4 

5 var. discoidea 

Anthemis sect. Anthemis 

6 A. arvensis 

subsp. incrassata 

7 

8 

9 A. confusa 

10 

11 A. gharbensis 

12 A. melampodina 

13 A. secundiramea 

varo cossyrensis 

14 A. tenuisecta 

15 A. ubensis 

Ga: Narbonne 

Co: Alistro 

HB Liège 

HB Liège 

36 .370 .166 60.0 1.80 

36 .361 .139 59.4 1.75 

Ma: Tazzeka Vogt & Oberprieler 37 .397 .119 60.4 1.63 

3937 

Ma: Taza Vogt & Oberprieler 18 .388 .132 60.9 1.55 

3945 

Ma: Tidighin Oeil4867 18 .397 .135 61.4 1.59 

Ma: Bab Taza Vogt 9652 & Ober- 18 .395 .107 60.4 1.42 

prieler 4088 

Hs: Sierra de las Vogt 9229 18 .341 .133 58 .6 1.52 

Nieves 

Ma: Tidirhine Vogt 9592 & Oberprieler 

18 .388 .109 60.9 1.35 

Tn: Tataouine 4028 

Vogt 13131 & 18 .433 .133 62.4 1.57 

Tn: Remada 

Oberprieler 7436 

Vogt 13049 & 20 .416 .136 61 .7 1.64 

Ma: Mamora 


Podlech 43489 18 .388 .143 60.7 1.58 

Eg: ElArish Podlech 49968 18+ 1 B .435 .131 62.9 1.58 

Tn: Cap Blanc Vogt 13791 & 18 .391 .125 60.7 1.56 


Ma: Tiznit Podlech 45076 18 .402 .118 61.3 1.34 

Tn: Kesra 

Vogt 13524 & 


19 .351 .116 59.2 1.46 

Anthemis sect. Maruta 

16 A. cotula 

HB Dijon 

18 .327 .119 57.7 1.53 

17 A. pseudocotula 

Cy: Paphos 

Vogt9000 

18+1B .362 .112 60.8 1.42 

Anthemis sect. Gota 

18 A. altissima 

Ga: Alzonne 

HB Liège 

18 

.350 .148 59.8 1 .53 

19 

Ga: Montolieu 

HB Liège 

18 

.355 .157 60.1 1.53 

20 

Ga: Sigean 

HB Liège 

18 

.339 .158 59.3 1.76 

21 

Ga: Villedubert 

HB Liège 

18 

.342 .152 59.5 1.49 

22 

It: Perinaldo 

HB Genua 

18 

.348 .144 59.5 1.55 

23 

Ga: Montolieu 

HB Liège 

18 

.342 .135 59.6 1.69 

Anthemis sect. Anthemaria 

24 A. tinctoria 

Au: Persenberg 

HB Univo Salzburg 

18 .373 .148 60.2 1.61

Bocconea 9 - 1998 37 

Table 2. Data on karyotype asymmetry (as defined by Barghi & al. 1989) used in 

our analysis of Anthemideae, in addition to the originai data of Table 1. 

Taxon Asl R Source 

Chrysantheminae 

25 Argyranthemum frutescens (L.) Schultz-Bip. 60.5 1.65 Uitz (1970: Table 7) 

26 Chrysanthemum segetum L. 56.2 1.69 Uitz (1970: Table 8) 

27 C. coronarium L. 54.7 1.46 Uitz (1970: Table 8) 

28 Ismelia vesicolor Casso 53.7 1.33 Uitz (1970: Table 8, sub 

Chrysanthemum carinatum) 

Achilleinae 

29 Achillea asplenifolia Vent. 60.8 1.55 Uitz (1970: Table 5) 

30 A. setacea Wildst. & Kit. 60.2 1.41 Uitz (1970: Table 5) 

31 Anacyclus clavatus (Dest.) Perso 57.7 1.44 Uitz (1970: Table 5) 

32 A. radiatus Loisel. 59.9 1.37 Uitz (1970: Table 5) 

33 A. pyrethrum (L.) Lagasca 56.5 1.33 Uitz (1970: Table 5) 

34 Chamaemelum nobile (L.) Ali. 57.7 1.21 Uitz (1970: Table 10, sub 

Ormenis nobilis) 

35 C. fuscatum (Brot.) Vasco 58.9 1.60 Uitz (1970: Table 10, sub 

Ormenis fuscata) 

36 C. mixtum (L.) Ali. 59.9 1.63 Uitz (1970: Table 10, sub 

Ormenis mixta) 

37 Cladanthus arabicus (L.) Casso 55.1 1.45 Uitz (1970: Table 8) 

Tanacetinae 

38 Tanacetum vulgare L. 59.3 1.53 Uitz (1970: Table 11) 

39 T. parthenium (L.) Schultz-Bip. 58 .0 1.23 Uitz (1970: Table 11) 

40 T. annuumL. 60.8 1.93 Uitz (1970: Table 11) 

Matricariinae 

41 Tripleurospermum perforatum (Merat) Lainz 62.6 1.56 Uitz (1970: Table 11 , sub 

T. inodorum) 

42 T. maritmum (L.) K. Koch 60.7 1.47 Uitz (1970: Table 12) 

43 T. disciforme (C. Meyer) Schultz-Bip. 60.4 1.63 Uitz (1970: Table 11) 

44 Matricaria recutita (L.) Rauschert 60.7 1.68 Uitz (1970: Table 10) 

45 M. matricarioides Porter ex Britton 56.1 1.93 Uitz (1970: Table 10) 

Leucantheminae 

46 Rhodanthemum arundanum (Boiss.) Wilcox & al. 56.4 1.51 Vogt(1991 : 67, sub 

Leucanthemum arundanum) 

47 Leucanthemum vulgare Lam. 59.3 1.24 Uitz (1970: Table 9) 

48 L. vulgare subsp. pujiulae Sennen 54.1 1.24 Vogt (1991: 67) 

49 Coleostephus myconis (L.) Reichenb. f. 57.3 1.30 Uitz (1970: Table 9, as 

Leucanthemum myconis) 

50 Lepidophorum repandum (L.) DC. 54.1 1.51 Uitz (1970: Table 9) 

51 Nivel/ea nivel/ei (Br.-BI. & Maire) Wilcox & al. 57.7 1.48 Vogt & Oberprieler (1996: 124) 

terminology of Levan & al. (1964) is used. Arm ratios of single chromosomes are expressed 

by the ratio long arm to short armo Karyotype asymmetry was assessed using the 

indices (Ah A 2 ) proposed by Romero Zarco (1986). 

Alternatively, for purposes of comparison of karyotype asymmetries of the present 

study with measurements given by Uitz (1970), the asymmetry indices Asl and R as de-


fined by Barghi & al. (1989) were computed. Here, intrachromosomal asymmetry (Asl) is 

expressed by the ratio of long arms in chromosome set to total chromosome length in the 

set multiplied by 100, and interchromosomal asymmetry (R) by the ratio of longest pair to 

shortest pair in the chromosome set. 

In the folIowing lists, specimens used for karyotype analysis are marked with an asterisk. 

Results of karyotype analyses are summarised in Table l, and morphometric chromosome 

data for each karyotype studied are given in Tables 3-16, where n is the number of 

different metaphase plates studi ed. 

Results 

Anthemis (subg. Anthemis, sect. Anthemis) ser. Anthemis 

Anthemis arvensis subsp. arvensis 

2n = 18 

-Ge, Sachsen-Anhalt, Brachwitz, HB Halle (cult. in HB Berlin No. 153-02-93-10). 

Numerous counts of the cosmopolitan Anthemis arvensis have been published, most of 

them without a mention of subspecies. Counts relating to A. arvensis subsp. arvensis were 

given by Kuzmanov & al. (1980) and Kuzmanov & al. (1981) for Bulgarian, Fernandes & 

Queir6s (1971, sub A. arvensis varo arvensis) for Portuguese, and Benedf i Gonziilez 

(1987) for Spanish plants. AlI counts yielded a diploid chromosome number of n = 9 or 

2n = 18. Kuzmanov & al. (1981) found the karyotype of this subspecies to consist of 2 

metacentric, 12 submetacentric and 4 satellited subtelocentric chromosomes (2m + 12sm + 

4st-sat). 

Anthemis arvensis subsp. incrassata 

n=9 

-Ma, Rif, Djebel Tassaot, 1600m, 25 Jun 1992, Vogt & Oberprieler4078. 

-Ma, Rif, Souk-Tleta-Taghramet - Ceuta, 450-500 m, 17 Apr 1993, Vogt 9854 & Oberprieler 

4302 (Fig. 4A). 

-Ma, Rif, Monts de Beni Hosmar, 1200 m, 21 Apr 1993, Vogt 10001 & Oberprieler 4449. 

-Ma, Rif, Chefchaouene, 730 m, 30 May 1993, Vogt 12037. 

-Ma, Rif, Monts de Beni Hosmar, 1200 m, 30 May 1993, Vogt 12038. 

2n = 18 

* - Ma, Rif, Djebel Tidirhine, 1550 m, 20 Jun 1992, Vogt 9592 & Oberprieler 4028. 

-Ma, Rif, Bab Berret, 1450 m, 25 Jun 1992, Vogt 9632 & Oberprieler 4068. 

*- Ma, Rif, Bab Taza - Djebel Talamssantane, 1420 m, 26 Jun 1992, Vogt 9652 & Oberprieler 4088 

(Fig. 4B). 

*- Hs, Malaga, Sierra de las Nieves, 900-1250 m, 26 May 1992, Vogt 9229. 

-Hs, Cadiz, El Puerto de Sta. Maria - Jerez de la Frontera, 30 m, 17 Apr 1993, Vogt 9711 & 

Oberprieler 4159. 

-Hs, Cadiz, Benolup, 20 m., 16 Apr 1989, Deil3603. 

-Si, Palermo, Scillato - Collesano, 520-800 m, 25 May 1994, Vogt 13885 & Oberprieler 8190. 

-lt, Grosseto, Monte Labbro, l Sep 1992, Marchetti (HB Siena) (cult. in HB Berlin No. 103-02- 

93-10).

Bocconea 9 - 1998 39 

Table 3. Chromosomes of Anthemis arvensis subsp. incrassata 

Chromo- Long arm Short arm Totallength Relative length Arm ratio Chromosome 

some pair (~m) (~m) (~m) (%) type 

Oì'U 6 (n = 4) 

I 3.43 2.62 6.05 13.8 1.30 m 

Il 3.06 2.33 5.39 12.3 1.31 m 

III 2.74 2.27 5.01 11.4 1.20 m 

IV 3.12 1.71 4.83 11.0 1.82 sm 

V 3.12 0.91 4.81 11.0 3.42 st-sat 

VI 2.71 1.98 4.69 10.7 1.36 m 

VII 3.13 1.36 4.49 10.3 2.30 sm 

VIII 2.42 1.94 4.36 10.0 1.24 m 

IX 2.76 0.72 4.26 9.7 3.83 st-sat 

OTU 7 (n = 3) 

I 3.97 3.11 7.08 13.9 1.27 m 

Il 3.44 2.95 6.39 12.6 1.16 m 

III 3.28 2.83 6.11 12.0 1.15 m 

IV 3.23 2.52 5.75 11 .3 1.28 m 

V 3.47 2.16 5.63 11.1 1.60 m 

VI 3.59 0.93 5.30 10.4 3.86 st-sat 

VII 2.97 2.29 5.26 10.4 1.29 m 

VIII 2.67 2.02 4.69 9.2 1.32 m 

IX 3.19 0.68 4.65 9.2 4.69 st-sat 

OTU 8 (n = 3) 

I 4.02 2.97 6.99 13.3 1.35 m 

Il 3.57 2.97 6.54 12.4 1.20 m 

III 4.16 1.19 6.13 11 .6 3.49 sl-sat 

IV 3.32 2.76 6.08 11 .5 1.20 m 

V 3.95 2.13 6.08 11.5 1.85 sm 

VI 3.06 2.51 5.57 10.6 1.21 m 

VII 3.81 0.80 5.39 10.2 4.76 st-sat 

VIII 3.51 1.67 5.18 9.8 2.10 sm 

IX 2.75 2.08 4.83 9.2 1.32 m 

Anthemis arvensis subsp. incrassata was studied cytologically by Brullo & al. (1978b) 

from Sicily and Benedi i Gonzalez (1987) from Spain, who found n = 9 or 2n = 18, respectively. 

Additionally, many counts published for A. arvensis concern this subspecies, 

e.g. those of Dahlgren & al. (1971) for Balearic or Pavone & al. (1981) for Sicilian plants 

because A. arvensis subsp. arvensis is rnissing there. Counts on N African plant material 

of "A. arvensis" were made by Humphries & al. (1978) who found 2n = 18 chromosomes, 

but the identity of the plants, originating from "Jbel Ayachi, cirque de Jaffar" in the High 

Atlas mountains where A. arvensis was never found to grow, is questionable. 

A deviating and presumably wrong count of n = 8 was published by Blanché & al. 

(1985) for plants from the Sierra Nevada in S Spain identified as A. granatensis which 

according to Benedi i Gonzalez (1987) is a synonym of A. arvensis subsp. arvensis and in 

my opinion, of A. arvensis subsp. incrassata. A triploid chromosome number of 2n = 27 

was found in B ulgarian plants by Kuzmanov & al. (1981), who found the karyotype of this 

population to consist of 7 metacentric, 14 submetacentric and 6 satellited subtelocentric


chromosomes. A further karyotype analysis was performed by Benedf i Gonzalez (1987) 

who indicates the chromosome formula 12m + 2M + 2st-sat + 2t-sat for plants from near 

Barcelona. Karyotype analyses performed for the present study yielded a chromosome 

formula of 10m + 4sm + 4st-sat for two Moroccan populations and a more symrnetric 

karyotype with 14m + 4st-sat for a S Spanish population (Table 3). 

Anthemis arvensis subsp. sphacelata 

n=9 

- It, Reggio Calabria, Montalto, 1600 m, 27 May 1994, Vogt 13983 & Oberprieler 8288. 

This interesting biennial to perennial subspecies of Anthemis arvensis endemie to the 

mountains of Sicily and S Italy, had been studied cytologically only once, by Brullo & al. 

(1978a) who found a diploid chromosome set of 2n = 18 chromosomes (12m + 4sm + 2tsat) 

in a Sicilian population. The present count from the Calabrian mountains corroborates 

this data. 

Anthemis (subg. Anthemis, sect. Anthemis) ser. Chrysanthae 

Anthemis gharbensis 

n=9 

-Ma, Arbaoua - Moulay Bousselham, lO m, 24 Apr 1993, Vogt 10161 & Oberprieler 4609. 

-Ma, Forèt de la Mamora, 120 m, 27 Apr 1993, Vogt 10259 & Oberprieler 4707. 

2n = 18 

* - Ma, Forèt de la Mamora, 120 m, 2 May 1987, Podlech 43489. 

The haploid and diploid chromosome numbers mentioned above were already reported 

earlier (Oberprieler 1994). The karyotype was found to be very symrnetrical, consisting of 

5 pairs of metacentic, 2 pairs of submetacentric, and 2 pairs of satel\ited subtelocentric 

chromosomes (10m + 4sm + 4st-sat, Table 4). 

Table 4. Chromosomes of Anthemis gharbensis (OTU 11; n = 8). 

Chromo- Long arm Short arm Totallength Relative length Arm ralio Chromosome 

some pair (IJm) (IJm) (IJm) (%) type 

4.54 3.60 8.14 14.4 1.26 m 

Il 3.95 3.18 7.13 12.6 1.24 m 

III 4.63 2.28 6.91 12.2 2.03 sm 

IV 3.44 2.93 6.37 11.2 1.17 m 

V 4.33 1.00 6.11 10.8 4.33 sI-saI 

VI 3.32 2.65 5.97 10.5 1.25 m 

VII 3.69 1.94 5.63 9.9 1.90 sm 

VIII 3.02 2.31 5.33 9.4 1.30 m 

IX 3.55 0.84 5.17 9.1 4.22 sI-saI

Bocconea 9 - 1998 41 

Fig. 4. Metaphase plates in Anthemis. - A: A. arvensis subsp. incrassata, PMC meiosis, n = 9 (Vogt 

9854 & Oberprieler 4302). B: A. arvensis subsp. incrassata, root tip mitosis, 2n = 18 (Vogt 9652 & 

Oberprieler 4088). C: A. maroccana subsp. maroccana, PMC meiosis, n = 9 + IB, arrowhead 

indicating B chromosome (Vogt 10181 & Oberprieler 4729). D: A. tenuisecta subsp. tenuisecta, 

root tip mitosis, 2n = 18 (?odlech 45076). E: A. confusa, root tip mitosis, 2n = 20 (Vogt 13049 & 

Oberprieler 7354). F: A. secundiramea var. secundiramea, root tip mitosis, 2n = 18 (HE Catania). 

- Scale bars = IO J.lm.


Anthemis maroccana subsp. maroccana 

n=9+0-1B 

-Ma, A1n-Cheggag, 620 m, 27 Apr 1993, Vogt 10181 & Oberprieler 4729. (Fig. 4C). 

-Ma, Middle At1as, Sefrou, 1060 m, 27 Apr 1993, Vogt 10284 & Oberprieler4732. 

-Ma, Middle Atlas, Sefrou, 1150m, 28 Apr 1993, Vogt 10296 & Oberprieler4744. 

Anthemis maroccana had never been investigated cytologically before. The three 

populations studied here ali belong to subsp. maroccana and proved to be diploid with n = 

9 chromosomes. 

Anthemis tenuisecta subsp. tenuisecta 

n=9 

-Ma, Agadir, Tiznit - Agadir, 180 m, 19 May 1993, Vogt 11905 & Oberprieler 6353. 

- Ma, Agadir, Agadir - Essaouira, 160m, 21 May 1993, Vo g t 11 924. 

2n = 18 

* - Ma, Agadir, Tiznit - Agadir, 25 Apr 1989, Podlech 45076. (Fig. 4D). 

-Ma, Agadir, Agadir- Essaouira, 120 m, 22 May 1993, Vogt 11930. 

The chromosome number of Anthemis tenuisecta, a species cIosely related to A. gharbensis, 

is reported here for the first time. A ~ in A. gharbensis and A. maroccana, a diploid 

chromosome set was found, with n = 9 and 2n = 18, respectively. The karyotype is a little 

more asymmetrical than in A. gharbensis due to the occurrence of an additional pair of 

submetacentric instead of metacentric chromosomes (8m + 6sm + 4st-sat, Table 5). 

Table 5. Chromosomes of Anthemis tenuisecta (OTU 14; n = 11). 

Chromo- Long arm Short arm Totallength Relative length 

some pair (lJm) (lJm) (lJm) (%) 

4.16 3.58 7.74 13.7 

Il 4.55 2.60 7.15 12.7 

III 3.64 3.12 6.76 12.0 

IV 3.99 2.27 6.26 11.1 

V 3.40 2.68 6.08 10.8 

VI 4.05 1.02 5.85 10.4 

VII 3.87 1.92 5.79 10.3 

VIII 3.08 2.42 5.50 9.8 

IX 3.93 0.71 5.42 9.6 

Arm ratio 

1.16 

1.75 

1.16 

1.75 

1.26 

3.97 

2.01 

1.27 

5.53 

Chromosome 

type 

m 

sm 

m 

sm 

m 

st-sat 

sm 

m 

st-sat 

Anthemis (subg. Anthemis, sect. Anthemis) ser. Melampodinae 

Anthemis melampodina 

2n = 18 + lB 

*-Eg, Sinai, Nizzana - El Arish, 130 m, 2 May 1991, Podlech 49968.

Bocconea 9 - 1998 43 

Fig. 5. Metaphase plates in Anthemis. - A: A. secundiramea var. cossyrensis. root tip mitosis, 2n = 

18 (Vogt 13791 & Oberprieler 8096). B: A. ubensis. PMC meiosis, n = 9, arrowhead indicating a 

bivaIent forrned by chromosomes differing markedly in length (Vogt 12216 & Oberprieler 6521). C: 

A. ubensis. root tip mitosis, 2n = 19 (Vogt 13524 & Oberprieler 7829). D: A. mauritiana subsp. 

faurei. PMC mitosis, n = 9 (Vogt 10927 & Oberprieler 5375). E: A. zaianica. PMC meiosis, n = 9 

(Vogt 14840 & Oberprieler 9149). F: A. abylaea. PMC meiosis, n = 18 (Vogt 10023 & Oberprieler 

4471). - Scale bars = lO ~m.


This seems to be the first count not only for this species but for Anthemis ser. Melampodina 

e as a whole. It coincides with nearly all records for A. sect. Anthemis. All cells 

checked proved to have a small accessory B-chromosome. The karyotype resembles that 

of other species of the section, with 5 pairs of metacentric, 2 pairs of submetacentric, and 2 

pairs of satellited subtelocentric chromosomes (Table 6). The intrachromosomal asymmetry 

index AI yielded the highest value of all species studied here. A comparably high 

value was found only in one population of A. confusa from S Tunisia. 

Table 6. Chromosomes of Anthemis melampodina (OTU 12; n = 5). 


some pair (Ilm) (Ilm) (Ilm) (%) 

type 

4.20 3.57 7.77 13.9 

Il 3.89 2.97 6.86 12.3 

III 4.49 2.19 6.68 12.0 

IV 4.86 0.96 6.60 11 .8 

V 3.48 2.50 5.98 10.7 

VI 3.32 2.41 5.73 10.3 

VII 4.11 0.81 5.70 10.2 

VIII 3.79 1.79 5.58 10.0 

IX 2.98 1.94 4.92 8.8 

1.17 

1.30 

2.05 

5.06 

1.39 

1.37 

5.07 

2.11 

1.53 

m 

m 

sm 

st-sat 

m 

m 

st-sat 

sm 

m 

Anthemis (subg. Anthemis, sect. Anthemis) ser. Secundirameae 

Anthemis confusa 

n=9 

- Tn, Gafsa, Redeyef - Tamerza, 680 m, 9 May 1994, Vogt 12747 & Oberprieler 7052. 

- Tn, Sidi-Bouzid, Jelma, 370 m, 17 May 1994, Vogt 13333 & Oberprieler 7638. 

2n = 18 

- Tn, Gafsa, Tozeur - El Oued, 45 m, 9 Apr 1980, Podlech 34145. 

*-Tn, Tataouine - Ghomrassen, 340 m, 14 May 1994, Vogt 13131 & Oberprieler 7436. 

2n = 20 

*- Tn, Tataouine - Remada, 450 m, 13 May 1994, Vogt 13049 & Oberprieler 7354 (Fig. 4E). 

This seems to be the first report of chromosome numbers for this species. The indications 

of n = 9 and 2n = 18 agree with all previous reports for members of Anthemis ser. 

Secundirameae. Plants from a population in S Tunisia were found to have an aneuploid 

chromosome number of 2n = 20, but other plants from that area had the normal chromosome 

set with 2n = 18 chromosomes (Table 7). As discussed in the taxonornic part of the 

present work, Le Houérou (1962) considered plants from this area to be morphologically 

intermediate between his A. pedunculata var. decumbens [= A. confusa] and the W Libyan 

A. glareosa. However, in the present revision they are assigned to A. confusa not to 

A. glareosa. The aneuploid chromosome number found in one of these plants may be due 

to the past or recent interbreeding of the two cIosely related species in SE Tunisia.

Bocconea 9 - 1998 45 

Table 7. Chromosomes of Anthemis confusa. 


some pair (~m) (~m) (~m) (%) type 

OTU 9 (n = 2) 

I 3.37 2.54 5.91 14.1 1.32 m 

Il 3.64 0.66 5.08 12.1 5.51 st-sat 

III 2.81 2.23 5.04 12.0 1.26 m 

IV 2.58 2.13 4.71 11 .2 1.21 m 

V 3.16 1.55 4.71 11.2 2.03 sm 

VI 3.18 0.58 4.54 10.8 5.48 st-sat 

VII 2.46 1.94 4.40 10.5 1.26 m 

VIII 2.79 1,.07 3.86 9.2 2.60 sm 

IX 2.23 1.53 3.76 9.0 1.45 m 

OTU 10 (n = 5) 

I (3x) 3.60 2.91 6.51 14.3 1.23 m 

Il 3.08 2.42 5.50 12.1 1.27 m 

III 3.68 1.76 5.44 11.9 2.09 sm 

IV 3.76 0.80 5.34 11.7 4.69 st-sat 

V 2.88 2.22 5.10 11.2 1.29 m 

VI 3.25 0.67 4.70 10.3 4.85 st-sat 

VII 3.09 1.54 4.63 10.2 2.00 sm 

VIII (3x) 2.60 1.92 4.52 9.9 1.35 m 

IX 2.29 1.69 3.98 8.7 1.35 m 

Anthemis secundiramea varo secundiramea 

2n = 18 

-Si, Noto, Vendicari, 16 Jun 1990, HB Catania (eult. in HB Berlin No. 301-04-92-10 & 298-06- 

91-10) (Fig. 4F). 

The chromosome number presented here for this variety of Anthemis secundiramea 

agrees with indications by Capi neri & al. (1976, sub A. secundiramea) for Sicilian materiai 

and Benedf i Gonzalez (1987) for plants from Menorca. One pair of satellite chromosomes 

was found in the studied plants of A. secundiramea var. secundiramea, whereas 

Brullo (in Bartolo & al. 1979) found none, and A. secundiramea varo cossyrensis (see 

below) has two. No Algerian or Tunisian plant material was available for cytological investigations. 

Anthemis secundiramea varo cossyrensis 

n=9 

- Tn, Bizerte, Cap Blanc, 10-30 m, 22 May 1994, Vogt 13791 & Oberprieler 8096. 

2n = 18 

* - Tn, Bizerte, Cap Blane, 10-30 m, 22 May 1994, Vogt 13791 & Oberprieler 8096 (Fig. 5A) 

-Si, Isole Egadi, Fariguana, 15 Sep 1990, HB Catania (eult. in HB Berlin No. 298-05-91-10 & 

301-02-92-10).


The finding of n = 9 and 2n = 18 chromosomes for this variety of Anthemis secundiramea 

agrees with a report by Brullo & al. (1977, sub A. secundiramea subsp. intermedia) 

. Brullo (in Bartolo & al. 1979) later found two pairs of satellited chromosomes in 

plants from Lampedusa which he considered to represent a separate subspecies, 

A. secundiramea subsp. lopadusana, on account of its karyogram (12m + 2sm + 4st-sat), 

as he had found that A. secundiramea subsp. secundiramea lacked satellites. The Tunisian 

plant material studied here shows two pairs of satellited subtelocentric chromosomes 

(Table 8), exaetly as the morphologieally similar plants from Lampedusa. This supports 

the inc1usion of A. secundiramea subsp. lopadusana in A. secundiramea varo cossyrensis 

deseribed from Pantellaria. 

Table 8. Chromosomes 01 Anthemis secundiramea var. cossyrensis (OTU 13; 

n = 6). 

Chromo- Long arm Short arm Totallength 

some pair (~m) (~m) (~m) 

Relative length 

(%) 

Arm ratio 

Chromosome 

type 

3.38 2.76 6.14 

Il 2.98 2.36 5.34 

III 3.58 0.76 5.12 

IV 2.67 2.34 5.01 

V 3.29 1.67 4.96 

VI 3.24 0.59 4.61 

VII 2.51 2.04 4.55 

VIII 2.73 1.52 4.25 

IX 2.29 1.64 3.93 

14.0 1.22 

12.2 1.26 

11.7 4.71 

11.4 1.14 

11.3 1.97 

10.5 5.49 

10.4 1.23 

9.7 1.79 

9.0 1.39 

m 

m 

st-sat 

m 

sm 

st-sat 

m 

sm 

m 

Anthemis secundiramea varo urvilleana 

n=9 

-Me, Paradise Bay, 27 Feb 1994, Mevert & Seifert s.n. (Herb. Oberprieler). 

This taxon, endemie to Malta, had been studied eytologically by Brullo & al. (1997). 

The ehromosome number reported here is in agreement with their findings. 

Anthemis ubensis 

n=9 

- Tn, Beja, Djebel Goraa, 730 m, 3 May 1994, Vogt 12216 & Oberprieler 6521 (Fig. 5B). 

- Tn, Beja, Djebel Goraa, 960 m, 3 May 1994, Vogt 12291 & Oberprieler6596. 

- Tn, Beja, Djebel Goraa, 960 m, 3 May 1994, Vogt 12305 & Oberprieler6610. 

- Tn, El Kef, Djebel Dyr, 1000 m, 4 May 1994, Vogt 12352 & Oberprieler 6657. 

- Tn, El Kef, Table de Jughurta, 1200-1270 m, 5 May 1994, Vogt 12467 & Oberprieler 6772. 

2n= 19 

*-Tn, FOfl~t de Kesra, 18 May 1994, Vogt 13524 & Oberprieler 7829 (Fig. 5C). 

- Tn, Zaghouan, Djebel Zaghouan, 23 May 1994, Vogt 13824 & Oberprieler 8129.

Bocconea 9 - 1998 47 

This interesting species from E Algeria and N Tunisia had not yet been studied cytologically. 

In morphological respects it holds a somewhat intermediate position between 

Anthemis pedunculata and A. secundiramea. The occurrence of accessory chromosomes in 

two of the seven plants studied (Table 9) may be a consequence of hybrid speciation. 

Meiosis, studied in five other plants, was found to be regular, with the formation of nine 

bivalents. In one such population I was able to observe the formation of a bivalent by two 

chromosomes differing markedly in length (Fig. 5B). This again suggests that the chromosome 

set of this species is relatively unbalanced. 

Table 9. Chromosomes of Anthemis ubensis (OTU 15; n = 3). 

Chromo- Long arm Short arm Totallength Relative length Arm ratio 

some pair (IJm) (IJm) (IJm) (%) 

3.60 2.80 6.40 13.5 1.28 

Il 3.18 2.57 5.75 12.2 1.23 

III 3.75 1.12 5.65 12.0 3.34 

IV 3.58 1.95 5.53 11.7 1.83 

V 2.91 2.43 5.34 11.3 1.19 

VI 2.82 2.17 4.99 10.6 1.29 

VII 3.10 0.98 4.86 10.3 3.16 

VIII 2.47 2.00 4.47 9.5 1.23 

IX (3x) 2.65 1.73 4.38 9.3 1.53 

Chromosome 

type 

m 

m 

st-sat 

sm 

m 

m 

st-sat 

m 

m 

Anthemis (subg. Anthemis, sect. Anthemis) ser. Bourgaeinianae 

Anthemis mauritiana subsp.faurei 

n=9 

-Ma, Saidia, lO m, 5 May 1993, Vogt 10872 & Oberprieler 5320. 

-Ma, Ras-el-Ma, 40 m, 5 May 1993, Vogt 10927 & Oberprieler 5375 (Fig. 5D) 

This seems to be the frrst report of a chromosome number for this species endemie to 

the Mediterranean coast of E Morocco. Bivalent formation in pollen mother celi meiosis 

was found to be regular. 

Anthemis stiparum subsp. sabulicola 

2n = 18 

- Ag, Laghouat, Ghardaia - El Golea, 400 m, 28 Mar 1981 , Podlech 35529. 

-Ag, Laghouat, Zelfana, 360 m, 30 Mar 1981 , Podlech 35571 . 

This subspecies of Anthemis stipa rum was studied by Podlech (1986, sub 

A. monilicostata subsp. stipa rum) who found the same chromosome number of 2n = 18 

using one of the specimens (Podlech 35571) of the present study. 

Anthemis zaianica 

n=9 

-Ma, Oulmès, 1310 m, 12 May 1995, Vogt 14840 & Oberprieler 9149. (Fig. 5E).


Fig. 6: Metaphase plates in Anthemis. - A: Anthemis abylaea, root tip mitosis, 2n = 36 (Vogt 

12045). B: A. maritima subsp. maritima, root tip mitosis, 2n = 36 (HB Liège). C: A. pedunculata 

var. pedunculata, PMC meiosis, n = 9 (Vogt 9414 & Oberprieler 3852). D: A. pedunculata var. 

pedunculata, root tip mitosis, 2n = 18 (Vogt 9414 & Oberprieler 3852). E: A. pedunculata varo 

pedunculata, PMC meiosis, n = 18 (Vogt & Oberprieler4026). F: A. pedunculata var. pedunculata, 

PMC meiosis, n = 18 (Vogt 14962 & Oberprieler 9271). - Scale bars = lO 11m.

Bocconea 9 - 1998 49 

No previous report of a chromosome number for this species newly described exists. 

The number of n = 9 chromosomes reported here agrees with that found in other species of 

Anthemis ser. Bourgeinianae and most members of A. sect. Anthemis. 

Anthemis (subg. Anthemis) sect. Hiorthia 

Anthemis abylaea 

n = 18 

-Ma, Djebel Fahies, 550-660 m, 20 Apr 1993, Vogt 9897 & Oberprieler4345. 

-Ma, Monts de Beni Hosmar, 1200 m, 21 Apr 1993, Vogt 10023 & Oberprieler4471 (Fig. 5F). 

-Ma, Monts de Beni Hosmar, 1200 m, 30 May 1993, Vogt 12039. 

2n = 36 

-Ma, Djebel Fahies, 550-660 m, 31 May 1993, Vogt 12045 (Fig. 6A). 

The chromosome number for this species endemie to the limestone mountains of the 

Tangier peninsula is reported here for the first time. 

Anthemis eretica subsp. columnae 

2n= 36 

- It, Avellino, Monte Vergine, 1220 m, 28 May 1994, Vogt 14004 & Oberprieler 8309. 

It was not possible to study N African plant material of this taxon cytologically, so a 

chromosome count from the locus cIassicus of this subspecies in S Italy is provided instead. 

The tetraploid level found confirms counts by Strid & Franzén (1983, sub Anthemis 

carpatica) and Papanicolaou (1984, sub A. eretica) on Greek plants that were later assigned 

to A. eretica subsp. columnae by Franzén (1986). 

Anthemis maritima subsp. maritima 

n = 18 

- Tn, Bizerte, Cap Blane, 10-30 m, 22 Jun 1994, Vogt 13792 & Oberprieler 8097. 

-Ga, Col de Tende, 270 m, 23 May 1995, Vogt 15352 & Oberprieler 9661 . 

2n= 36 

- Tn, Bizerte, Cap Blane, 20 m, 22 May 1994, Vogt 13813 & Oberprieler 8118. 

-Hs, Cab ° Roehe, 5-JOm, 17 Apr 1993, Vogt 9733 & Oberprieler4181. 

*-Ga, Aude, Narbonne, HB Liège (eult. in HB Berlin No. 194-25-92-10). 

*-Co, Alistro, HB Liège (eult. in HB Berlin No. 194-27-92-10) (Fig. 6B). 

The tetraploid chromosome numbers found in Tunisian, Spanish and French material 

(Table lO) confirms counts of n = 18 or 2n = 36, respectively, made on plants of the same 

taxon from the Mediterranean coasts of France (Larsen 1954, Rashid 1974), Italy 

(Capi neri & al. 1976), Sicily (Brullo & Pavone 1978, Brullo & al. 1988), and Mallorca 

(Benedf i Gonzalez 1987), and from the Atlantic coasts of Portugal (Fernandes & Queir6s 

1970, 1971) and Spain (Luque 1984, Benedf i Gonzalez 1987).


Deviating, diploid chromosome numbers for Anthemis maritima were given by Harling 

(1950) and Mitsuoka & Ehrendorfer (1972) who studied plant material from botanical 

gardens, without documented provenance. Aneuploid chromosome numbers of n = 10 and 

2n = 20 for a population of A. maritima from Menorca were reported by Benedf i GonzaIez 

(1987), which suggests that diploid chromosome numbers can indeed be found in 

natural populations of A. maritima. 

Table 10: Chromosomes of Anthemis maritima subsp. maritima. 

Chromo- Long arm $hort arm Totallength Relative length Arm ratio Chromosome 

some pair (11m) (11m) (11m) (%) type 

OTU 1 (n = 1) 

I 4.07 3.91 7.98 14.6 1.04 m 

4.38 3.41 7.79 14.2 1.28 m 

Il 4.77 2.71 7.48 13.6 1.76 sm 

Il 4.19 2.95 7.14 13.0 1.42 m 

III 3.88 2.83 6.71 12.2 1.37 m 

III 3.76 2.67 6.43 11.7 1.40 m 

IV 3.60 2.64 6.24 11.4 1.36 m 

IV 3.60 2.48 6.08 11 .1 1.45 m 

V 3.10 2.91 6.01 11 .0 1.06 m 

V 3.37 2.52 5.89 10.7 1.33 m 

VI 3.99 1.28 6.05 11.0 3.11 st-sat 

VI 3.99 0.85 5.62 10.3 4.69 st-sat 

VII 3.84 1.90 5.74 10.5 2.02 sm 

VII 3.06 2.48 5.54 10.1 1.23 m 

VIII 3.72 0.78 5.28 9.6 4.76 st-sat 

VIII 3.26 0.74 4.78 8.7 4.40 st-sat 

IX 2.33 2.09 4.42 8.1 1.11 m 

IX 2.83 1.63 4.46 8.1 1.73 sm 

OTU 2 (n=4) 

I 3.53 2.64 6.17 14.7 1.33 m 

I 3.56 1.85 5.41 12.9 1.92 sm 

Il 2.99 2.42 5.41 12.9 1.23 m 

Il 2.95 2.18 5.13 12.2 1.35 m 

III 3.31 1.15 5.24 12.5 2.87 sm-sat 

III 3.34 0.93 5.05 12.0 3.59 st-sat 

IV 2.68 2.20 4.88 11 .6 1.21 m 

IV 2.65 2.13 4.78 11.4 1.24 m 

V 2.55 2.01 4.56 10.9 1.26 m 

V 2.96 1.76 4.72 11.2 1.68 m 

VI 2.96 0.89 4.63 11 .0 3.32 st-sat 

VI 2.49 0.86 4.13 9.8 2.89 sm-sat 

VII 2.73 1.64 4.37 10.4 1.66 m 

VII 2.36 1.91 4.27 10.2 1.23 m 

VIII 2.1 9 1.84 4.03 9.6 1.19 m 

VIII 2.51 1.37 3.88 9.2 1.83 sm 

IX 2.08 1.76 3.84 9.1 1.18 m 

IX 2.03 1.49 3.52 8.4 1.36 m

Bocconea 9 - 1998 51 

Anthemis pedunculata (subsp. pedunculata) varo pedunculata 

n=9 

-Ma, Middle Atlas, Azrou - AIn Leuh, 1585 ID, Il Jun 1992, Vogt 9414 & Oberprieler 3852 bis 

(Fig. 6C). 

-Ma, Middle Atlas, Azrou - Ain Leuh, 1580 ID, Il Jun 1992, Oberprieler 3849. 

-Ma, Middle Atlas, FOUID Kheneg, 1880 ID, 12 Jun 1992, Vogt 9440 & Oberprieler 3878. 

-Ma, Djebel Tazzeka, Gouffe de Friovato, 1200 ID, 15 Jun 1992, Vogt 9468 & Oberprieler 3905. 

- Ma, Rif, Djebel Tizirene, 1700 ID, 24 Jun 1992, Vogt 9628 & Oberprieler 4064. 

-Ma, Rif, Djebel Lakraa, 1900-2000 ID, 26 Jun 1992, Vogt & Oberprieler4115. 

-Ma, Monts des Beni Snassen, Tanezzert - Taforalt, 900-950 ID, 9 May 1993, Vogt 11448 & 


-Ma, Middle Atlas, Source de l'OuID-er-Rbia, 1630 ID, 28 May 1993, Vogt 12002. 

-Ma, Middle Atlas, AIn-Leuh - Azrou, 1900 ID, 13 May 1995, Vogt 14949 & Oberprieler 9258 

bis. 

-Ma, Middle Atlas, AIn-Leuh - Azrou, 1980 ID, 13 May 1995, Vogt 14927 & Oberprieler 9236. 

-Ma, Monts des Beni-Snassen, Djebel Foughal, 1280 ID, 16 May 1995, Vogt 15264 & Oberprieler 

9573. 

2n = 18 

-Ma, Taza, AhermouIDou (Ribat-el-Kheyr) - Jebel Bou Iblane, 1430-1550 ID, 25 Jun 1989, 

Podlech 46964. 

-Ma, ibid., 1680 ID, 25 Jun 1989, Oberprieler 1908. 

-Ma, Middle Atlas, Azrou - AIn Leuh, 1585 ID, Il Jun 1992, Vogt 9414 & Oberprieler 3852 bis 

(Fig. 6D). 

-Ma, Middle Atlas, FOUID Kheneg, 1880 ID, 12 Jun 1992, Vogt 9440 & Oberprieler 3878. 

-Ma, Middle Atlas, Col du Zad, 2100 ID, 12 Jun 1992, Vogt 9442 & Oberprieler 3880. 

-Ma, Djebel Tazzeka, Gouffe de Friovato, 1200 ID, 15 Jun 1992, Vogt 9468 & Oberprieler 3905. 

*- Ma, Djebel Tazzeka, Taza - Bab-Bou-Idir, 1420 ID, 17 Jun 1992, Vogt & Oberprieler 3945. 

-Ma, Rif, Djebel Tizirene, 1700 ID, 24 Jun 1992, Vogt 9628 & Oberprieler4064. 

n = 18 

-Ma, Djebel Tazzeka, 1900 ID, 16 Jun 1992, Vogt 9487 & Oberprieler 3923. 

-Ma, Rif, KetaIDa - Djebel Tidirhin, 1750 ID, 20 Jun 1992, Vogt & Oberprieler4026 (Fig. 6E). 

-Ma, Rif, TaleIDbote - Djebel Tassaot, 1565 ID, 25 Jun 1992, Vogt 9647 & Oberprieler 4083. 

-Ma, Rif, Djebel TalaIDssantane, 1765-1900 ID, 26 Jun 1992, Vogt 9668 & Oberprieler4105. 

. I 

-Ma, Monts des Beru Snassen, Tanezzert - Taforalt, 900 ID, 9 May 1993, Vogt 11466 & Oberprieler 

5914. 

- Ma, Anti-Atlas, Djebel Lekst, 1500 ID, 15 May 1993, Vogt 11766 & Oberprieler 6214. 

-Ma, Anti-Atlas, Djebel Lekst, 1550-1570 ID, 16 May 1993, Vogt 11829 & Oberprieler 6277. 

-Ma, Middle Atlas, Col du Zad, 2125 ID, 26 May 1993, Vogt 11988. 

-Ma, Tahout-ou-Fillali, 1990 ID, 26 May 1993, Vogt 11948. 

-Ma, Middle Atlas, AIn-Leuh - Source de l'OuID-er-Rbia, 1790 ID, 28 May 1993, Vogt 12006. 

-Ma, Middle Atlas, Ain-Leuh - Azrou, 1900 ID, 13 May 1995, Vogt 14949 & Oberprieler 9258. 

-Ma, Middle Atlas, Ain-Leuh - Azrou, 1900 ID, 13 May 1995, Vogt 14949 & Oberprieler 

9258 ter. 

-Ma, Middle Atlas, AIn-Leuh - Azrou, 1830 ID, 13 May 1995, Vogt 14962 & Oberprieler 9271 

(Fig. 6F). 

-Ma, Middle Atlas, Ain-Leuh - Azrou, 1720 ID, 14 May 1995, Vogt 15005 & Oberprieler 9314. 

-Ma, Middle Atlas, AIn-Leuh - Azrou, 1700 ID, 14 May 1995, Vogt 15010 & Oberprieler 9319.


Table 11. Chromosomes of Anthemis pedunculata varo pedunculata. 



OTU 3 (n = 5) 

I 2.90 2.33 5.23 13.4 1.24 m 

2.53 2.29 4.82 12.3 1.10 m 

Il 2.67 1.96 4.63 11.8 1.36 m 

Il 2.46 2.03 4.49 11.5 1.21 m 

III 3.03 0.77 4.58 11.7 3.93 st-sat 

III 2.74 0.66 4.18 10.7 4.15 st-sat 

IV 2.78 1.52 4.30 11 .0 1.82 sm 

IV 2.65 1.40 4.05 10.4 1.89 sm 

V 2.44 1.83 4.27 10.9 1.33 m 

V 2.29 1.79 4.08 10.4 1.27 m 

VI 2.49 0.62 3.89 10.0 4.01 st-sat 

VI 2.66 0.54 3.98 10.2 4.92 st-sat 

VII 2.13 1.81 3.94 10.1 1.17 m 

VII 2.13 1.59 3.72 9.5 1.33 m 

VIII 2.00 1.54 3.54 9.1 1.29 m 

VIII 2.07 1.40 3.47 8.9 1.47 m 

IX 2.54 1.12 3.66 9.4 2.26 sm 

IX 2.22 0.97 3.19 8.2 2.28 sm 

+ (V) 2.50 1.66 4.16 10.6 1.50 m 

OTU4(n=14) 

I 3.90 3.28 7.18 14.2 1.18 m 

Il 4.13 2.14 6.27 12.4 1.92 sm 

III 3.39 2.72 6.11 12.1 1.24 m 

IV 2.99 2.53 5.52 11.0 1.18 m 

V 3.96 0.78 5.52 11 .0 5.07 st-sat 

VI 3.46 1.75 5.21 10.3 1.97 sm 

VII 2.88 2.26 5.14 10.2 1.27 m 

VIII 3.42 0.72 4.92 9.8 4.75 st-sat 

IX 2.60 2.03 4.63 9.2 1.28 m 

-Ma, Middle Atlas, Azrou - Ifrane, 1750 m, 14 May 1995, Vogt 15016 & Oberprieler 9325. 

-Ma, Monts des Beni-Snassen, Gorge du Zegzel - Nn-Almou, 800-900 m, 15 May 1995, Vogt 

15123 & Oberprieler 9432. 

-Ma, Monts des Beni-Snassen, Nn Almou, 1180 m, 16 May 1995, Vogt 15188 & Oberprieler 

9497. 

-Ma, Monts des Beni-Snassen, Djebel Foughal, 1200 m, 16 May 1995, Vogt 15227 & Oberprieler 

9536. 

-Ma, Monts des Beni-Snassen, Oulad Jabein-Fouaga, 11 80 m, 16 May 1995, Vogt 15280 & 


2n = 36 

-Ma, Taza, Ahermoumou (Ribat-el-Kheyr) - Jbel Bou-Iblane 1680 m, 25 Jun 1989, Oberprieler 

1895. 

-Ma, Jebel Bou Iblane, Tizi Bouzabel, 2040 m, 26 Jun 1989, Oberprieler 1967. 

- Ma, Tizi-n-Tahout-ou-Fillali, 2070 m, 4 lui 1989, Podlech 47622.

Bocconea 9 - 1998 53 

-Ma, Middle Atlas, Foum Kheneg, 1880 m, 12 Jun 1992, Vogt 9440 & Oberprieler 3878bis. 

-Ma, Djebel Tazzeka, 1780 m, 16 Jun 1992, Vogt & Oberprieler 3937. 

-Ma, Rif, Talembote - Djebel Tassaot, 1600 m, 25 Jun 1992, Vogt 9638 & Oberprieler 4074 (Fig. 

7A). 

-Ma, Rif, Talembote" Djebel Tassaot, 1565 m, 25 Jun 1992, Vogt 9647 & Oberprieler 4083. 

-Ma, Rif, Talembote - Djebel Tassaot, 1565 m, 25 Jun 1992, Vogt & Oberprieler 4084. 

-Ma, Rif, Djebel Talarnssantane, 1765-1900 m, 26 Jun 1992, Vogt 9668 & Oberprieler 4105. 

2n = 37 

* - Ma, Djebel Tazzeka, 1780 m, 16 Jun 1992, Vogt & Oberprieler 3937 (Fig. 7B). 

Vogt & Oberprieler (1993, sub Anthemis pedunculata subsp. tuberculata) were the first 

to find tetraploid chromosome numbers in A. pedunculata. Before, only diploid numbers 

had been reported, both for N African (Galland 1985, sub A. pedunculata; Galland 1991, 

sub A. pedunculata subsp. pedunculata) and Spanish populations (Aparicio & Silvestre 

1985; Blanché & al. 1985, sub A. tuberculata; Benedf i Gonzaléz 1987, sub A. tuberculata 

subsp. tuberculata). The counts listed above demonstrate that in Morocco both cytotypes 

grow not only in the same geographical region but in close proximity, even in the 

same locality (Table Il). 

Anthemis pedunculata (subsp. pedunculata) varo discoidea 

n=9 

-Ma, Rif, Djebel Tidirhin, 2200 m, 20 Jun 1992, Vogt & Oberprieler4012 (Fig. 7C, D). 

2n = 18 

-Ma, Rif, Bab Bagla, 2000 m, 8 Jun 1987, Deil2648 (Fig. 7E). 

*-Ma, Rif, Djebel Tidighin, 2000 m, lO Ju11989, Deil4867 (Fig. 7F). 

This is the first report of a chromosome number for this variety, found in the W parts of 

the Betic Range in S Spain as well as in the C Rif and C High Atlas mountains of Morocco. 

Unlike Anthemis pedunculata varo pedunculata, only diploid chromosome numbers 

were found (Table 12). 

Table 12. Chromosomes of Anthemis pedunculata varo discoidea (OTU 5; n = 3). 



4.99 4.16 9.15 13.4 1.19 m 

Il 5.75 3.05 8.80 12.9 1.88 sm 

III 4.59 3.85 8.44 12.4 1.19 m 

IV 5.37 2.57 7.94 11 .6 2.08 sm 

V 4.25 3.15 7.40 10.9 1.34 m 

VI 5.27 1.14 7.19 10.5 4.62 st-sat 

VII 4.06 3.06 7.12 10.4 1.32 m 

VIII 3.69 2.85 6.54 9.6 1.29 m 

IX 4.02 0.97 5.77 8.5 4.14 st-sat

54 


Anthemis pedunculata subsp. atlantica 

n=9 

- Tn, Kasserine, Djebel Chambi, 1540 m, 7 May 1994, Vogt 12606 & Oberprieler 6911. 

n = 18 

- Tn, Siliana, Forèt de Kesra, 1030-1100 m, 18 May 1994, Vogt 13510 & Oberprieler 7815. 

- Tn, Siliana, Forèt de Kesra, 1040-1100 m, 18 May 1994, Vogt 13523 & Oberprieler 7828. 

This is the first report of a chromosome number for this taxon, endemie to the mountainous 

areas of E Algeria and N Tunisia. 

Anthemis pedunculata subsp. turolensis 

n=9 

-Hs, Cuenca, Beteta, 1350 m, 20 May 1995, Vogt 15350 & Oberprieler 9659. 

This taxon, endemie to the mountains of C Spain, had been studied cytologically only 

once (Benedf i Gonz,Hez 1987), and found to be diploid. The present count corroborates 

these data. 

Anthemis punctata subsp. punctata 

n = 18 

- Tn, Beja, Djebel Goraa, 780 m, 3 May 1994, Vogt 12238 & Oberprieler 6543. 


- Tn, El Kef, TabIe de Jughurta, 1200-1270 m, 5 May 1994, Vogt 12466 & Oberprieler 6771 . 

- Tn, Zaghouan, Djebel Zaghouan, 750-850 m, 23 May 1994, Vogt 13823 & Oberprieler 8128. 

2n= 36 


This species, confined to the mountains of NE Algeria and N Tunisia, had not been 

studied cytologically. The four specimens studied were all found to be tetraploid, same as 

the closely related Anthemis cupaniana from Sicily (see below). 

Anthemis cupaniana 

2n= 36 

-Si, Madonie, Collesano - Petralia, 1590 m, 26 May 1994, Vogt 13938 & Oberprieler 8243. 

-Si, Madonie, Quacella, 28 Jul 1990, HB Catania (cult. in HB Berlin No. 298-04-91-10). 

The findings of a tetraploid chromosome number confrrrns earlier reports by Brullo & 

al. (1978a), Brullo & al. (1988), and Devesa & al. (1988). Though no locality is given, it 

is probable that Capineri' s (1968) report of 2n = 36 for "A. punctata" was based on Sicilian 

material and also relates to A. cupaniana.

Bocconea 9 - 1998 55 

c 

D 

Fig. 7. Metaphase plates in Anthemis. - A: A. pedunculata varo pedunculata, root tip mitosis, 2n = 

36 (Vogt 9638 & Oberprieler 4074). B: A. pedunculata varo pedunculata, root tip mitosis, 2n = 37 

(Vogt & Oberprieler 3937). C: A. pedunculata varo discoidea, PMC meiosis, n = 9 (Vogt & 

Oberprieler 4012). D: A. pedunculata varo discoidea, PMC meiosis, n = 9 (Vogt & Oberprieler 

4012). E: A. pedunculata varo discoidea, root tip mitosis, 2n = 18 (Deil 2648). F: A. pedunculata 

varo discoidea, mot tip mitosis, 2n = 18 (Deil4867). - Scale bars = lO ~m .


Anthemis (subg. Anthemis) sect. Maruta 

Anthemis cotula 

2n = 18 

*-Ga, HB Dijon (cui t. in HB Berlin No. 116-02-93-10). 

The report of 2n = 18 chromosomes for Anthemis cotula fits the numerous previous 

counts published for this wide-spread species. 1-2 accessory B-chromosomes were observed 

by Yavin (1970) on Greek and Yugoslavian material. The French plants studied 

here were found to have two paiis of satellited subtelocentric chromosomes, which agrees 

with observations by Yavin (1970), Mitsuoka & Ehrendorfer (1972), Bartolo & al. (1978), 

Kuzmanov & al. (1981), and Benedf i Gonzalez (1987). In Portuguese populations, Fernandes 

& Queir6s (1971) found karyotypes with tandem satellites but also one population 

with only one pair of satellited chromosomes; while Uitz (1970) figures a karyotype without 

any satellites. The karyotype found in the present study consists of 6 pairs of metacentric, 

1 pair of submetacentric, and 2 pairs of satellited subtelocentric chromosomes 

(Table 13). Interchromosomal asymmetry (Al) has the lowermost value of alI samples of 

the present study. No plants from N Africa were available for study. 

Table 13. Chromosomes of Anthemis cotula (OTU 16; n = 14). 

Chromo- Long arm Short arm 

some pair (IJm) (IJm) 

Totallength Relative length 

(IJm) (%) 

Arm ratio 

Chromosome 

type 

2.71 

Il 2.42 

III 2.69 

IV 2.22 

V 2.44 

VI 2.06 

VII 2.24 

VIII 1.97 

IX 1.77 

2.19 

1.97 

0.75 

1.82 

1.40 

1.71 

0.66 

1.57 

1.43 

4.90 13.8 

4.39 12.4 

4.22 11 .9 

4.04 11.4 

3.84 10.8 

3.77 10.6 

3.68 10.4 

3.54 10.0 

3.20 9.0 

1.23 

1.22 

3.58 

1.21 

1.74 

1.20 

3.39 

1.25 

1.23 

m 

m 

st-sat 

m 

sm 

m 

st-sat 

m 

m 

Anthemis pseudocotula 

2n = 18 + 1B 

*-Cy, Paphos-Harbour, 3-20 m, 2 May 1991, Vogt 9000. 

Same as Yavin (1970) who studied plants from Palestine, we found a chromosome 

number of 2n = 18 plus a small accessional B-chromosome on a Cyprian representative of 

this species. N African plants of Libyan provenance were studied by Brullo & al. (1990), 

who found no B-chromosome. The karyotype of Anthemis pseudocotula had not yet been 

studied. Unlike A. cotula, only one of its two subtelocentric chromosome pairs is satellited 

(Table 14), but as discussed above for A. cotula, this character may show considerable 

variation.

Bocconea 9 - 1998 

57 

Table 14. Chromosomes of Anthemis pseudocotula (OTU 17; n = 3). 

Chromo- Long arm Short arm 

some pair (I.lm) (I.lm) 

3.49 2.71 

Il 3.16 2.69 

III 2.98 2.47 

IV 3.47 1.06 

V 3.58 1.51 

VI 2.75 2.26 

VII 3.66 1.14 

VIII 2.82 1.60 

IX 2.38 2.00 

Totallength Relative length Arm ratio 

(I.lm) (%) 

6.20 13.4 1.28 

5.85 12.62 1.17 

5.45 11 .7 1.20 

5.31 11.4 3.27 

5.09 11 .0 2.37 

5.01 10.8 1.21 

4.80 10.3 3.21 

4.42 9.5 1.76 

4.38 9.4 1.19 

Chromosome 

type 

m 

m 

m 

st-sat 

sm 

m 

st 

sm 

m 

Anthemis (subg. Cota) sect. Cota 

Anthemis altissima 

2n = 18 

*-Ga, Aude, Alzonne, HB Liège (cult. in HB Berlin No. 194-17-92-10). 

*-Ga, Aude, Montolieu, HB Liège (cult. in HB Berlin No. 194-18-92-10). 

*-Ga, Aude, Sigean, HB Liège (cult. in HB Berlin No. 194-19-92-10). 

*-Ga, Aude, Villedubert, HB Liège (cult. in HB Berlin No. 194-20-92-10). 

*-Ga, Aude, Montolieu, HB Liège (cult. in HB Berlin No. 152-07-93-10). 

*-It, Imperia, Perinaldo, 520 m, HB Genova (cult. in HB Berlin No. 268-01-92-10). 

My chromosome counts for Anthemis altissima agree with all previous indications, by 

Delay (1971) for French, Capineri & al. (1978) for Italian, Stephanou & Georgiadis 

(1982) for Greek, Kuzmanov & Georgieva (1977) and Kuzmanov & al. (1980) for Bulgarian, 

Strid (1980) for Turkish, and Podlech & Dieterle (1969) for Afghanian material. 

Karyotypes have been reported by Uitz (1970) and Kuzmanov & al. (1981). As they correctly 

indicate, the karyotype of A. altissima differs markedly from those of the other Anthemis 

species studied by me (Table 15): The two · pairs of satellite chromosomes are not 

subtelocentric but submetacentric, the arm ratio being on average 2.48 (vs. 4.03) for the 

long and 2.68 (vs. 4.36) for the short pair of satellited chromosomes. With relative lengths 

of (on average) 12.5 % for the long and 11.6 % for the short satellited pair they are among 

the long or medium chromosomes of the set, not among the medium to short ones (11 .2% 

and 9.8% on average, respectively) as in the other Anthemis species 

Anthemis austriaca 

2n = 18 

-Au, Niederosterreich, Erlauftal, 220 m, HB Salzburg (cult. in HB Berlin No. 007-11-93-10). 

No N African plants were available for study. The diploid chromosome number of 2n = 

18 chromosomes found on material of Austrian provenance agrees with most previous 

reports for this species, e.g. by Kuzmanov & Kozuharov (1970), Kuzmanov & al. (1980), 

and Kuzmanov & al. (1981) for Bulgarian, Stephanou & Georgiadis (1982) for Greek, and


Vachova & Majovsky (1980) for Slovakian plants. Uitz (1970) reports diploid chromosome 

numbers for several garden provenances of Anthemis austriaca, along with a 

tetraploid number for material from Turkey. Although A. austriaca supposedly belongs to 

the same section as A. altissima, both Uitz (1970) and Kuzmanov & al. (1981) found its 

satellited chromosomes to be subtelocentric not submetacentric. 

Table 15. Chromosomes of Anthemis altissima. 



OTU18(n=4) 

I 4.95 4.04 8.99 14.4 1.22 m 

Il 4.99 2.20 7.97 12.7 2.26 sm-sat 

III 5.11 1.88 7.77 12.4 2.71 sm-sat 

IV 3.82 3.31 7.13 11.4 1.15 m 

V 3.71 3.02 6.73 10.8 1.22 m 

VI 3.31 2.85 6.16 9.9 1.16 m 

VII 3.53 2.61 6.14 9.8 1.35 m 

VIII 3.87 2.03 5.90 9.4 1.90 sm 

IX 4.20 1.68 5.88 9.4 2.50 sm 

OTU 19 (n = 13) 

I 5.35 4.16 9.51 14.7 1.28 m 

Il 5.45 2.06 8.29 12.8 2.64 sm-sat 

III 4.24 3.59 7.83 12.1 1.18 m 

IV 5.06 1.75 7.59 11.7 2.89 sm-sat 

V 3.84 3.15 6.99 10.8 1.21 m 

VI 4.24 2.26 6.50 10.0 1.87 sm 

VII 3.42 2.86 6.28 9.7 1.19 m 

VIII 4.22 2.00 6.22 9.6 2.10 sm 

IX 3.26 2.52 5.78 8.9 1.29 m 

OTU 20 (n = 6) 

I 4.78 4.09 8.87 15.0 1.16 m 

Il 4.66 1.79 7.23 12.2 2.60 sm-sat 

III 3.72 3.17 6.89 11.7 1.17 m 

IV 4.52 2.31 6.83 11.6 1.95 sm 

V 4.16 1.59 6.53 11 .1 2.61 sm-sat 

VI 3.50 2.81 6.31 10.7 1.24 m 

VII 3.24 2.51 5.75 9.7 1.29 m 

VIII 3.76 1.97 5.73 9.7 1.90 sm 

IX 2.73 2.31 5.04 8.5 1.18 m 

OTU 21 (n = 12) 

I 4.98 4.02 9.00 14.5 1.23 m 

Il 4.94 2.19 7.91 12.8 2.25 sm-sat 

III 3.99 3.40 7.39 11 .9 1.17 m 

IV 4.77 1.81 7.36 11.9 2.63 sm-sat 

V 3.67 2.97 6.64 10.7 1.23 m 

VI 4.05 2.20 6.25 10.1 1.84 sm 

VI I 3.29 2.75 6.04 9.8 1.19 m 

VIII 4.24 1.79 6.03 9.7 2.36 sm 

IX 3.03 2.45 5.48 8.8 1.23 m

Bocconea 9 - 1998 59 

Table 15 (continued). 



OTU 22 (n = 12) 

I 4.58 3.78 8.36 14.6 1.21 m 

Il 4.48 1.87 7.13 12.4 2.39 sm-sat 

III 3.69 3.12 6.81 11.9 1.18 m 

IV 4.26 1.59 6.63 11.6 2.67 sm-sat 

V 3.44 2.85 6.29 11.0 1.20 m 

VI 3.84 1.97 5.81 10.1 1.94 sm 

VII 3.18 2.53 5.71 10.0 1.25 m 

VIII 3.74 1.67 5.41 9.4 2.23 sm 

IX 2.95 2.35 5.30 9.3 1.25 m 

OTU 23 (n = 3) 

I 5.04 3.94 8.98 14.4 1.27 m 

Il 4.90 1.78 7.46 11.9 2.75 sm-sat 

III 3.89 3.46 7.35 11.8 1.12 m 

IV 4.78 2.34 7.12 11.4 2.04 sm 

V 3.71 3.20 6.91 11.1 1.15 m 

VI 4.37 1.72 6.87 11.0 2.54 sm-sat 

VII 3.49 2.85 6.34 10.1 1.22 m 

VIII 4.17 2.14 6.31 10.1 1.94 sm 

IX 2.97 2.35 5.32 8.5 1.26 m 

Anthemis (subg. Cota) sect. Anthemaria 

Anthemis tinctoria 

2n = 18 

*-Au: Niederosterreich, Persenberg, 200 m, HB Salzburg (cult. in HB Berlin No. 007-14-93-10). 

The present count agrees with the many previous reports for this species, e.g. by Strid 

(1980) for Turkish, Strid & Franzén (1981) for Greek, Capineri (1971) for Italian, Kuz- 

Table 16. Chromosomes of Anthemis tinctoria (OTU 25; n = 4). 



5.23 4.30 9.53 14.5 1.21 m 

Il 4.60 3.99 8.59 13.1 1.15 m 

III 4.08 3.49 7.57 11 .5 1.16 m 

IV 4.10 3.17 7.27 11.1 1.29 m 

V 4.74 2.29 7.03 10.7 2.06 sm 

VI 4.79 1.39 6.96 10.6 3.44 st-sat 

VII 3.76 3.06 6.82 10.4 1.22 m 

VIII 4.04 2.04 6.08 9.3 1.98 sm 

IX 4.24 0.89 5.91 9.0 4.76 st-sat


manov & Kozuharov (1970) and Van Loon & Van Setten (1982) for Bulgarian, and Rostovceva 

(1983) for Sibirian plant material. The karyotype was described by Kuzmanov & 

al. (1981) to consist of one pair of metacentric, six pairs of submetacentric, and two pairs 

of satellited subtelocentric chromosomes. I found 5 pairs of metacentric, 2 pairs of submetacentric, 

and 2 pairs of satelIited subtelocentric chromosomes (Table 16). 

Discussion 

Karyotypes pr.oved to be rather uniform throughout the genus. Most species have five 

pairs of metacentric, two pairs of submetacentric, and two pairs of satellited subtelocentric 

chromosomes in the diploid chromosome sets. The most deviating karyotype was found in 

Anthemis altissima of A. sect. Cota, where satelIited chromosomes are submetacentric 

rather than subtelocentric. Following Stebbins' s (1971) classification of karyotypes with 

respect to interchromosomal and intrachromosomal asymmetry, ali species studied, even 

A. altissima, falI into class 2A. As Romero Zarco (1986) has shown usirig karyotypes of 

representatives of Aveneae (Gramineae), Stebbins's classification is rather coarse when 

taxa with minute differences in karyotype asymmetry are studied. He therefore proposed a 

more sensitive method to assess interchromosomal and intrachromosomal asymmetry, 

using two indices. The results for the 24 OTUs studied are shown in Fig. 8. The x axis is 

formed by Romero's Al and indicates intrachromosomal asymmetry increasing from left to 

right, while Romero's A 2 is reported on the y axis and shows interchromosomal 

(karyotype) asymmetry increasing from bottom to top. In this diagram we find A. altissima 

welI separated from the other representatives of the genus, which is caused by the combination 

of low Al and high A 2 values, i.e. high intrachromosomal and low interchromosomal 

symmetry. A. sect. Anthemis proves to be very variable in respects to karyotype symmetry: 

While populations of A. confusa and A. melampodina show a pronounced intrachromosomal 

asymmetry, chromosomes are much more symmetrical in A. ubensis 

(OTU 15) and the Spanish population of A. arvensis (OTU 7), which reach the same values 

of intrachromosomal symmetry as A. altissima. The karyotypes of A. sect. Hiorthia are 

less variable in their intrachromosomal but more variable in interchromosomal symmetry. 

The plants of A. pedunculata (OTU 3, 4, 5) and one of A. maritima (OTU 2) falI within 

the range of A. sect. Anthemis, but the other population of A. maritima (OTU 1) has a 

karyotype with a very high interchromosomal asymmetry. The perennial representative of 

A. subg. Cota, A. tinctoria (OTU 24), is welI nested within the group of the perennials of 

A. subg. Anthemis (OTUs 1-5). The annuals of A. sect. Maruta, A. pseudocotula and 

A. cotula (OTUs 16 and 17), show rather low interchromosomal and interchromosomal 

asymmetry, but are stili close to the representatives of A. sect. Anthemis. 

According to Heywood & Humphries (1977) and Bremer & Humphries (1993), x = 9 is 

the commonest base number in the Anthemideae. Other base numbers occur only in a few 

genera. Descending dysploidy is found in Cancrinia (x = 7) of the Cancriniinae, Pentzia 

(x = 8, 6) of the Matricariinae, Ursinia (x = 8, 7) and Athanasia (x = 8, KalIersjo 1991) of 

the Ursiniinae, and Artemisia (x = 9, 8, 7, 6) of the Artemisiinae. Ascending dysploidy is 

reported for Lasiospermum (x = 9, lO), Osmitopsis, Inezia, and Thaminophyllum (x = lO) 

of the Thaminophyllinae, and Hilliardia, Cotula sect. Cotula, Soliva (x = lO), Leptinella 

(x = 13), and Cotula sect. Strongylosperma (x = 17) of the Cotula group in Matricariinae 

(Bremer & Humphries 1993). Until now, approximately 280 chromosome number reports

Bocconea 9 - 1998 61 

had been published for Anthemis species. All indicate that the base chromosome number 

in this genus is x = 9 and most chromosome counts in the present study are in agreement 

with this assumption. The deviating counts of 2n = 19, 20, 37 reported for plants of Anthemis 

ubensis, A. confusa, and A. pedunculata, respectively, are interpreted as resulting 

from aneuploid changes. 

Polyploidy is common in the tribe. Extensive polyploid complexes are found in Achillea 

(2-lOx), Artemisia (2-12x), Dendranthema (2-lOx), Leptinella (-I2x), and Leucanthemum 

(2-22x, Vogt 1991). Polyploid series in Anthemis are less marked, and only members 

of the A. eretica group reach the hexaploid (A. carpatica, Kiipfer 1974, Baltisberger 1993; 

A. eretica subsp. pontica, Oberprieler unpubl.) or octoploid level (A . carpatica, Kiipfer 

1974). Polyploidy is almost exclusively found in the perennials of A. sect. Hiorthia. Reports 

of polyploids in the other sections are sparse and confined to triploids in A. arvensis 

(Kuzmanov & al. 1981) and A. triumfetti (Gonzalez Zapatero & Elena Rossell6 1986) and 

one reported occurrence of a tetraploid in A. austriaca (Uitz 1970). Five N African taxa of 

Anthemis, all members of A. sect. Hiorthia, were found to be tetraploid: A. abylaea, 

,17 

A 2 

,1 6 

20 19 

4----- -- -- 8 

,15 

\ 21 ./ 

\6 

,14 

,13 

i 

2rJ! 

-,d 

~ ... , 

4 

 

13 

'V 

........................... 

~. 

·········'i7.12 

................... 


A. punctata, and A. eretica subsp. columnae consistently so, A. pedunculata (present 

study) and A. maritima (Benedf i Gonzalez 1987) comprising both diploid and tetraploid 

cytotypes. All other N African representatives of Anthemis studied cytologically are 

diploids. 

Aneuploid changes in the chromosome number were observed in three N African species. 

Plants from two populations of Anthemis ubensis were found to have 2n = 19 chromosomes, 

2n = 20 chromosomes were counted in a population of A. confusa, and 2n = 37 

chromosomes were found in plants from a tetraploid population of A. pedunculata in Morocco. 

In all cases, the accessory chromosomes were metacentric or submetacentric. The 

occurrence of aneuploidy on a diploid level contrasts with Vogt's (1991) observations on 

Spanish Leucanthemum, that aneuploid chromosome numbers occur only at the tetraploid 

and higher polyploid levels but never among diploids. Aneuploidy on a diploid level in 

Anthemis, however, had already been reported by Uitz (1970; A. ruthenica, 2n = 20), 

Kuzmanov & al. (1981; A. orbelica, 2n = 22; A. stribrnyi, 2n = 21), and Benedf i Gonzalez 

(1987; A. maritima, 2n = 20). Hypertetraploid chromosome numbers, found once in 

the present study, had been reported also by Kuzmanov & al. (1981; A. hinkovae, 2n = 39) 

and Kiipfer (1974; A. saxatilis, 2n = 37, 38, 39). It is plausible that ascending aneuploid 

changes are better tolerated than descending ones, since the gai n of an extra chromosome 

affects viability less ser10us than a loss (Grant 1981 : 363). Hypodiploids (monosomics or 

nullisomics), therefore, are unlikely to be found under natural conditions. Unlike in diploids, 

however, aneuploidy at higher ploidy levels may al so be due to the loss of chromosomes. 

Hypotetraploid numbers were reported by Benedf i Gonzalez (1987; A. saxatilis, 

2n = 35) and Kiipfer (1974; A. saxatilis, 2n = 35). Hypohexaploidy was found by Baltisberger 

(1993; A. carpatica, 2n = 53), and hypo-octoploidy by Kiipfer (1974; A. carpatica, 

2n = 66, 68, 70). According to Grant (1981: 363) this "polyploid drop", i.e. the loss of one 

or more chromosomes in a polyploid, is often tolerated because of the duplication of genetic 

material in the homologous chromosomes. Gottschalk (1976: 356) and Khush (1973: 

Il) explained aneuploid chromosome numbers by uneven distribution of chromosomes 

during mitotic or meiotic cell divisions. When non-disjunction events in the germ line of 

somatic tissues or during meiosis affects the chromosome number of gametes, the accessory 

chromosome(s) will be found in alI cells of the offspring. In higher polyploids, especially 

autopolyploids, the uneven distribution of chromosomes during meiosis is likely 

enhanced by the formation of multivalents, whereas in diploids it may be caused by distorted 

pairing of homologous chromosomes and the formation of univalents during meiosis, 

consequent to hybridisation. It is hardly a fortuitous coincidence that the two hyperdiploid 

chromosome numbers reported here were found in populations which belong to 

species likely to be of hybrid ori gin (A. ubensis) or supposedly influenced by hybridisation 

in the corresponding area (A. confusa). 

B chromosomes are accessory chromosomes that differ from the basic chromosomal 

complement (A chromosomes) by their smaller size, varying number, and greater degree 

of heterochromatisation (Stebbins 1971). They are considered to be fragments homologous 

to parts of A chromosomes, and suspected to originate from them through disturbance 

during cell divisions, but they do not pair with the homologous A chromosomes 

during meiosis (Nagl 1980). B chromosomes may vary in number during ontogeny or in 

different tissues of the same plant and are considered to have but limited taxonomic significance 

(Stuessy 1990). In the present study, B chromosomes were observed in Anthemis

Bocconea 9 - 1998 63 

maroccana, A. melampodina, and A. pseudocotula. B chromosomes in Anthemis species 

had been reported previously by Yavin (1970) for A. cotula (1-2 B) and A. pseudocotula 

(1 B), Uitz (1970) for A. carpatica (1-4 B), Benedf i Gonzalez (1987) for A. pedunculata 

varo pedunculata (1 B; sub A. tuberculata varo tuberculata) and A. pedunculata subsp. 

turolensis (1 B; sub A. tuberculata subsp. turolensis), and Georgiou (1990) for A. werneri 

subsp. werneri (1 B) and A. peregrina subsp. peregrina (1-4 B). 

Karyotypes were studied comprehensively in the present work. Comparable surveys 

were made previously only by Uitz (1970) who studied 8 and Kuzmanov & al. (1981) who 

studied 21 Anthemis species. Mitsuoka & Ehrendorfer (1972) also presented detailed 

analyses of karyotypes and furthermore studied experimental hybrids and their meiotic 

chromosome pairing. Uitz (1970) found prominent differences in structure and total size of 

karyotypes among Anthemideae, in parallel to phylogenetic relationships between species, 

sections and genera as deduced from other characters. 

1,9 

R 

1,8 

1,7 

1,6 

1,5 

1,4 

1,3 

1,2 

52 

26 

,.+-- - - - ___ _ 

50 ,. 46 

•.........../- ... .................*.... 

.,. 

: i 

j 

1J/'/ 

.' 

51 

... ~ 

48 .... ~7 

.. .. u ......u .... e 

49 

23 

6 

25 

-----:(> 

/~40 

/,/


When compared with other genera of the Anthemideae (Fig. 9, lO), the sections of Anthemis 

cluster together due to their asymmetrical chromosomes (intrachromosomal asymmetry, 

Asl) and moderately bimodal karyotype (interchromosomal asymmetry, R). Most 

other genera have a more symmetrical karyotype, but in some representatives of Matricaria 

and Tanacetum even greater asymmetry is found. 

Some interesting patterns emerge when one compares karyotype differentiation in the 

different subtribes of Anthemideae, as defined by Bremer & Humphries (1993) on morphological 

grounds. 

The Chrysantheminae, which Bremer & Humphries (1993) considered to be the sister 

group of Anthemidinae, are represented by three genera (Fig. 9). While the perennial Argyranthemum 

(OTU 25) was found to have quite similar symmetry values to Anthemis, the 

annual Chrysanthemum (OTUs 26 and 27) and Ismelia (OTU 28) show very low intrachromosomal 

asymmetry. 

The Achilleinae, as defined by Bremer & Humphries (1993), also exhibit a considerable 

increase of chromosomal symmetry from perennial to annual life form (Fig. lO): The 

perennial species of Achillea (OTUs 29 and 30) fall within the range of Anthemis, the 

perennial (OTU 33) and annual species (OTUs 31 and 32) of Anacyclus and the unispecific 

annual genus Cladanthus (OTU 37) stand apart, with Chamaemelum (OTUs 34- 

36) linking Anthemis to Anacyclus and exhibiting considerable intrageneric variation. Of 

the three Chamaemelum species studied, the perennial C. nobile (OTU 34) has the karyotype 

least similar to Anthemis, although Mitsuoka & Ehrendorfer (1972) reported that it is 

easily crossed with A. cotula, and found frequent chromosome pairing during meiosis and 

high pollen fertility in the resulting hybrid. They therefore considered the two genera to be 

cJosely related. However, since only a single intergeneric cross was made, and bearing in 

mind that crossability is considered a primitive trait and the formation of reproductive 

barriers progressive (Funk 1985, but see discussion in Stuessy 1990: 201), crossability of 

C. nobile and A. cotula may be a symplesiomorphic feature rather than reflecting true 

phy logenetic affinity. 

In the Leucantheminae (Fig. 9), a considerable shift is detectable from the moderately 

asymmetrical karyotype of the long-lived perennial Rhodanthemum (OTU 46) to the very 

symmetrical karyotypes of the short-lived perennials of Leucanthemum (OTU 47 and 48) 

and the annual Coleostephus myconis (OTU 49). Lepidophorum (OTU 50), a unispecific 

annual genus considered by Bremer & Humphries (1993) to be a basai member of their 

Leucantheminae, shows largely symmetrical chromosomes and on karyological, but also 

on fruit morphological grounds (Uitz 1970), may rather belong to the Chrysantheminae. 

Tanacetum, as the only representative of Bremer & Humphries's (1993) provisional 

subtribe "Tanacetinae ", was represented in Uitz's (1973) studies by three species, the 

perennials T. vulgare (OTU 38) and T. parthenium (OTU 39) and the annual T. annuum 

(OTU 40). They show similar intrachromosomal symmetry values (Fig. lO) but differ 

markedly from each other in their interchromosomal asymmetry, the two perennials having 

a moderately to extremely symmetrical karyotype, the annual an extremely bimodal one. 

The same trend is noted in Bremer& Humphries's (1993) subtribe Matricariinae (Fig. 

lO), represented by the perennials of Tripleurospermum (OTUs 41-43) and the annuals of 

Matricaria (OTUs 44 and 45): Tripleurospermum falls within range of Anthemis, Matricaria 

(at least the species M. recutita OTU 44) shows highly symmetrical chromosomes 

but a very asymmetrical karyotype.

Bocconea 9 - 1998 65 

The relatively similar karyotypes found in the predominantly perennial genera Anthemis, 

Achillea, Argyranthemum, Tanacetum, and Tripleurospermum led both Uitz 

(1970) and Mitsuoka & Ehrendorfer (1972) to consider these genera as rather centrai 

(basaI) within the Anthemideae. From these, progressive karyotype divergence led to 

mostly more symmetrical, but sometimes to more asymmetrical (e.g. Matricaria) karyotypes. 

This assumption is supported by morphology. The mentioned genera ali belong to 

different subtribes in Bremer & Humphries's (1993) classification, and it is more parsimonious 

to consider similarities of karyotypes as the sympiesiomorphic condition rather than 

suppose an independent origin of similar karyotypes from quite dissimilar ones within 

each subtribe. Anthemis, therefore, wouId beiong to the basaI stock of genera in the Anthemideae, 

showing reiations to Tanacetum as well as to Achilleinae, Chrysantheminae, 

and Matricariinae through their basaI genera Achillea, Argyranthemum, and Tripleurospermum, 

respectiveIy. 

2,0r--------------------------------------------------, 

1 ,8 

1,6 

R 

45 

+. 

". 

........... 

..... 

......... .............. 

............ ...... 

...•, ...... 

..... 

........ 

40 

.11 

........ 

................. 

. ...• / 

. ......... 44 

.+ 

/36 ......... 

............ 

······· ..... 41 

.. :t> 

............ 

1,4 

37 

31 

O 

..... 

33 

// 

i 

34 

3~i 

56 

58 

60 62 64 

Fig. lO. Intrachromosomal (Asl) and interchromosomal (R) asymmetry in representatives of 

Achilleinae (solid line; O = Achillea, O = Anacyclus, bo. = Chamaemelum, V = Cladanthus), 

Tanacetinae (broken line; = Tripleurospermum, 

+ = Matricaria). OTU numbers and data sources are expicited in Table l and 2. The 

indices Asl and R were assessed using the formulae of Barghi & al. (1989).


Indications of gross structural differences in the karyotypes of Anthemideae are very 

sparse. Small structural changes, along with karyotype differentiation due to hybridisation, 

are considered as the main mechanism underlying karyotype divergence (Mitsuoka & 

Ehrendorfer 1972). As Nagl & Ehrendorfer (1974) found, karyotype differentiation is 

more rapid in annuals than in perennials perhaps because annuals show shorter celI cycles 

and enhanced celI elongation, thus a more rapid development. 

Mitsuoka & Ehrendorfer (1972) stated that structural differentiation of chromosomes in 

the Anthemideae, aIready occurs at the infraspecific level. They gave Anthemis cotula as 

an example of a species in which the number and formation of satellites is very variable. 

The same is observed in A. secundiramea where Brullo (in Bartolo & al. 1979) found a 

Sicilian population of A. secundiramea varo secundiramea to lack satellited chromosomes, 

our own material from Sicily revealed one pair, and plants of A. secundiramea var. 

cossyrensis from Tunisia, two pairs of satellited chromosomes. In A. arvensis subsp. incrassata, 

considerable differences in karyotype symmetry between N African and S Spain 

populations were detected in the present study. 

11. Pollen morphology 

First detailed light rnicroscopic studies of Anthemideae pollen were made by W odehouse 

(1926, 1935) on Anthemis cotula and Chamaemelum nobile. He described the pollen 

as echinate, having a co arse-granular, two-Iayered exine. The thicker inner layer appeared 

to him to be built of coarse radiaI striae, while the outer, much thinner layer was 

found to have very fine radiaI striae. He considered the exine sculpturing as a mai n difference 

within the tribe, with echinate pollen grains with sharply pointed spines characterising 

Anthemis together with other insect-pollinated genera (Leucanthemum, Chrysanthemum, 

Tanacetum) and non-echinate pollen grains with vesti gal or entirely absent spines 

being typical for wind-pollinated genera like Artemisia (Woodhouse 1935). 

Fig. Il. Anthemis gharbensis (Vogt 10161 & Oberprieler 4609). - A: Acetolysed pollen grain in 

polar view; equatorial inside diameter (E) as measured far the comparative analysis of pollen size 

indicated. B: Acetolysed pollen grain in equatorial view; inside length of polar axis (P) indicated.

Bocconea 9 - 1998 67 

Stix (1960), in her comprehensive light microscopical work on pollen morphology of 

Compositae, described her "Anthemis type", to which she also assigned representatives of 

Achillea, Chamaemelum, Chrysanthemum, Leucanthemum, Cotula, and Matricaria, as 

having a tegillate sexine. She found the inner, coarsely striate layer of the sexine to consist 

of rather thick and distally branched infrategillary baculae, while the outer, finely striate 

layer is formed of fine pila with heads mostly fused together (intertegillary baculae). Her 

findings were later corroborated by transmission electron microscopy (TEM) and scanning 

electron microscopy (SEM) (e.g. Skvarla & al. 1977, Vezey & al. 1994; the latter authors 

use the term "double tectum" for the outer layer of baculae which they misleadingly call 

"infratectal columellae", while the infrategillary baculae are called "basaI columellae"). 

Besides the mentioned authors, pollen of Anthemis was studied by Cigurjaeva & Tereskova 

(1983), Benedf i Gonzalez (1987), Fedoroncuk & Savitskii (1988), and De Leonardis 

& al. (1991). In all species studied, pollen grains were found to be rather uniformly 

spheroidal, trizonocolporate and spiny, with tenuimarginate, sharply pointed colpi and 

lalongate, tenuimarginate, sharply pointed ora. No qualitative features were found to define 

subgroups. However, pollen dimensions vary markedly between species. Benedf i 

Gonzaléz (1987) found annual representatives of A. subg. Anthemis to have conspicuously 

smaller pollen than the perennial representatives of this subgenus. 

Vogt (1991) has demonstrated that dimensions of pollen grains are correlated with 

ploidy level in Leucanthemum. I therefore aimed to find out whether diploids and 

tetraploids in Anthemis as well can be separated by pollen measurements, i.e. whether 

pollen dimensions can help to assess of chromosome numbers, e.g. for types and other 

herbarium specimens, especially from presently inaccessible areas. For that purpose, it was 

thought desirable to exclude variation of exine thickness and spine length by using inside 

measures of equatorial diameter and length of polar axis rather than total equatori al diameter 

and totallength of polar axis (Fig. Il). 


A total of 104 pollen samples (Table 17) were taken from herbarium specimens representing 

35 N African Anthemis taxa. All 25 species accepted in the present revision were 

included in the survey. Pollen was treated with the acetic anhydride/sulfuric acid 

(acetolysis) mixture ofErdtman (1960), washed with destilled water, mounted in glycerine 

jelly, and the slides were sealed with Paraplast. The material was studied with an Leitz 

Dialux 20 microscope, and measurements were made using an eyepiece micrometer. 

Equatorial inside diameter (E) of pollen grains was measured in polar view in at least 50 

grains per sample; inside length of the polar axis (P) was measured in equatorial view in at 

least 30 grains per sample. In both cases, the inner boundary of the nexine was used as 

starting and ending points for the measurements (Fig. Il). 

Results and discussion 

Pollen grains of all studied representatives of the genus were found to be uniformly 

spiny, spheroidal, and trizonocolporate. Dimensions of pollen grains vary conspicuously. 

Equatorial inside diameters (E) were found to range between 12.5 fIm and 25.4 fIm, and 

the inside length of polar axis (P), between 13.3 fIm and 27.5 fIm. The index of equatorial 

diameter and polar axis (PIE), however, was found to be fairly constant and usually falls


Table 17. Pollen grain dimensions (mean ± standard deviation) in Anthemis. - E: 

Equatorial inside diameter; P: Inside length of polar axis; ploidy, when inferred 

fram pollen measurements, is given in square brackets. 

OTU specimen ploidy E (IJm) P (IJm) PIE 

Anthemis sect. Hiorthia 

A. abylaea 

1 Ma: Beni Hosmar, 29.6.1930, Maire (P) [4x) 19.51 t 0.92 21.15t 1.04 1.08 

A. eretica subsp. columnae 

2 Ag: Dj. Tamesguida, 6. 1890, Battandier(G) [Type of [?) 19.61 t 0.84 20.31 t 1.04 1.04 

A. numidica) 

A. maritima subsp. maritima 

3 Hs: Cabo Roche, Vogt 9733 & Oberprieler4181 (B) 4x [?) 18.83 t 0.69 18.86tO.61 1.00 

4 Tn: Cap Blanc, Vogt 13792 & Oberprieler 8097 (B) 4x 22.01 t 1.50 21 .29 t 1.27 0.97 

5 Ga: Col du Tende, Vogt 15352 & Oberprieler 9661 (B) 4x 20.67 t 0.55 22.36 t 1.14 1.08 

6 Ag: La Calle, 2.5. 1914, Clavé (G) [?) 18.18tO.73 19.38 t 0.96 1.07 

7 Ag: Bone, 14.5.1906, Romieux(G) [4x) 20.21 t 0.83 22.24 t 0.89 1.10 

8 Sa: Su Palloon, 31.5. 1983, Charpin (G) [?) 18.60t 1.50 19.57 t 1.54 1.05 

A. maritima subsp. bolosii 

9 Ag: Philippeville, 11.5. 1853, Cosson (P) [2x] 17.50t 0.62 17.68 t 0.58 1.01 

10 Ag: Slora, 23.6. 1930, Maire (P) [2x] 17.01 t 0.70 18.14±0.91 1.07 

11 Ag: Philippeville, 5. 1858, Choulette (MPU) [2x] 16.77 ± 0.67 18.50± 0.56 1.10 

12 Ag: Slora, 5. 1913, Battandier(MPU) [2x) 16.58 t 0.77 17.00tO.77 1.03 

A. maritima subsp. pseudopunctata 

13 ~g : Cap Aokas, 5. 1896, Reverchon (B) [Type) [?) 19.31 t 0.76 20.50 t 1.01 1.06 

A. pedunculata varo pedunculata 

14 Herb. Desfontaines (P) [Type) [2x) 16.12 t 0.62 18.99t 0.72 1.18 

15 Hs: Sierra Nevada (G) [4x) 19.66 ± 0.69 20.81 t 0.97 1.06 

16 Ag: Tiarel, Pomel (P) [Type of A. granulata) [?) 17.90 t 1.03 19.56t 1.13 1.09 

17 Ag: Miliana, 6. 1856, Pomel(MPU) [Type of A. tenuisecta) [2x) 17.21 t 1.07 17.97 t 0.97 1.04 

18 Ma: Tazzeka, 6. 1925, Maire (P) [Type of A. laeviscula) [?) 18.43 t 0.55 18.90tO.71 1.03 

19 Hs: Sierra Nevada (G) [Type of A. tuberculata varo [2x] 19.27tO.81 16.74 t 0.66 0.87 

microcephala) 

20 Ma: Dj. Touchka, 1876, Ibrahim (MPU) [Type of [?) 19.51 t O.71 20.07 t 0.83 1.03 

A. punctata varo maroccana) 

21 Ag: Bossuel, 6.6. 1927, Faure (MPU) [Type of [?) 18.58 t 0.86 20.13 t 0.83 1.08 

A. punctata varo microcepha/a) 

22 Hs: Sa. de los Filabres, 28.7. 1995, Rico (SALA 59886) [2x) 18.58 t 0.93 18.26 t 1.22 0.98 

23 Ag: Yakouren, Podlech 39294 (G) [2x) 16.37 t 0.45 16.13 t 0.55 0.99 

24 Ag: Aurès, Podlech 38937 (MSB) [2x] 17.75 t 0.60 18.17 t 0.57 1.02 

25 Ag: Tizi N'Kouilal, Podlech 39137 (MSB) [?) 18.94t 0.85 18.63 t 0.81 0.98 

26 Ag: Takerbouzl, Podlech 3899

Bocconea 9 - 1998 69 

Table 17. (continued) 


42 Ma: Foum Kheneg, Vogt 9440 & Oberprie/er 3878b (B) 4x 18.44 ± 0.65 23.90± 0.93 1.30 

43 Ma: Tahout·ou·Fillali, Vogt 11948 (B) 4x 18.67 ± 0.87 22.39 ± 0.97 1.20 

44 Ma: Dj. Lekst, Vogt 11829 & Oberprie/er 6277 (B) 4x 20.47 ± 1.08 20.43 ± 1.33 1.00 

45 Ma: Col du Zad, Vogt 11988 (B). 4x 18.64 ± 0.75 23.50 ± 0.89 1.26 

46 Ma: Ain Leuh, Vogt 14949 & Oberprie/er 9258 (B) 4x 21 .12 ± 0.91 21.38 ± 0.90 1.01 

47 Ma: Azrou . Ilrane, Vogt 15016 & Oberprie/er 9325 (B) 4x 21 .99± 0.95 25.07 ± 1.06 1.14 

48 Ma: Ain Leuh, Vogt 14927 & Oberprie/er 9236 (B) 2x[4xj 20.83 ± 1.42 22.79± 1.37 1.09 

A. peduncu/ata var. discoidea 

49 Ma: Ril, Azrou, 26.6. 1926, Maire(MPU) [Type 01 [2xj 17.00 ± 0.77 16.92 ± 0.79 1.00 

A. peduncu/ata I. eradiata) 

50 Hs: Sa. de las Nieves (G) [Type) [4xj 20.24 ± 0.78 20.65 ± 0.95 1.02 

51 Ma: Assil·n·Arous, Grane 170 (RNG) [2xj 18.23 ± 0.94 17.99 ± 0.99 0.99 

A. peduncu/ata subsp. atlantica 

52 Ag : Dj. Tougour, Ba/ansa 967 (G) [?) 17.68± 1.08 19.54± 1.23 1.11 

53 Ag : Kenchala, Pome/ (MPU) [Type) [?) 18.24 ± 0.88 21.14± 1.29 1.16 

54 Ag: Dj. Babor, Pod/ech 39380 (MSB) [2x) 16.47 ± 1.07 16.42± 1.01 1.00 

55 Ag : Dj. Gouffi, 10.7.1861, Gosson (P) [2x) 16.44 ± 0.56 17.62±0.66 1.07 

56 Ag : Dj. Gouffi, 10.7.1861, Gosson (P) [2x) 16.42 ± 0.68 18.25± 1.00 1.11 

57 Ag: Dj. Babor, 20.7.1880, Gosson (P) [4x) 19.86 ± 1.26 21.64 ± 1.15 1.09 

58 Ag: Dj. Babor, 7.1897, Reverchon (P) [2xj 17.48 ± 0.76 18.85±0.86 1.08 

59 Ag: Roumel, Reboud 1677 (G) [?) 18.38 ± 0.47 19.11 ±0.72 1.04 

60 Ag : Oued Zenati, 27.5.1911, G/ay (P) [4xj 20.02 ± 0.92 20.94 ±0.86 1.05 

61 Agl Rummel, 28.5.1840, Durieu (P) [?) 18.75±0.99 19.15± 1.63 1.02 

62 Ag: Dj. Chelia, Pod/ech 38760 (MSB) [2xj 17.95± 1.34 18.94± 1.04 1.06 

63 Ag: Oued Abdi, 4.6. 1853, Gosson (P) [2xj 17.77 ± 0.85 18.44 ± 0.64 1.04 

64 Ag: Dj. Cheliah, 11.6.1853, Gosson (P) [4x) 21.20± 1.12 21.32 ± 0.97 1.01 

65 Ag: Adghar Amellal, 1.7. 1880, Gosson (P) [2x) 15.27± 0.65 16.08±0.91 1.05 

66 Tn: Dj. Semata, 21.5.1887, Letourneux(P) [?) 19.67 ± 0.84 20.19±0.76 1.03 

67 Tn: Souk el Djema, 28.5.1887, Letourneux (P) [?) 19.92 ± 0.69 20.01 ± 0.83 1.00 

68 Tn: Dj. Meghila, 16.5.1887, Letourneux (P) [4x] 20.63 ± 1.22 22.03± 1.25 1.07 

A. peduncu/ata subsp. cJausonis 

69 Ag: Castiglione, G/auson (MPU) (Type] [?) 18.61 ± 1.17 18.39± 1.38 0.99 

70 Ag: Zeralda, Battandier(MPU) [2xj 16.66 ± 0.75 17.08 ± 1.09 1.03 

A. peduncu/ata subsp. turo/ensis 

71 Hs: Beteta, Vogt 15350 & Oberprie/er 9659 (B) 2x 15.98 ± 0.53 15.72 ± 0.74 0.98 

A. punctata subsp. punctata 

72 Ag: Tamesguida, Battandier(P) [Type 01 A. punctata [?) 18.33 ± 0.70 19.03 ± 0.66 1.04 

varo baborensis) 

73 Tn: Dj. Zaghouan, Vogt 13823 & Oberprie/er 8128 (B) [?) 19.05 ± 0.68 18.10±0.71 0.95 

74 Tn: Dj. Dyr Vogt 12353 & Oberprie/er 6658 (B) [?) 18.67 ± 0.59 18.76± 1.11 1.00 

75 Ag: Mecid Sidi Aicha, 8.1880, Reboud(P) [?) 18.10±0.75 19.17 ± 0.93 1.06 

A. punctata subsp. kaby/ica 

76 Ag : Tirourda, 6.1882, Battandier[Type) [?) 18.41 ±0.67 19.24 ± 0.62 1.05 

77 Ag: Tala Guilel, Davis 53190 (RNG) [?] 18.73 ± 0.96 18.61 ± 1.10 0.99 

78 Ag : Tirourda, 7.1896, Reverchon (G) [?) 19.76 ± 0.66 20.47± 0.66 1.04 

79 Ag: Tirourda, 9.7. 1909, Saint-Lager(G) [2xj 17.95 ± 0.98 18.10± 1.31 1.01 

Anthemls sect. Anthemis 

Anthemis ser. Anthemis 

A. arvensis subsp. incrassata 

80 Ma: Tetouan, Vogt 10061 & Oberprie/er4509 (B) [2xj 16.42 ± 0.90 16.35 ± 0.69 1.00 

Anthemis ser. Chrysanthae 

A. boveana 

81 Ag: Oran, 9.5.1907, Faure (B) [2x) 16.83 ± 0.76 17.22 ± 0.73 1.02


Table 17. (continued). 


A. chtysantha 

82 Ag: Bou-Sler, Dubuis n° 10733 (MSB) [?) 18.95 ± 1.38 19.32 ± 0.86 1.02 

A. gharbensis 

83 Ma: Arbaoua, Vogi 10161 & Oberprie/er4609(8) [Type) [2x) 17.48 ± 0.88 16.93 ± 0.82 0.97 

A. maroccana subsp. maroccana 

84 Ma: Selrou, Vogt 10296 & Oberprie/er4744 (B) [2x) 16.22 ± 0.73 16.01 ± 0.65 0.99 

A. maroccana subsp. aguilarii 

85 Ma: Azib de Ktama, Font Quer441 (G) [2x) 17.04 ± 0.90 17.07±0.83 1.00 

A. tenuisecta subsp. tenuisecta 

86 Ma: Essaouira. Vogt 11930 (B) [2x) 16.75± 1.09 17.99 ± 0.99 1.07 

87 Ma: Tiznit, Vogi 11905 & Oberprie/er 6353 (B) [2x) 15.26 ± 1.78 16.56±2.19 1.09 

A. tenuisecta subsp. jahandiezii 

88 Ma: Sali, Jahandiez 115 (8) [2x) 15.04 ± 0.56 15.39±0.75 1.02 

Anthemis ser. Secundirameae 

A. confusa 

89 Tn: Gabès, Kralik 356 (G) [Type) [2x) 16.94 ± 0.97 16.44 ± 0.64 0.97 

90 Tn: Galsa. Vogt 12712 & Oberprie/er 7017(B) [2x) 16.22 ± 1.31 16.01 ± 1.06 0.99 

A. cyrenaica 

91 Li: Benghasi, Ruhmer 182 (GOET) [2x) 16.11 ±0.68 16.12 ± 0.67 1.00 

A. g/areosa 

92 Li: Misurata. Davis 49829 (RNG) [2x) 15.33±0.83 15.50± 0.59 1.01 

A. secunairamea varo secundiramea 

93 Tn: La Haouaria. Davis 56896 & Lamond (RNG) [?) 19.42 ± 1.45 19.90 ± 1.22 1.03 

A. taubertii 

94 Li: Baiada. Davis 49985 (RNG) [2x) 16.23± 0.72 16.00±0.69 0.99 

A. ubensis 

95 Tn: Kesra, Vogt 13524 & Oberprie/er 7829 (B) 2x 17.22± 1.22 16.56± 0.98 0.96 

Anthemis ser. Bourgaelnlanae 

A. mauritiana subsp. faurei 

96 Ma: Saidia, Vogi 10872 & Oberprie/er 5320 (B) 2x[?) 19.88 ± 1.04 19.89 ± 0.95 1.00 

A. monilicostata 

97 Ag: Sa"ida, 20.3.1872, Warion (G) [?) 19.58± 1.07 20.79± 1.14 1.06 

A. stiparum subsp. stiparum 

98 Ag: Allou, Pod/ech 34022 (MSB) [?) 18.67± 0.89 18.60 ± 0.75 1.00 

A. stiparum subsp. sabulico/a 

99 Ag: Ghardaia, Podlech 35529 (MSB) [2x) 17.58 ± 1.28 17.32 ± 1.19 0.99 

A. stiparum subsp. intermedia 

100 Ag: Ain Ben Khelil, Kralik n° 194 (GO ET) [Type) [2x) 16.93 ± 0.95 16.65 ± 0.98 0.98 

A. zaianica 

101 Ma: Tougroulmès, Vogt 14840 & Oberprie/er 9149 (B). 2x 18.21 ±0.93 18.04±0.94 0.99 

Anthemls sect. MaTuta 

A. cotula 

102 Ma: Imasinen, 13.6. 1929. Maire (P) [2x) 16.58 ± 0.75 16.17 ± 0.64 0.98 

A. pseudocotu/a 

103 Li: Timimi, Davis 50308 (RNG) [2x) 17.09±0.99 16.96± 1.02 0.99 

Anthemls sect. Cata 

A. austriaca 

104 Tn: Tataouine, Vogi 13151 & Oberprieler 7456 (B) [2x) 16.13±0.96 16.28 ± 0.80 1.01

Bocconea 9 - 1998 71 

within the range of 0.9-1.1 (see Table 17). Deviating, prolate-spheroidal pollen grains with 

indices of 1.2-1.3 were observed only in three tetraploid populations of Anthemis peduneulata 

varo peduneulata from the Middle Atlas mountains of Morocco (OTUs 42, 43, 45). 

Pollen size differences between diploids and tetraploids are well marked: equatorial inside 

diameter (E) is (12.5-)15.3-19.9(-22.5) 11m in diploids, (16.7-)18.4-22.0(-25.4) 11m in 

tetraploids. (Values in brackets denote the range of single pollen grains, non-bracketed 

values, the range of means.) The inside length of the polar axis (P) was found to be 

(13.3-)15.7-20.6(-22.5) 11m in diploids and (16.6-)18.1-25.1(-27.5) 11m in tetraploids. In 

spite of considerable overlap of values, the combination of both measures makes it possible 

to discriminate between diploids and tetraploids (Fig. 12). The overlap zone corresponds 

to three representatives of tetraploid Anthemis maritima (OTU 3) and A. punetata 

(OTUs 73-74) and two diploid representatives of A. peduneulata (OTU 35) and A. mauritiana 

(OTU 96). Within this overlap zone pollen measurements do not permit to assess 

with confidence the ploidy level of the OTUs. Another exceptional case is OTU 48, found 

to have a diploid chromosome number but whose pollen dimensions fall fully within the 

range of tetraploid representatives of A. peduneulata. This is explained by the fact that at 

the locality in question - as in many places in Morocco - one finds a mixture of diploid 

24 

23 

E 

47 

22 o 40 o 

o 

21 

20 o 

73 35 

46 

o 48 

o 

41 44 o 

o o 

19 o 3 o 

7~ 43 45 

o o o 42 

o 

o 

18 99 38 39 

83 00 38 

95 o o o 

o 

17 86 

o 

71 80 37 

16 

000 

15 

87 

o 

14 P 

14 16 18 20 22 24 26 

Fig. 12. Pollen dimensions of OTUs with known chromosome number (O diploids; 

= tetraploids). E = equatorial inside diameter (~m); P = inside length of polar axis (~m). 

Numbers of OTU s refer to numbers gi ven in Table 17.


and tetraploid plants of A. pedunculata, and the plant studied palynologically was not the 

same individuai studied cytologically. In the following, the range of pollen dimensions of 

known diploids and tetraploids was used to assess ploidy levels in cytologically unchecked 

specimens. 

In contrast to Benedi i Gonzalez (1987), who had studied Spanish representatives of 

Anthemis, I could detect no difference between pollen sizes of annual and perennial diploids. 

Since Benedi i Gonzalez (1987) included spines in his measurements of equatorial 

diameter and length of polar axis, this seerning contradiction obviously reflects differences 

in the thickness of the exine or the length of spine s, between the annuals and perennials. 

Further measurements of exine thickness and spine length in annuals and perennials of 

Anthemis are needed to confirm this assumption. However, since the annual representatives 

of Anthemis are usually considered to be autogamous while the perennial representatives 

are allogamous, the reduction of exine thickness and, moreover, the reduction of 

spine length in annuals is plausible. 

Fig. 13 displays pollen dimensions of annuals of Anthemis sect. Anthemis, sect. Maruta, 

and sect. Cota. AH representatives of the two latter sections and most of sect. Anthemis 

Fig. 13. Pollen dimensions in Anthemis sect. Anthemis (O = A. ser. Anthemis; O = A. ser. 

Bourgaeinianae; a = A. ser. Chrysanthae; V = A. ser. Secundirameae), A. sect. Cota (

Bocconea 9 - 1998 73 

clearly fall within the range of typical diploids, but five OTUs are found in the overlap 

zone between diploids and tetraploids: A. secundiramea var. secundiramea (OTU 93) of 

A. ser. Secundirameae, A. chrysantha (OTU 82) of A. ser. Chrysanthae, and A. mauritiana 

(OTU 96), A. monilicostata (OTU 97), and A. stipa rum subsp. stipa rum (OTU 98) of 

A. ser. Bourgaeinianae. The last mentioned series seems to be characterised by distinctly 

larger pollen grains than most other annual representatives of the genus. 

In Anthemis maritima (Fig. 14), representatives of A. maritima subsp. maritima and subsp. 

pseudopunctata were found to fall within the pollen size range of tetraploids or in the 

overlap zone between diploids and tetraploids, while the four studied specimens of 

A. maritima subsp. bolosii have conspicuously smaller pollen grains. Since this subspecies 

also has smaller achenes, capitula, ray florets, etc. than A. maritima subsp. maritima and 

subsp. pseudopunctata, it is reasonable to assume that A. maritima subsp. bolosii is a 

diploid taxon. Diploid chromosome numbers in A. maritima have been indicated by 

Benedf i GonzaIez (1987) from Menorca (Balearic Islands) and by Harling (1950) and 

Mitsuoka & Ehrendorfer (1972) for plants of unstated provenance. 

24 

23 

E 

22 

o 

21 

20 

o 

7 

o 

19 

o 

18 

17 

12 

o 

o 

10 

o 11 

o 

16 

15 

14 

14 16 

18 

20 

22 

p 

24 26 

Fig. 14. Pollen dimensions in Anthemis maritima subsp. maritima (O), subsp. bolosii CO), and 

subsp. pseudopunctata (L::.).E = equatorial inside diameter (Jlm); P = inside length of polar axis 

(Jlm). Dotted lines delimit the area of overlap between diploid (bottom left) and tetraploid (top 

right) OTUs (see Fig. 12).


Fig. 15 and Fig. 16, plotting pollen dimensions found in the perennial Anthemis pedunculata 

- A. punctata complex and A. eretica subsp. columnae, suggest ploidy levels for 

some of the studied herbarium specimens. As found by cytological study, A. pedunculata 

varo pedunculata is represented by diploids and tetraploids. The type specimens of 

A. pedunculata (OTU 14) and A. tuberculata varo microcephala (OTU 19) definitively fall 

within the group of diploids by their pollen size. The type specimens of A. granulata 

(OTU 16), A. laeviuscula (OTU 18), A. punctata varo maroccana (OTU 20), and 

A. punctata varo microcephala (OTU 21) were found to have intermediate pollen dimerisions, 

so that no suggestion on their ploidy level is possible. A specimen from the Sierra 

Nevada (OTU 15), labelled as "neotype" of A. tuberculata by Benedf i Gonzalez in 1984 

(but not designated neotype in Benedf i Gonzalez 1987), has rather large pollen suggesting 

tetraploidy. No concrete data on tetraploidy in Spanish representatives of A. pedunculata 

(= A. tuberculata) yet exist, but counts for this taxon are very scarce (Aparicio & Silvestre 

1985, BIanché & al. 1985, Benedf i Gonzalez 1987), and further cytological studies of 

Spanish plants may welI yield similar results as for Morocco, where both cytotypes were 

found to grow side by side in most populations studied. The same may apply for A. pedunculata 

varo discoidea, since the pollen dimensions of the type of that name from the Sierra 

24 

23 

E 

.. ........•. 

..................... .. 

34 

° 

22 '. '. 

21 

20 

19 

18 

17 

...... 

.................. 

............•........... 

............................... 

~" 

...... ......•.. 

......... 

41 44 

..... ° 050 

······6.7 o 60 

i~ .. 1!f 

o" ;~? 

21 ° 

..... ~28 61 

··.... 2J9 . ..... ........ 

53 

" 

40 

° 

sa 

" 

57 

" 

48 

° 

43 

............•. ° 

...... 

45 

° 42 

° 

47 

° 

16 

15 

14 

14 

65 

" " ~ ......... . 

16 18 20 

....... 

........ 

....•. 

22 

... .... 

.......... 

24 

p 

....... 

26 

Fig. 15. Pollen dimensions in Anthemis pedunculata varo pedunculata (O), varo discoidea (O), 

subsp. atlantica (..6.), subsp. clausonis (V), and subsp. turolensis (

Bocconea 9 - 1998 75 

de Grazalerna in S Spain (OTU 50), also fall within the range of the tetraploids, aIthough 

no tetraploid chrornosome count yet exists for this taxon, ali its studied Moroccan representatives 

being unequivocally diploid. Pollen dirnensions suggest that A. pedunculata 

subsp. atlantica (OTUs 52-68) similarly occurs on two ploidy levels, but ali representatives 

of A. pedunculata subsp. clausonis (OTUs 69-70) and subsp. turolensis (OTU 71) 

fall within the pollen size range of the diploids, which in the latter case confirrns the results 

of cytological studies by Benedf i Gonzalez (1987) and rnyself. 

In Anthemis punctata, A. abylaea, and A. eretica subsp. columnae, even in plants suspected 

to be tetraploid due to their morphological characteristics, or known to be tetraploid 

(OTUs 73-74), pollen dimensions did not normally reach the dimensions typical for 

tetraploids, but fell within the overlap zone between diploids and tetraploids, or even within 

diploid values (OTU 79). Obviously, pollen dirnensions are not unrestrictedly comparab1e 

between different species groups of Anthemis. 

24 

23 

E 

.......•. 

.. ...•. 

.... 

22 

21 

20 

19 

....... 

.... ...... 

.......•. ..... 

73 

..•..•. .... 

........ 78 

2 ... 

v ··· ....... o 

.......... 

................•. 

18 

o 

.............................• 

17 

......•. 

..... 

16 

.....•. 

15 

14 

14 

16 

18 

20 

22 

24 

p 

26 

Fig. 16. Pollen dimensions in Anthemis abylaea (O), A. eretica subsp. columnae (V), A. punctata 

subsp. punctata (O), and subsp. kabylica (.D.). E = equatorial inside diameter (flm); P = inside 

length of polar axis (flm) . Dotted lines delimit area of overlap between diploid (bottom left) and 

tetraploid (top right) OTUs (see Fig. 12).


12. Species delimitation in tbe Anthemis boveana group 

The Anthemis boveana group, together with the more remote species A. chrysantha, 

forms A. ser. Chrysanthae within sect. Anthemis, being characterised by the yellow colour 

of the ray florets and endemic to the W Mediterranean region. While A. chrysantha is 

restricted to the surroundings of Oran (Algeria) and Cartagena (Spain), the A. boveana 

group - formerly treated as a single species, A. boveana - is also found in the vicinity of 

Oran but is furthermore scattered throughout Morocco, especially in the N and W parts of 

the country. The species group varies considerably throughout its range, which has led to 

the description of several species, varieties and forms. Gay (in Bory & Durieu 1848-1850) 

described A. boveana from the coastal areas around Oran. Ball (1873: 365) added 

A. tenuisecta from the N foothills of the High Atlas S of Marrakech in Morocco. Battandier 

& Pitard (in Pitard 1918, 1931) considered plants collected in the surroundings of Fès 

in N Morocco to be rather similar to A. boveana but distinct enough to deserve the status 

of a new species, A. maroccana. Maire (1923) found it impossible to separate the Moroccan 

taxa from the Algerian ones on the specific level and, together with his new 

A. boveana varo jahandiezii, treated them as varieties of A. boveana. Further infraspecific 

taxa were added by Maire & Sennen (in Maire 1933: A. boveana varo aguilarii) to accornrnodate 

plants from the Rif mountains, and Maire (in Jahandiez & Maire 1934: 

A. boveana [varo tenuisecta] f. elongata) for deviating populations in the F6ret de la 

Mamora near Rabat. Anticipating the result presented here, the latter taxon was raised to 

specific'rank by Oberprieler (1994) as A. gharbensis. 

To assess morphological variability within and between populations of the Anthemis 

boveana group and to assist in defining taxonomic limits, numerical techniques were used. 

Additionally, attention was paid to qualitative morphological and anatomical features of 

the achene, which are known to be of considerable importance for taxon delimitation in 

this group. 


33 specimens (OTUs) representative of the entire morphological and geographical 

range of the Anthemis boveana group (Table 18) were measured or scored for 26 morphological 

characters (Table 19, Fig. 17). The OTUs were either single herbarium specimens 

or sets of duplicates from the same gathering. On each specimen 3-10 measurements 

were made for each character, the results averaged, and the mean values used in the . 

multivariate analysis (Table 20). Principal component analysis (PCA) was performed 

using the SPSS FOR WINDOWS software package. Missing values (marked by a question 

mark in Table 20) were replaced by the mean value of ali other OTUs for the corresponding 

character. 

Results and discussion 

Principal component analysis (PCA) based on the complete data set (Table 20) yielded 

six components with eigenvalues higher than 1, accounting for 77.6 % of the total variance 

in the data set. Loadings of characters on the first three components, which explain 58.8 % 

of the total variance, are presented in Table 21. The frrst principal component (PC 1; 28.6 

% of the total variance) is dominated by variables describing the dimensions of involucres 

(invol, oibl, mibl, iibl), ray and disc florets (rayl, rayw, discl, discm), and the shape of pale

Bocconea 9 - 1998 77 

Table 18. OTUs used in the multivariate analysis of the Anthemis boveana 

group. 

No. 

Location 

Specimen 

A. boveana 

1 Ag: Oran, au Cagneret. 

2 Ag: Oran, Cap Canastel, 100 m. 

3 Ag : Broussailles avoisinant le pont de la Macta. 

4 Ag: Pelouses rocailleuses à Gambetta. 

5 Ag: Oran. 

6 Ag : Oran, au Cagneret. 

7 Ag: Canastel, près Oran. 

8 Ag: [ ... ) ad promontorium Canastel, 100 m. 

A. maroccana subsp. aguilé!rii 

9 Ma: Alias Rifain, Azib de Ketama, 1600 m. 

10 Ma: Rif SW: Suberain de Rhomara, 1450 m. 

11 Ma: Atlas Rifain, à Isaguen, 1650 m. 

A. maroccana subsp. maroccana 

12 Ma: Middle Atlas, Massif du Kandar S Sefrou. 

13 Ma: Meknès, vers D'Khissa, bords des champs. 

14 Ma: entre Kasbet el Hadjeb et Dar Ca'id Ito. 

15 Ma: In lapidosis calcareis prope jugum Zegotta. 

A. gharbensis 

16 Ma: Casablanca. 

17 Ma: Mamora, 5km N Sidi-Allal-Bahraoui. 

18 Ma: Mamora. 

19 Ma: La Mamora. 

20 Ma: Mamora, Sidi-Allal-el-Babraui. 

21 Ma: Rharb, Arbaoua - Moulay Bousselham. 

22 Ma: C. El Araix. 

23 Ma: circa Tingidem: Charf-el-Agab. 

A. tenuisecta subsp. jahandiezii 

24 Ma: entre l'Oued Tensift & Souk et Tleta. 

25 Ma: Safi. 

A. tenuisecta subsp. tenuisecta 

26 Ma: Djebileh, au N de Marrakech. 

27 Ma: c. 10km N Tamri. 

28 Ma: Safi, Tamanar, loco dicto Tamri. 

29 Ma: 6 km S Ait-ou-Mribeta. 

30 Ma: 15 km N Tiznit. 

31 Ma: c. 14.3 km N Tiznit. 

32 Ma: route d'Agadir à Tiznit. 

33 Ma: à l'embouchure de l'O. Massa. 

9/30 May 1907, Faure (B) 

Maire n° 7489 (B) 

Balansa n° 13 (GOET; P; G) 

13 May 1909, Faure (G) 

May 1889, s. coli. (Herb. Girod) (G) 

9/30 May 1907, Faure (G) 

13 May 1936, Faure (G) 

23 Apr 1934, Maire (B, G) 

20 Jun 1934, Sennen & Mauricio n° 

9406 (MPU-AtN, G) 

11 Jul 1959, Sauvage (MPU-AfN) 

6 Ju11932, Sennen & Mauricio (MPU-AfN) 

Vogt 10296 & Oberprieler4744 (B) 

Jahandiez 159 (G) 

Benoist 550 (P) 

21 Apr 1926, Maire (P) 

May 1887, Méllerio (P) 

Podlech 43489 (MSB, G) 

15 May 1939, Sauvage (MPU-Sauvage) 

23 Apr 1928, Braun-Blanquet (MPU 

Braun-Blanquet) 

Vogt 10259 & Oberprieler 4707 (B) 


Font Quer n° 666 (G) 

27 Apr 1924, Maire (MPU-AfN, P, G) 

Romieux 1348 (G) 

Jahandiez 115(B, G) 

21 Apr 1921, Wilczek(G) 

Vogt 11930 (B) 

Fernandez Casas & al. 9171 (B, G) 

Podlech 42958 (MSB) 



Maire & Wilczek 551 (G) 

Sauvage 3998 & al. (MPU-Sauvage) 

tips (pa\et), whereas PC 2 (21.1 % of total variance) is dominated by variabIes describing 

the Iength/width ratio of invo\ucral bracts (oibi, mibi, iibi) and Ieaves (Ieafi), and the number 

of ray florets (rayno). On PC 3, which exp\ains onIy 9.0 % of the total variance (and


therefore strongly falls back behind the first two principal components), there are only few 

variables with high loadings describing length and shape of ultimate leaf segments (ulsl, 

ulsi) and dimensions of achenes (acheI, achew). 

Table 19. Characters used in the numerical analysis of the Anthemis boveana 

group (Ieafl and coron not used) and the A. punctata - A. pedunculata complex 

(rayw and discm omitted). See Fig. 17 for details of measurements. 

Symbol 

Leaves 

leafl 

leafi 

ulsl 

ulsi 

Capitula 

invol 

rayno 

Characters (unit of measure / character states) 

length of basai leaves (m m) 

length-width ratio of basalleaves (Ieafl / leafw) 

length of ultimate leaf segments of basalleaves (mm) 

length-width ratio of ultimate leaf segments of basalleaves (ulsl / ulsw) 

diameter of involucrum (mm) 

number of ray florets per capitulum 

Involucral bracts 

oibl length of outer involucral bracts (mm) 

oibi length-width ratio of outer involucral bracts (oibl /oibw) . 

oibs shape of outer involucral bracts (1 = ovate, 2 = elliptical, 3 = obovate) 

mibl length of middle involucral bracts (mm) 

mibi length-width ratio of middle involucral bracts (mibl / mibw) 

mibs shape of middle involucral bracts (1 = ovate, 2 = elliptical, 3 = obovate) 

iibl length of inner involucral bracts (mm) 

iibi length-width ratio of inner involucral bracts (iibi / iibw) 

iibs shape of inner involucral bracts (1 = ovate, 2 = elliptical, 3 = obovate) 

Pales 

palel 

palei 

pales 

palet 

length of pales (mm) 

length-width ratio of pales (pale I / palew) 

shape of outer pales (1 = ovate, 2 = elliptical, 3 = obovate) 

shape of tip of pales (1 = tricuspidate, 2 = acute, 3 = acuminate, 4 = blunt, 

5 = emarginate) 

Ray florets 

rayl length of ray florets, including tubular part (mm) 

rayw width of ray florets (m m) 

rayi length-width ratio of ray floret limbs (raylb / rayw) 

Disc florets 

discl totallength of disc florets (mm) 

disci ratio totallength / basai part of disc florets (discl / discb) 

discm length of distai appendages of corolla lobes (mm) 

Achenes of peripheral disc florets 

achei length of achenes, eXcluding corona (m m) 

achew width of achenes (mm) 

coron maximal length of corona (mm)


79 

leafl 

lulsw 

r11(f\mmm 

1 2 3 4 5 

palet 

coron 

acheI 

Fig. 17. Measurements made for the numerical analysis ofthe Anthemis boveana group and the 

A. pedunculata - A. punctata complex (Table 19).


Table 20. Data far the multivariate analysis af the Anthemis boveana group 

(character symbals are explained in Table 19 and Fig. 17). 

OTU lesti ulsl ulsl Invol rayno olbl olbl oibs mlbl mlbl mibs iibl iibi 

1 

2 

3 

4 

5 

6 

7 

1.80 1.90 2.80 9.17 

1.70 2.00 2.80 10.33 

2.10 2.20 3.50 9.67 

1.70 2.50 3.80 10.67 

1.90 3.10 4.30 10.00 

1.70 2.40 2.70 10.83 

1.70 1.90 2.60 10.33 

8 2.40 2.30 2.40 11 .67 

9 2.20 1.60 3.10 9.10 

10 1.60 2.00 2.60 8.33 

11 2.20 1.80 3.00 8.00 

12 1.60 2.10 2.80 11 .00 

13 1.90 2.50 4.80 10.33 

14 2.30 1.90 3.20 9.83 

15 1.90 1.70 3.40 10.83 

16 2.50 1.60 3.10 8.77 

17 2.50 2.10 3.60 8.50 

18 3.20 1.80 3.20 8.93 

19 2.90 2.10 3.30 9.00 

20 2.40 2.20 3.80 9.23 

13.00 3.17 2.50 1.17 4.13 2.70 2.50 4.10 2.68 

13.33 3.70 3.00 1.50 4.80 2.47 2.67 4.70 2.59 

12.33 2.90 2.67 1.00 4.67 3.21 2.33 4.27 3.43 

13.33 3.00 2.41 1.00 4.00 3.20 1.67 4.10 2.93 

14.00 3.27 2.80 1.00 3.97 2.70 1.83 4.20 2.99 

11 .673.272.651 .174.272.77 2.17 3.93 2.89 

10.00 2.73 2.16 1.17 3.77 2.06 2.00 3.90 2.29 

12.33 3.47 2.45 1.17 4.47 2.51 2.17 4.93 2.25 

? 2.57 3.69 1.17 3.17 2.78 1.83 3.83 2.68 

9.00 2.203.161.172.933.042.172.803.11 

7.00 1.50 2.14 1.00 2.60 2.17 2.00 2.80 2.80 

13.67 3.30 3.53 1.50 4.60 3.39 2.33 4.30 3.61 

13.00 2.90 3.33 1.33 3.43 3.55 2.17 3.37 3.63 

13.00 2.03 2.19 1.00 3.80 2.95 2.00 3.97 2.79 

13.00 2.97 3.41 1.17 3.83 3.30 1.67 4.20 3.25 

12.33 2.67 3.41 1.50 3.07 3.44 1.83 3.50 3.83 

12.67 2.13 3.32 1.33 2.87 3.22 1.67 3.13 3.37 

12.33 2.27 3.70 1.17 2.87 3.19 1.50 3.23 4.10 

13.00 2.60 3.30 1.75 3.90 3.55 2.00 3.65 3.85 

12.00 2.63 3.45 1.33 3.27 3.40 1.83 3.13 3.78 

21 2.40 2.10 3.20 9.67 13.67 3.07 4.29 1.17 3.57 4.19 1.50 3.80 4.40 

22 2.30 1.90 2.90 9.33 12.00 2.80 3.48 1.17 3.47 3.26 2.50 3.17 3.97 

23 2.10 2.50 3.80 11.00 13.50 3.00 3.75 1.25 3.40 3.57 2.00 3.80 3.15 

24 1.70 2.00 3.10 9.00 13.00 2.25 2.65 1.25 3.25 2.72 2.25 3.35 3.71 

25 1.60 1.80 3.80 7.07 12.33 2.33 3.07 1.33 3.20 3.51 2.00 2.73 4.04 

26 1.60 1.60 3.20 6.50 8.00 1.75 1.86 1.75 2.40 2.21 2.00 2.60 2.91 

27 1.40 2.60 4.30 8.67 7.33 2.43 2.97 1.33 2.77 2.87 2.00 2.97 3.31 

28 1.50 2.70 4.20 7.00 7.33 1.73 2.54 1.1 7 2.37 3.10 2.1 7 2.20 2.75 

29 1.30 1.90 2.60 6.50 5.67 2.73 2.85 1.83 3.07 2.97 2.00 2.90 3.48 

30 1.90 1.80 2.40 6.40 7.00 2.03 2.46 1.1 7 2.67 2.50 2.00 2.63 2.76 

31 1.30 2.50 3.70 7.37 7.33 2.93 3.28 1.33 3.17 3.19 2.17 3.23 3.52 

32 1.70 1.80 2.80 9.17 10.00 2.27 2.92 1.50 2.83 3.18 1.83 3.27 4.12 

33 1.30 2.20 4.70 6.50 7.00 2.37 2.99 1.33 2.93 3.04 2.00 2.93 3.40 

The resulting arrangement of OTUs in the two-dimensional factor space of PC 1 : PC 2 

and PC 1 : PC 3 is shown in Fig. 18-19. It shows a go od correlation of morphological 

variation and geographical origino With the exception of Anthemis boveana varo jahandiezii 

(OTUs 24-25), variation is found to be much smaller within groups of specimens 

from the same area than between these groups. This supports the assumed morphological 

distinctness of the taxa as here circumscribed. 

In terms of PC 1, major differences were found between the Algerian populations of 

Anthemis boveana (OTUs 1-8) on the one hand and the W and N Moroccan populations of 

A. gharbensis (OTUs 16-23), A. tenuisecta (OTUs 26-33), A. boveana varo aguilarii


81 


iibs palel palei pales palet rayl rayw rayi discl disci discm achei achew OTU 

2.83 3.33 3.42 3.00 2.00 9.37 4.07 1.91 2.87 1.73 0.24 1.27 0.78 

2.83 3.33 3.63 2.17 2.17 11 .80 3.77 2.68 3.33 1.89 0.28 1.27 0.76 2 

2.67 3.20 3.70 2.67 2.00 9.67 4.07 2.02 2.93 2.17 0.29 1.13 0.72 3 

2.83 3.83 2.95 2.83 2.33 9.93 3.87 2.06 3.27 1.75 0.19 1.33 0.78 4 

2.50 3.47 3.74 2.50 2.00 9.23 3.53 2.15 3.10 1.94 0.18 1.31 0.61 5 

2.50 3.80 3.64 2.00 2.00 9.27 3.53 2.10 3.27 1.69 0.26 1.18 0.87 6 

2.83 3.53 2.49 2.33 2.00 13.57 5.17 2.14 3.60 1.72 0.21 1.25 0.69 7 

2.67 3.77 2.83 2.67 2.83 15.17 6.70 1.91 3.50 1.78 0.24 1.33 0.85 8 

2.17 3.20 4.06 2.50 2.83 7.43 3.77 1.59 3.10 2.27 0.19 1.80 1.00 9 

2.33 3.50 4.81 2.33 2.50 5.10 2.77 1.42 3.17 2.07 0.13 1.85 1.03 10 

2.00 3.60 4.00 2.50 3.00 6.70 3.70 1.49 3.50 2.33 0.25 1.85 1.05 11 

2.50 3.57 4.80 2.17 3.00 8.17 4.30 1.52 3.27 2.01 0.21 1.57 0.87 12 

2.33 3.50 5.00 2.00 3.00 10.37 4.07 2.00 3.47 1.96 0.23 1.58 0.82 13 

2.67 3.40 5.40 2.50 2.17 8.03 3.50 1.81 3.03 2.17 0.25 1.80 0.78 14 

2.50 3.70 5.64 2.00 2.50 6.80 3.07 1.78 2.87 2.15 0.23 1.52 0.83 15 

2.00 3.07 6.33 2.17 3.83 7.87 3.00 2.10 2.30 2.88 0.12 1.17 0.55 16 

2.33 3.17 

2.33 3.30 

2.25 3.45 

2.00 3.73 

2.17 3.57 

3.00 3.03 

3.00 3.65 

2.50 2.95 

5.04 2.17 

5.25 2.50 

4.93 2.50 

5.33 2.17 

7.07 2.00 

5.37 2.33 

4.89 2.00 

4.52 2.25 

2.17 2.47 5.77 2.00 

3.00 2.80 2.96 3.00 

2.50 2.77 4 .15 2.00 

2.50 2.13 3.39 2.50 

2.67 3.10 3.96 2.33 

2.67 2.93 4.06 2.17 

2.33 3.43 4.13 2.50 

2.17 3.33 4.79 2.00 

2.33 2.97 4.71 2.17 

3.83 8.70 3.43 2.05 2.83 1.98 0.13 

3.50 9.27 3.90 1.89 2.67 2.06 0.15 

4.00 12.35 3.95 2.72 2.75 2.12 0.14 

3.33 8.57 3.50 2.02 2.90 2.03 0.14 

4.50 9.73 4.07 1.95 3.23 1.87 0.17 

4.67 9.83 3.63 2.24 2.70 2.29 0.13 

3.75 10.25 4.05 2.19 2.60 2.38 0.16 

3.25 10.25 4.40 1.96 2.60 2.00 0.19 

1.35 

1.20 

? 

? 

1.23 

1.40 

? 

1.38 

0.51 

0.58 

? 

? 

0.57 

0.65 

? 

0.75 

17 

18 

19 

20 

21 

22 

23 

24 

3.17 6.40 2.57 1.99 2.17 2.17 0.15 ? ? 25 

3.50 5.45 3.35 1.27 2.60 2.08 0.20 1.02 0.68 26 

3.67 6.83 3.80 1.46 2.57 1.84 0.18 1.27 0.71 27 

3.00 5.37 3.27 1.28 2.17 1.91 0.20 1.05 0.65 28 

3.67 6.10 3.23 1.52 2.60 2.11 0.13 1.21 0.67 29 

3.17 6.43 3.47 1.53 2.57 1.93 0.15 1.13 0.66 30 

3.00 7.23 3.37 1.80 2.70 1.80 0.13 1.22 0.69 31 

3.00 7.87 4.90 1.28 2.87 1.88 0.14 1.18 0.71 32 

3.00 7.30 3.80 1.60 2.03 1.98 0.15 1.27 0.65 33 

(OTUs 9- l l), and var. jahandiezii (OTUs 24-25) on the other hand, specimens from C 

Morocco assigned to A. maroccana (OTUs 12-15) forming a transitional cluster between 

the two groups. Variable loadings on PC 1 indicate that this pattern mainly results from 

differences in size of involucres, ray and disc tlorets, and shape of tips of paleso PC 2 

mainly contributes to the dismembering of the W & N Moroccan taxa, resulting in a 

marked polarity between the NW Moroccan A. gharbensis and the SW Moroccan 

A. tenuisecta. A. boveana var. jahandiezii forrns a transitional cluster between the two 

taxa. A. boveana var. aguilarii falls within the range of A. tenuisecta on PC 1 and PC 2, 

but is conspicuously set off from the latter in terms of PC 3, due to differences in achene


Table 21. Variable loadings on PC 1 to PC 3 

of principal component analysis (PCA) of the 

Anthemis boveana group (character symbols 

are explained in Table 19 and Fig. 17). 

Dominating loadings on each principal component 

are shown in bold-face type. 

Character 

PC 1 

PC2 

PC3 

leafi 

ulsl 

ulsi 

invol 

rayno 

oibl 

oibi 

oibs 

mibl 

mi bi 

mibs 

iibl 

iibi 

iibs 

palel 

palei 

pales 

palet 

rayl 

rayw 

rayi 

discl 

disci 

discm 

achei 

achew 

-.003 

.244 

-.294 

.747 

.354 

.649 

- .331 

-.365 

.817 

-.392 

.377 

.840 

- .592 

.476 

.593 

- .555 

.302 

-.654 

.699 

.613 

.383 

.751 

- .464 

.733 

.290 

.338 

.560 

.117 

.146 

.519 

.640 

.494 

.749 

.076 

.388 

.762 

-.242 

.391 

.600 

-.254 

.417 

.661 

-.443 

.378 

.461 

.080 

.666 

.077 

.228 

-.229 

- .498 

- .564 

.418 

-.549 

-.452 

.157 

- .176 

- .184 

.180 

-.243 

- .122 

- .033 

- .356 

.102 

.070 

- .397 

.342 

.315 

- .051 

.058 

- .093 

.088 

-.266 

.402 

.274 

.048 

.573 

.509 

dimensions and length and form of ultimate leaf segments. 

An even clearer picture of the relations between the taxa in the Anthemis boveana 

group emerges when qualitative characters are taken into consideration. A conspicuous 

difference between the Algerian and Moroccan representatives of this group concerns the 

tip of the paleso A. boveana from Algeria has pales with cuspidate tips formed by the 

protruding midrib; plants from the Moroccan populations, even the somewhat intermediate 

A. maroccana, have pales with acute to blunt or even emarginate, translucent, yellow 

tinged, and somewhat hooded tips which are never formed by the midrib of the paleso 

Since the sister taxon of the A. boveana group, A. chrysantha, possesses pales with cuspidate 

tips, there is go od reason to assume the yellow and hooded tips to be the derived 

character state and, hence, the Moroccan taxa to form a monophyletic group.

Bocconea 9 - 1998 83 

A further dismembering of that monophyletic group is achieved by using morphological 

and anatomical characters of the achenes. As discussed earlier (Oberprieler 1994), the NW 

Moroccan populations of the Anthemis boveana group are characterised by disc achenes 

that are unique not only within the species group but among alI N African representatives 

of the genus. The IO acute-angled ribs are smooth, devo id of tubercles, and bear characteristicalIy 

formed mucilage celIs which are strongly elongated and form continuous slime 

ribbons on the narrow ridges (Fig. 3B, 70). Smooth achenes are also found in A. boveana 

varo jahandiezii and in the N populations of A. tenuisecta, but there the mucilage celIs are 

short, isodiametric or but slightly elongate, and the ribs are more rounded and less prominent 

than in the NW Moroccan populations. These were newly described as A. gharbensis 

(Oberprieler 1994). 

A further distinct entity among the Moroccan populations of the Anthemis boveana group 

is A. maroccana (from the C Moroccan highlands around Fès, the adjacent foothills of the 

Middle Atlas, and the Rif mountains) plus A. boveana varo aguilarii (from higher altitudes 

in the Rif mountains). Both taxa tend to have very large achenes (A . maroccana: 

3~--------------------------------------------------------~ 

PC2 

2 

o 

-1 

-2 

21 

",19 

~ .-._. 23 

. ,·'2~ - ~ 

.' l:; 

. 

~/ / 

25 ._ -._ ~ 2:f 


1.5-1.8 mm; A. boveana varo aguilarii: 1.8-1.9 mm), when achenes are less than 1.6 mm . 

long in alI other taxa of the species group as weII as in A. chrysantha. The large achenes 

are therefore a synapomorphic character state of the two taxa which are treated as subspecies 

of A. maroccana here (subsp. maroccana and subsp. aguilarii), differing in achene 

size, corona length, and number of ray florets per head, and being geographicalIy and altitudinalIy 

separated. 

A third cIear cut entity is Anthemis tenuisecta from the coastaI plains around Cap Ghir 

and Agadir in SW Morocco, with some inland populations around Marrakech. As confirmed 

by PCA, this species is characterized by rather small capitula, very short ray florets 

with nearly circular limbs, relatively broad leaves with very low index values, and achenes 

usualIy not exceeding 1.3 mm in length. Additional qualitative characters further emphasise 

the distinctness of this taxon. In contrast to alI other members of the A. boveana group 

and A. chrysantha, the disc achenes of A. tenuisecta usualIy have a very thin pericarp 

lacking scIerenchymatic tissue (Fig. 84). Moreover, A. tenuisecta has a characteristic procumbent 

habit, with stems which branch repeatedly and bear a relatively high number of 

3~----------------------------------------------------. 

PC3 

2 

16 

t;,. -- - 

... 

18 21 i 

_r..--~ .i 

11 

'J'V 

"! 

15 

~ .. 

. ... ~ 

o 

: 0'- 

1~ ./ 

~ 

-1 

PC 1 

-2~ ________ -. ________ -.r-________ .-________ ~ ________ ~ 

-2 -1 o 2 3 

Fig. 19. Anthemis boveana group. Ordination of OTUs (O = A. boveana, = A. maroccana subsp. 

maroccana; " = A. maroccana subsp. aguilarii; b.. = A. gharbensis; t> = A. tenuisecta subsp. 

tenuisecta;

Bocconea 9 - 1998 85 

heads. The leaves are usualIy sessile, i.e. their bases bear entire or dissected teeth which 

grade into primary leaf lobes along the main axis, while the other species usualIy lack 

basai teeth and their leaves are distinctly petiolate. Plants of A. tenuisecta from around 

Cap Ghir, in the N part of the species range, have smooth achenes with very rounded and 

low ribs (Fig. 84, bottom), whereas in the southern populations, around Agadir and Safi, 

and inland around Marrakech, the sculpturing of achenes is more pronounced and even 

strongly tuberculate achenes are found (Fig. 84, top). 

Anthemis boveana varo jahandiezii, which grows in a small area of the coastal plain 

around Safi, is geographicalIy and morphologicàlly transitional between A. tenuisecta and 

A. gharbensis. With the latter it shares the erect, sparsely branched habit and the relatively 

long ray florets, but it lacks the uniquely sculptured achenes of A. gharbensis with their 

extremely elongate mucilage celIs. From A. tenuisecta it differs in habit and in the sclerenchymatic 

tissue of its achenes (Fig. 88), but it shares the rather sessile leaves and the small 

heads of the latter. AdditionalIy, achenes with rounded and rather low ribs as in 

A. boveana var. jahandiezii are also found in populations of A. tenuisecta around Cap 

Ghir. It is mainly the latter feature that suggests inclusion of A. boveana varo jahandiezii in 

A. tenuisecta, as an independent subspecies. 

13. Species delimitation in the Anthemis pedunculata - A. punctata complex 

The Anthemis pedunculata - A. punctata complex comprises alI short- to long-lived 

perennial representatives of the genus in N Africa except A. maritima and A. eretica, 

which also belong to the section A. sect. Hiorthia but hold a relatively remote taxonomic 

position. Members of the complex are widely distributed throughout the mountainous 

areas of Spain, Morocco, Algeria, and Tunisia, and show considerable variation throughout 

that range, which has led to the description of a multitude of taxa of specific or infraspecific 

rank. 

The first to be described was Anthemis punctata, a species with very large flower heads 

from the mountains around Tunis, diagnosed and figured by Vahl (1791). Soon after, Desfontaines 

(1799) described A. pedunculata, with comparatively smaller heads, based on 

plant material presumably collected in W Algeria. Boissier (1838) added A. tuberculata, a 

small-headed plant with tuberculate achenes from the Sierra Nevada in S Spain. Mainly 

based on characters concerning leaf dissection (A. tenuisecta Pomel [non BalI]), colour of 

the membranous margins of involucral bracts (A. clausonis), shape of involucral bracts 

(A. atlantica), and sculpturing of achenes (A. granulata), Pomel (1874-1875) named further 

Algerian taxa of the group at species rank. Battandier (in Battandier & Trabut 1892) 

recognised one more Algerian species similar to A. punctata, A. kabylica, raising in rank a 

variety of A. montana described a few years before (Battandier in Battandier & Trabut 

1888-1890) from the Djurdjura mountains. In Morocco, Humbert & Maire (in Maire 

1927) described A. laeviuscula as a segregate of A. pedunculata, but soon later Maire (in 

J ahandiez & Maire 1934) sunk it to a variety under the latter species. The last two species 

added to the group were the Spanish A. turolensis and A. guardielae, described by Caballero 

(1942a, b).

Bocconea 9 - 1998 85 

heads. The leaves are usually sessile, i.e. their bases bear entire or dissected teeth which 

grade into primary leaf lobes along the main axis, while the other species usually lack 

basaI teeth and their leaves are distinctly petiolate. Plants of A. tenuisecta from around 

Cap Ghir, in the N part of the species range, have smooth achenes with very rounded and 

low ribs (Fig. 84, bottom), whereas in the southern populations, around Agadir and Safi, 

and inland around Marrakech, the sculpturing of achenes is more pronounced and even 

strongly tuberculate achenes are found (Fig. 84, top). 

Anthemis boveana varo jahandiezii, which grows in a small area of the coastal plain 

around Safi, is geographically and morphologicàlly transitional between A. tenuisecta and 

A. gharbensis. With the latter it shares the erect, sparsely branched habit and the relatively 

long ray florets, but it lacks the uniquely sculptured achenes of A. gharbensis with their 

extremely elongate mucilage cells. From A. tenuisecta it differs in habit and in the sclerenchymatic 

tissue of its achenes (Fig. 88), but it shares the rather sessile leaves and the small 

heads of the latter. Additionally, achenes with rounded and rather low ribs as in 

A. boveana var. jahandiezii are also found in populations of A. tenuisecta around Cap 

Ghir. It is mainly the latter feature that suggests inclusion of A. boveana varo jahandiezii in 

A. tenuisecta, as an independent subspecies. 

13. Species delimitation in the Anthemis pedunculata - A. punctata complex 

The Anthemis pedunculata - A. punctata complex comprises all short- to long-lived 

perenni al representatives of the genus in N Africa except A. maritima and A. eretica, 

which also belong to the section A. sect. Hiorthia but hold a relatively remote taxonomic 

position. Members of the complex are widely distributed throughout the mountainous 

areas of Spain, Morocco, Algeria, and Tunisia, and show considerable variation throughout 

that range, which has led to the description of a multitude of taxa of specific or infraspecific 

rank. 

The first to be described was Anthemis punctata, a species with very large flower heads 

from the mountains around Tunis, diagnosed and figured by Vahl (1791). Soon after, Desfontaines 

(1799) described A. pedunculata, with comparatively smaller heads, based on 

plant material presumably collected in W Algeria. Boissier (1838) added A. tuberculata, a 

small-headed plant with tuberculate achenes from the Sierra Nevada in S Spain. Mainly 

based on characters concerning leaf dissection (A. tenuisecta Pomel [non BalI)), colour of 

the membranous margins of involucral bracts (A. clausonis), shape of involucral bracts 

(A. atlantica), and sculpturing of achenes (A. granulata), Pomel (1874-1875) named further 

Algerian taxa of the group at species rank. Battandier (in Battandier & Trabut 1892) 

recognised one more Algerian species similar to A. punctata, A. kabylica, raising in rank a 

variety of A. montana described a few years before (Battandier in Battandier & Trabut 

1888-1890) from the Djurdjura mountains. In Morocco, Humbert & Maire (in Maire 

1927) described A. laeviuscula as a segregate of A. pedunculata, but soon later Maire (in 

J ahandiez & Maire 1934) sunk it to a variety under the latter species. The last two species 

added to the group were the Spanish A. turolensis and A. guardielae, described by Caballero 

(1942a, b).


The species complex around Anthemis pedunculata and A. punctata has never been 

comprehensively revised. Only geographically restricted treatments in Floras are available. 

Battandier (in Battandier & Trabut 1888-1890) treated A. kabylica and A. punctata as 

varieties of the polymorphic and wide spread A. montana, known today as A. eretica, 

while A. pedunculata (including Pomel's illegitimate A. tenuisecta) and A. tuberculata 

(with Pomel's A. atlantica in synonymy) were separated at species level, the latter to differ 

by its strongly tuberculate achenes and its abruptly acuminate paleso A. granulata was 

recognised as a variety of A. tuberculata, differing from the typical variety by its achenes 

having a short adaxial corona. A. clausonis, characterised by its pale membranous margins 

of involucral bracts, was kept as an independent species, its resemblance to 

A. pedunculata being however stated. 

The treatrnent of the Anthemis pedunculata - A. punctata group in Battandier & Trabut 

(1905) differs markedly from Battandier's previous account: A. kabylica and A. punctata 

were considered independent species, while A. montana was left with only one variety 

endemie to the Babors mountains of Algeria, formally described as A. numidica 

(Battandier & Trabut 1888-1890). A. pedunculata was expanded to include two varieties: 

one was based on the Algerian A. clausonis; the other, the annual varo decumbens, had 

been described in the meantime by Bonnet & Barratte (1896) from Tunisia and included 

A. stipa rum, described by Pomel (1874-1875) from Algeria. A. tuberculata, which !ike 

A. pedunculata was said to compri se annual, biennial and perenni al plants, was treated 

again as a independent species with not further subdivision. 

A further deviating assessment ofthe species group was proposed by Battandier (1910), 

who treated Anthemis kabylica as a variety of A. punctata and added a new variety 

(A. punctata varo baborensis). A. clausonis and A. granulata (the latter in contrast to the 

treatment in Battandier & Trabut 1888-1890, where it is placed near A. tuberculata) are 

considered to be independent microspecies ("petites espèces") related to A. pedunculata. 

A. tuberculata, was again is kept separate, but A. stiparum, A. sabulicola, and A. decumbens 

were placed in its vicinity as "types secondaires". 

Maire (in Jahandiez & Maire 1934) provided a taxonomic treatment of the group for 

Morocco, in which Anthemis tuberculata was treated as a second subspecies of 

A. pedunculata, both subspecies being represented by numerous varieties, e.g. A. laeviuscula 

as variety of A. pedunculata subsp. pedunculata and A. granulata as variety of 

A. pedunculata subsp. tuberculata. A. punctata, was said to have two varieties in Morocco: 

varo maroccana in the W High Atlas and varo abylaea in the W Rif mountains. 

Mainly based on Battandier's and Maire's works, the treatment of the Anthemis pedunculata 

- A. punctata complex by Quézel & Santa (1963) contributed !ittle to the understanding 

of its taxonomy: A. kabylica was omitted; A. punctata varo punctata and var. 

baborensis were kept separate from A. pedunculata, which was considered to comprise 

subsp. pedunculata (with a varo clausonis), subsp. tuberculata and subsp. granulata. 

Pottier-Alapetite (1981) adopted Battandier & Trabut's (1905) broad concept of Anthemis 

pedunculata. While A. punctata was kept as an independent species, her 

A. pedunculata comprises not only the usually perenni al taxa varo clausonis and subsp. 

tuberculata but also the annual varo decumbens and subsp. glareosa from S Tunisia. 

The taxonomy of the European representatives of the Anthemis pedunculata - 

A. punctata complex is far less comp!icated than that of its N African members. After the 

description of A. tuberculata and its two varieties varo microcephala and var. discoidea by

Bocconea 9 - 1998 87 

Boissier (1838, 1840) and a presumably erroneous indication of A. punctata by Willkomm 

(1893), no further taxa were noted for the Iberian peninsula until Caballero (1942a, b) 

described two new species, A. turolensis and A. guadielae. Fernandes (1976) sank the 

latter into the sylionymy of A. tuberculata and treated the former as a subspecies of the 

same. Later (Fernandes 1983), she confirmed this concept and further discussed the differences 

between A. tuberculata and A. pedunculata: she considered A. tuberculata to have 

larger achenes with marked and strongly tuberculate ribs, narrow pales which gradually 

taper into dark tips, hemispherical receptacles, and disc tlorets with only inconspicuously 

spongy basai parts; A. pedunculata to have smaller and less conspicuously tuberculate 

achenes, broad pales with abruptly acuminate tips, subconical receptacles, and disc tlorets 

with strongly intlated, spongy basai parts. Since she did not use the type material of 

A. pedunculata for her comparison but a specimen from E Algeria (which I consider to belong 

to A. pedunculata subsp. atlantica), these differences are not however appropriate to 

distinguish the two entities. In contrast to Fernandes (1976, 1983), Benedf i Gonz:Hez 

(1987) considered A. guadielae to be a synonym of A. tuberculata subsp. turolensis, said 

to differ from subsp. tuberculata by broader pales with tips usually not tinged with black 

or brown, smaller leaves, and smaller achenes. He did not further discuss the differences 

between A. tuberculata and A. pedunculata or the variation within the latter species. The 

treatment of A. tuberculata in modern Spanish Floras is variable: Valdés & al. (1987) 

listed it as an independent species, Bolòs & Vigo (1995) gave it the status of a subspecies 

of A. pedunculata. 

As mentioned above, Willkomm (1893) definitively indicated Anthemis punctata, 

which Willkomm & Lange (1865) had listed among their "species inqirendae", for Spain. 

However, the specimen he cited was found by Benedf i Gonzalez (1987) to fall within the 

morphological range of A. tuberculata. The latter author, however, considers one specimen 

from Sevilla province in S Spain to belong to A. punctata. Plants from Sicily with 

rather large tlower heads, similar to plants from Tunisia, were considered to represent 

A. punctata by Gussone (1843) and Fiori & Béguinot (1903), or were treated as an independent 

species, A. cupaniana (e.g. by Nyman 1878-1882, Pignatti 1982). Fernandes 

(l975a, 1976) placed the Sicilian taxon as a subspecies under A. punctata, differing from 

the Tunisian subsp. punctata by its relatively more stout, smooth and indistinctly ribbed 

achenes. 

The different taxonomic concepts brietly summarised above, make it obvious that 

variation within the N African representatives of the group is stili poorly understood, but 

that the group's centre of diversity lies there, so that any conclusion solely or predominantly 

based on plant material from outside Africa is at the very best a crude approximation. 

The present study is based on more than 500 specimens from the N African area of the 

group. In addition, plant material from the Iberian peninsula and from Sicily was studied. 

Besides morphology, cytological aspects were duly considered, since Vogt & Oberprieler 

(1993) had found that polyploidy played a role in the evolution of the group, plants of 

Anthemis pedunculata s.I., at least in Morocco, showing two different ploidy levels (2x, 

4x). As for the A. boveana group, a multivariate approach was made to assess morphological 

variation within the A. pedunculata - A. punctata complex. Further information for the 

interpretation of morphological and cytological findings was drawn from the application of 

the molecular technique of DNA amplification from arbitrary primers (RAPD). It soon


Table 22. List of OTUs used in the numerical (PCA) and/or molecular (RAPO) 

analysis of the Anthemis punctata - A. pedunculata complex. Indication of ploidy 

is reported fram Table 18. 

No. ploidy Location Specimen Analysis 

A. abylaea 

1 4x Ma: Ojebel Fahies. Vogt 12045 (B) PCA+ RAPO 

2 [4x] Ma: Beni-Hosmar. 29 Jun 1930, Maire (MPU) [Type] PCA 

A. pedunculata varo pedunculata 

3 [2x] [N Africa] (G) [Type] PCA 

4 [2x] Hs: Sa. de los Filabres. Rico (SALA 59886) PCA+ RAPO 

5 [4x] Hs: Sierra Nevada. Boissier (G) PCA 

6 2x Ma: Ojebel Tizirene. Vogt 9628 & Oberprieler 4064 (B) PCA+ RAPO 

7 2x Ma: Ojebel Foughal. Vogt 15264 & Oberprieler 9573 (B) PCA+ RAPO 

8 2x Ma: Foum Kheneg. Vogt 9440 & Oberprieler 3878 (B) PCA+ RAPO 

9 2x Ma: Arn Leuh - Azrou. Vogt 9414 & Oberprieler 3852 (B) PCA 

10 4x Ma: Ojebel Tazzeka. Vogt 9487 & Oberprieler 3923 (B) PCA+ RAPO 

11 4x Ma: Ojebel Tassaot. Vogt 9647 & Oberprieler 4083 (B) PCA+ RAPO 

12 4x Ma: Ojebel Foughal. Vogt 15280 & Oberprieler 9589 (B) PCA+ RAPO 

13 4x Ma: Foum Kheneg. Vogt 9440 & Oberprieler 3878 (B) PCA+ RAPO 

14 4x Ma: Ojebel Lekst. Vogt 11829 & Oberprieler 6277 (B) PCA 

15 4x Ma: Tahout-ou-Fillali. Oberprieler 3478 (Herb. Oberprieler) PCA 

16 [?] Ma: Ojebel Tazzeka. 18 Jun 1925, Maire (P) [Type of PCA 

A. laeviscula] 

17 [?] Ma: Ojebel Ouensa. 22 Jun 1875, Ibrahim (MPU) [Type of PCA 

A. punctata varo maroccana] 

18 [2x] Ag: Tiaret. Pomel (P) [Type of A. granulata] PCA 

19 [2x) Ag: Yakouren. Podlech 39294 (MSB) PCA 

20 [2x] Ag: Col Telmet. Podlech 38937 (MSB) PCA+ RAPO 

21 [2x) Ag: Takerbouzt. Podlech 38994 (MSB) PCA+ RAPO 

22 [2x] Ag: Ojebel Babor. Dubuis (MSB) PCA 

23 [2x] Ag: Miliana. Jun 1856, Pomel (MPU) [Type of PCA 

A. tenuisecta] 

24 [?] Ag: Tizi N 'Kouilal. Podlech 39137 (MSB) PCA 

25 [?] Ag: Bossuet 6 Jun 1927, Faure (MPU) [Type of PCA 

A. punctata varo microcephala] 

A. pedunculata varo discoidea 

26 2x Ma: Ojebel Tidirhine. Vogt 9576 & Oberprieler 4012 (B) PCA+ RAPO 

A. pedunculata subsp. atlantica 

27 [?] Ag: Khenchala. Jun 1874, Pome/(P) [Type] PCA 

28 [2x] Ag: Ojebel Babor. Podlech 39380 (MSB) PCA+ RAPO 

29 [?] Ag: Ojebel Tougour. Balansa 967 (MPU) PCA 

30 [4x] Ag: Ojebel Cheliah. 11 Jun 1853, Perraudière (P) PCA 

31 4x Tn: Foret de Kesra. Vogt 13510 & Oberprieler 7815 (B) PCA+ RAPO 

32 2x Tn: Ojebel Chambi. Vogt 12606 & Oberprieler 6911 (B) PCA+ RAPO 

A. pedunculata subsp. turolensis 

33 2x Hs: Beteta - Mazegoza. Vogt 15350 & Oberprieler 9659 (B) PCA+ RAPO


89 


No. ploidy Location 

Specimen 

Analysis 

A. punctata subsp. punctata 

34 [?] Ag: Tamesguida. 

35 

36 

4x 

4x 

Tn: Ojebel Zaghouan. 

Tn: Ojebel Oyr. 

A. punctata subsp. kaby/ica 

37 [?] Ag: Tirourda. 

38 [?] Ag: Tala Guilef. 

A. cupaniana 

39 Si: Palermo, S. Martino. 

A. ubensis 

40 2x Tn: Forét de Kesra. 

41 2x Tn: Ojebel Oyr. 

42 2x Tn: Ojebel Goraa. 

43 2x Tn: Table de Jugurtha. 

A. confusa Pomel 

44 2x Tn: Tataouine - Chenini. 

45 2x Tn: Redeyef - Tamerza. 

46 2x Tn: Jelma. 

A. stiparum subsp. stiparum 

47 2x Ag: Aflou - El Bayadh. 

A. stiparum subsp. sabu/icola 

48 2x Ag: El Golea - Ghardaia. 

A. zaianica 

49 2x Ma: Oj. Tougroulmès. 

Battandier (P) [Type of A. punctata varo 

baborensis] 



Jun 1882, Battandier (P) [Type] 

Davies 53190 (RNG) 

Todaro 1102 (P) [Type] 











PCA 

PCA+ RAPO 

PCA+ RAPO 

PCA 

PCA 

PCA 

PCA+ RAPO 

PCA+ RAPO 

PCA 

PCA 

RAPO 

RAPO 

RAPO 

RAPO 

RAPO 

RAPO 

became obvious that there are not only problems of taxon delimitation within the 

A. pedunculata - A. punctata complex but also with the delimitation of this complex 

against other morphologically and/or genetically related species and species groups. In 

particular, the role of the Tunisian annuals (A . ubensis and A. confusa of the present study) 

often treated as A. pedunculata var. decumbens was difficult to interpret and these taxa 

were therefore inc1uded in our morphological and/or molecular analysis, together with a 

Moroccan population of annuals (A. zaianica) found to be morphologically intermediate 

between A. pedunculata s.1. and the annuals of A. ser. Bourgaeinianae, and with another 

species of that series, A. stipa rum. 


Numerical analysis oJ morphological characters. - 43 specimens (OTUs) representative 

of the entire morphological and geographical range of the Anthemis pedunculata - 

A. punctata complex, plus the N Tunisian A. ubensis, were measured or scored for 26 

morphological characters (Table 19, Fig. 17). The OTUs (Table 22) were either single 

herbarium specimens or sets of duplicates from the same gathering. Type material of most 

of the relevant names was included in this study. On each specimen 3-5 measurements


were made for each character, the results averaged, and the mean values used in the multivariate 

analysis (Table 23). Principal component analysis (PCA) was performed using the 

SPSS FOR WINDOWS software package. Missing values (marked by a question mark in 

Table 23), were replaced by the mean value of ali other OTUs for the corresponding character. 

Plant material for RAPD analysis. - 26 OTUs were selected for use in the RAPD 

analysis (see Table 22). 20 were representatives of the Anthemis pedunculata - 

A. punctata complex, plus A. ubensis, also used in the morphological analysis. To these, 

specimens of A. confusa, A. stiparum, and A. zaianica (OTUs 44-49) were added to address 

the questions of the relations of these species with the A. pedunculata - A. punctata 

complex. Single plants of the selected specimens were used as source for DNA extraction. 

DNA extraction. - DNA was extracted from dried and crushed leaf material accarding 

to a method devised by Hellwig & al. (unpubl.). 30-40 mg dry plant tissue was used. Lysis 

was performed in 0.8 mi 2 % CTAB-buffer at 60°C for 20 min. A first centrifugation for 

lO min at 15800 g at room temperature cleaned the supernatant from celi debris. The following 

steps use buffers of Quiagen Inc. 1 VoI. QA-ns buffer (double concentrated 

Quiagen QBT buffer without NaCl) was added to the supernatant. A second centrifugation 

(lO min at 15800 g at room temperature) followed. The supernatant was then loaded onto 

a Quiagen tip 20 column equilibrated previously with buffer QBT and washed 4 times with 

l m1 QC buffer. DNA was eluted with 1115 fll QF buffer and precipitated by addition of 

0.7 VoI. (805 fll) isopropanol. After mixing, the precipitate was sedimented during 30 min 

centrifugation at 15800 g at room temperature. Pellets were washed with 0.4 mi ice co Id 

ethanol and sedimented again by a lO min centrifugation at 15800 g. The supernatant was 

discarded and the pellets were air-dried. The not too dry pellets were then resuspended in 

50 fll TE buffer pH 8.0. 

RAPD amplification. - Polymerase chain reactions (PCR) were perfarmed in 50 111 of 

the buffer supplied by the Taq polymerase manufacturer (Appligene), 0.2 mM dATP, 0.2 

mM dGTP, 0.2 mM dCTP, 0.2 mM TIP, 0.25 flM primer, l fll template DNA (c. 2-8 

ng/fll) and 2.5 units of Taq polymerase (Applygene). Primers AOI to A12 of Primer Kit A 

from OperonTechnologies Inc. (U.S.A.) were used in the 12 independent PCR reactions; 

prinier sequences are given in Table 24. Amplification was carried out in a thermocycler 

programmed far 1 min denaturation at 95°C, followed by 35 cycles of 1 min at 93°C, 30 

sec annealing at 35°C, 1 min extension at 72°C, and a last additional 2.5 min extension at 

n°c. Products of amplification were separated on 3 % TAE-agarose gels by electrophoresis, 

stained with ethidium bromide, and documented with a Polaroid camera. Lengths of 

amplification products were calculated using ÀlHind III and pBRlAlu I markers. Since a 

well-known problem with RAPD analyses is the difficulty to reproduce RAPD patterns 

exactly (Oxelman 1996), ali DNA templates far one primer were run with the same master 

mix of reaction components in the same terrnocycler at the same time. 

Data analysis. - Following recommendations by Koch (1995), only amplification 

products having lengths between 300 bp and 1500 bp were scored and used for the calculations. 

Shorter or longer amplification products are often of artificial nature, and their

Bocconea 9 - 1998 91 

amplification is highly random. Co-migrating bands shared by at least two of the OTUs 

were considered informative and used to build up the present/absent data matrix given in 

Table 25 . Question marks denote cases of faint band patterns due to failed DNA amplification. 

From the data matrix a distance matrix (DAB = l - SAB) was computed using Jaccard's 

association index (SAB = N AB / (NA + NB); where NAB = number of joint amplification 

bands; NA = number of bands in OTU A; NB = number of bands in OTU B) as the 

measure of similarity (Sneath & Sokal 1973: 131). For this operation the computer package 

SYN-T AX III (Podani 1988) was used. Dendrogram construction was performed by 

using the neighbor-joining method (Saitou & Nei 1987) provided in the computer package 

PHYLIP Version 3.5c (Felsenstein 1994). Since this method produces an unrooted tree, the 

tree was rooted using the midpoint-rooting option (program RETREE of PHYLIP) which 

places the root of the tree to the longest interior branch. For bootstrap analysis, 100 replicates 

of the originai data set were obtained running the bootstrap option of PHYLIP 

(program SEQBOOT). For each replicate a distance matrix was computed with SYN-TAX 

III, and a dendrogram constructed as above. The resulting 100 dendrograms were summarised 

in a majority-rule consensus tree with PHYLIP (program CONSENSE) and midpointrooted 

(program RETREE). 

Numerical analysis of morphological characters 

Principal component analysis (PCA) of morphological data yielded seven components 

with eigenvalues higher than 1, accounting for 78.8 % of the total vari ance of the data set. 

Character loadings on the frrst three components of PCA are listed in Table 26. The first 

principal component (PC l; 30.9 % of the total variance) is dominated by variables describing 

dimensions of leaves (leafl, ulsl), involucres (invol, oibl, mibl, iibl), pales (palel), 

ray and disc florets (rayl, discl), and the number of ray florets per capitulum (rayno). PC 2 

(16.0 % of the total vari ance) falls back markedly behind PC 1. The only characters with 

high loadings are those describing the shape of inner involucral bracts (iibs), the shape of 

pales (palei, pales), and the shape of the tips of pales (palet). PC 3, accounting for 10.4 % 

of the total variance, is dominated by variables describing the shape of the outer involucral 

bracts (oibi, oibs) and length of achenes (achei). 

As seen on plots of the frrst three principal components of the PCA (Fig. 20-21), PC I, 

which is dominated by size features, separates fairly well Anthemis pedunculata and 

A. ubensis on the left from A. punctata, A. cupaniana, A. punctata subsp. kabylica, and 

A. abylaea, characterized by long basai leaves, large involucres, long pales, and long ray 

and disc florets, and placed further right. However, the differences are not too clear cut; 

the type specimen of A. punctata subsp. kabylica (OTU 37), in particular, is very close to 

representatives of A. pedunculata s.1. 

The two subgroups resulting from ordination on PC 1 are further subdivided on PC 2 

(Fig. 20). Within the Anthemis punctata group of taxa we find a clear discontinuity between 

A. punctata subsp. punctata (OTUs 34-36) and A. cupaniana (OTU 39) on the one 

hand, and A. punctata subsp. kabylica (OTUs 37-38) and A. abylaea (Otus 1-2) on the 

other. This is mainly due to the shape of the inner involucral bracts and pales which are 

found to be elliptical in outline in A. punctata subsp. kabylica and A. abylaea but obovate 

in A. punctata subsp. punctata and A. cupaniana. The type of A. punctata varo baborensis 

(OTU 34) is definitively closer to A. punctata subsp. punctata than to subsp. kabylica.


Table 23. Data far the multivariate analysis af the Anthemis punctata - A. pedunculata 

camplex (character symbals are explained in Table 19 and Fig. 17). 

OTU leafl leafi ulsl ulsi inval rayna aibl aibi aibs mibl mibi mibs iibl 

1 49.00 2.04 2.88 3.21 16.83 18.00 4.57 2.77 1.17 6.33 3.77 1.67 6.77 

2 51.50 3.32 3.14 2.97 13.80 22.00 4.00 2.98 1.00 5.35 3.37 1.25 6.90 

3 33.33 3.36 2.36 3.13 9.80 16.50 2.75 2.82 1.00 3.85 4.17 1.75 4.10 

4 18.50 3.38 1.60 3.18 8.20 10.67 2.13 2.13 1.00 3.03 2.85 1.173.03 

5 25.33 3.62 1.74 2.98 10.33 ? 2.47 1.85 1.00 3.53 2.61 1.67 4.23 

6 28.25 2.20 2.14 2.40 10.33 12.33 3.07 2.87 1.33 4.00 3.24 1.67 3.87 

7 45.75 2.36 2.20 2.85 12.40 12.60 3.97 2.67 1.00 4.97 3.24 1.50 5.23 

8 15.00 2.50 1.40 2.38 8.67 10.33 2.70 2.38 1.00 3.37 3.16 1.67 3.63 

9 23.40 3.27 1.63 2.59 

10 29.25 2.43 1.96 2.47 

11 35.40 3.01 2.34 2.61 

12 37.67 3.73 2.26 2.76 

13 36.75 2.86 2.90 2.69 

14 27.00 3.25 1.83 2.28 

15 38.00 2.93 2.22 2.22 

16 35.67 3.39 2.14 2.58 

17 40.50 4.05 1.78 2.12 

18 24.00 3.52 1.40 2.17 

19 43.60 2.77 2.07 3.51 

20 33.67 3.16 2.36 2.56 

21 22.00 3.00 1.68 2.42 

22 27.00 3.00 2.13 2.77 

23 48.67 2.98 2.93 5.56 

24 31 .17 2.99 2.22 2.41 

25 27.00 3.00 1.70 2.04 

26 22.75 2.14 2.20 2.52 

27 20.50 3.12 1.64 2.32 

28 30.00 3.33 1.42 2.48 

29 25.50 2.81 2.05 2.67 

30 26.50 4.10 1.68 2.52 

31 43.80 2.83 2.42 2.83 

32 29.80 3.74 1.77 2.74 

33 20.83 3.40 1.28 2.49 

34 37.00 2.83 4.22 3.87 

35 51 .80 3.16 2.76 2.97 

36 46.50 3.61 2.98 2.65 

37 28.60 2.42 1.94 2.12 

38 59.75 2.91 3.16 2.48 

39 65.00 2.29 4.20 3.46 

40 28.00 2.65 1.53 2.44 

41 28.83 3.57 1.50 2.54 

42 21.00 2.78 1.57 2.97 

43 26.33 3.10 1.32 2.17 

10.00 13.00 2.90 2.95 1.00 3.37 3.26 1.50 3.73 

10.33 13.67 3.63 3.35 1.00 4.70 3.70 1.50 4.63 

10.67 11 .33 2.83 1.78 1.00 4.20 2.70 1.83 4.47 

11.67 13.60 3.57 3.06 1.00 4.40 2.82 1.33 5.03 

9.33 13.50 3.33 2.96 1.00 3.97 3.25 1.50 4.63 

11 .67 13.17 3.23 2.86 1.00 3.60 3.18 1.67 4.33 

10.83 12.25 2.53 2.85 1.00 3.00 3.34 1.50 3.83 

9.50 13.00 2.80 2.49 1.00 3.38 2.81 1.50 4.15 

12.83 16.67 3.30 2.76 1.00 3.95 2.63 1.00 5.15 

10.40 13.50 2.80 2.91 1.00 4.10 3.05 1.75 3.98 

11 .50 13.25 2.83 2.87 1.50 3.13 2.50 2.17 3.43 

10.60 12.67 3.53 3.12 1.00 4.37 3.22 1.67 5.43 

10.00 15.80 3.07 2.84 1.50 3.67 3.54 2.00 3.70 

10.83 14.80 3.13 2.55 1.00 4.07 3.00 1.67 4.43 

11 .17 ? 2.53 2.73 1.00 3.00 2.80 1.50 4.10 

11 .33 18.20 3.97 3.01 1.50 4.73 3.31 2.00 5.17 

9.75 ? 3.00 2.40 1.00 4.05 2.82 1.75 3.70 

8.00 0.00 2.03 1.76 1.00 3.07 2.31 1.33 3.10 

10.00 ? 2.65 2.52 1.00 3.35 2.35 1.50 3.45 

9.80 14.00 2.57 1.93 1.17 3.50 2.04 2.33 4.03 

10.83 13.00 3.07 2.48 1.00 4.07 2.64 2.00 4.63 

10.67 12.00 3.10 2.50 1.00 4.43 3.70 1.83 4.40 

13.17 16.33 3.90 2.93 1.17 4.70 2.93 1.83 5.40 

12.50 14.50 3.93 2.93 1.00 4.70 2.63 1.33 4.87 

9.63 14.803.132.551.173.732.741.833.83 

11 .80 13.00 3.75 2.42 1.00 4.25 2.43 1.25 5.05 

12.83 14.00 3.47 2.68 1.00 5.00 2.94 1.83 5.60 

14.83 17.17 3.70 2.12 1.00 4.87 2.72 1.67 6.60 

11 .00 23.00 3.30 2.81 1.00 4.28 2.82 1.67 5.47 

17.20 19.33 4.95 3.19 1.25 5.60 3.61 1.75 6.45 

15.33 21 .00 3.60 2.40 1.00 4.70 2.61 1.50 5.30 

11 .50 14.83 3.90 3.39 1.17 5.00 3.31 2.00 5.93 

10.83 13.17 3.70 2.68 1.17 4.37 2.70 1.83 5.23 

11 .00 10.50 3.40 2.60 1.00 4.33 2.92 1.83 4.30 

11 .0013.173.272.91 1.174.102.372.174.68

Bocconea 9 - 1998 93 


iibi iibs palel palei pales palet rayl rayi discl disci achei achew coro n OTU 

4.06 2.00 4.82 5.97 2.00 2.00 17.47 2.76 3.73 2.25 2.68 1.27 005 

4.20 2.50 5.23 7.60 2.00 2.00 15.65 2.69 3.23 2.26 2.16 1.23 0.00 

3.92 2.25 2.76 5.06 2.00 2.00 9.70 2.86 2.60 2.36 1.60 0.70 0.15 

2.97 2.00 2.77 5.29 2.00 2.00 9.17 2.14 2.63 2.34 1.67 1.03 0.20 

2.96 2.17 3.28 5.22 2.00 1.85 7.15 1.54 2.60 2.60 2.19 1.15 0.30 

3.07 2.00 3.44 6.50 2.00 2.00 7.50 2.06 2.78 2.32 1.95 0.98 0.05 

3.83 1.83 4.06 6.70 2.00 2.00 14.53 2.54 3.07 2.31 ? ? ? 

3.48 2.33 2.86 6.65 2.00 2.00 8.00 1.67 2.25 2.40 1.83 0.98 0.23 

3.68 1.67 2.52 7.70 2.00 2.00 9.27 2.39 2.63 2.29 1.73 0.87 0.06 

3.78 2.00 3.78 6.88 2.00 2.00 9.60 2.05 3.00 2.37 2.42 1.16 0.15 

3.15 2.17 3.70 6.01 2.00 2.00 10.03 1.91 3.30 2.36 1.83 0.92 0.10 

3.44 2.00 3.68 5.05 2.00 2.00 10.73 2.04 2.88 2.76 1.93 0.97 0.08 

3.67 2.00 3.76 6.17 2.00 2.00 13.43 2.78 2.93 2.39 1.85 0.93 0.14 

3.64 2.00 3.58 7.90 2.00 2.00 8.88 1 .88 3.05 2.90 2.22 1.08 0.24 

4.26 1.83 3.06 5.15 2.00 1.70 9.50 2.32 2.90 2.18 1.90 0.99 0.21 

3.81 1.75 4.22 7.56 2.00 2.00 8.30 1.64 2.93 2.61 2.23 1.10 0.23 

4.17 2.00 3.80 7.71 2.00 2.00 16.10 2.91 2.97 2.37 2.20 1.07 0.15 

2.37 2.50 3.15 4.79 2.25 2.00 11 .90 2.06 2.83 2.90 1.80 1.00 0.30 

2.97 2.17 3.22 5.11 2.00 1.80 11 .50 2.63 3.25 2.65 1.62 0.93 0.02 

3.47 2.67 4.22 4.97 2.00 1.90 13.80 2.20 3.22 2.49 1.73 0.93 0.00 

3.02 2.67 3.25 6.39 2.00 2.00 11 .20 2.50 3.40 2.54 1.77 0.92 0.08 

3.05 2.83 3.46 5.12 2.50 1.50 12.83 2.76 2.90 2.46 1.57 0.92 0.00 

3.29 2.25 3.73 6.60 2.10 2.00 10.80 3.00 2.93 2.58 2.05 0.94 0.21 

4.04 2.00 3.90 7.71 2.00 2.00 12.13 2.68 2.93 2.75 1.53 0.83 0.16 

2.31 3.00 2.98 4.11 2.30 1.60 8.43 2.21 2.63 2.26 1.61 0.96 0.27 

2.38 2.00 2.76 4.92 1.80 2.00 ? ? 2.83 2.52 2.18 1.00 0.28 

2.11 2.50 3.90 4.30 2.40 1.30 12.30 2.62 3.00 2.50 1.75 0.75 0.06 

1.87 3.00 3.34 3.45 2.70 1.00 9.60 2.32 2.63 2.24 1.79 0.80 0.00 

2.90 3.00 4.08 4.68 2.30 1.70 10.95 2.26 2.98 2.90 2.00 1.03 0.20 

3.27 2.67 3.28 4.55 2.50 1.50 8.55 2.40 2.83 2.23 2.08 0.94 0.11 

2.66 2.83 4.36 3.92 2.20 1.60 13.95 2.58 3.35 2.39 1.96 0.99 0.10 

2.61 2.50 3.63 3.37 3.00 1.13 10.76 2.05 3.17 2.19 1.98 0.85 0.38 

2.81 2.00 2.86 5.16 2.00 1.90 9.60 2.18 2.73 2.46 ? ? ? 

2.28 2.75 4.21 5.53 2.10 1.50 14.00 2.55 3.17 3.17 1.85 0.95 0.10 

2.62 2.67 4.92 5.15 2.00 1.40 12.27 2.77 3.35 2.95 2.20 1.08 0.73 

2.96 2.67 4.32 4.35 2.50 1.10 13.93 2.61 3.45 2.43 2.25 0.94 0.35 

3.87 2.00 4.88 9.56 2.00 2.00 11.40 2.91 3.23 2.55 2.15 0.95 0.30 

4.78 2.25 4.90 8.17 2.00 2.00 19.60 2.47 3.80 2.86 ? ? 0.50 

2.65 2.00 4.70 5.11 2.38 1.13 13.05 2.13 3.27 2.55 2.03 1.02 0.47 

3.62 2.33 4.20 3.95 2.10 1.20 8.77 2.26 2.95 2.41 1.55 0.76 0.02 

2.91 2.17 4.22 3.29 2.70 1.10 10.20 2.17 3.15 2.49 1.39 0.77 0.04 

2.69 2.33 4.03 3.09 2.75 1.13 8.80 1.82 3.03 2.12 1.90 0.90 0.82 

2.25 3.00 3.38 3.21 2.90 1.00 7.80 2.10 2.75 2.48 ? ? ? 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43


On the left side of the plot (Fig. 20), taxa belonging to Anthemis pedunculata s.l. are 

also scattered along PC 2 and form a transitional series, ranging from plants with extremely 

low scores (indicating narrow inner involucral bracts and narrow pales that gradually 

taper to the tip) to those with very high scores (indicating broad inner involucral 

bracts and broad pales with tricuspidate or abruptly acuminate tips). This morphological 

variation shows a clear correlation with geographical origin: Plants from the W part of the 

distributional area of the group (Spain alld Morocco) are placed at the negative end of PC 

2, while the positive end is formed by populations of A. ubensis (OTUs 40-43) from Tunisia 

and populations from E Algeria ascribed to A. pedunculata subsp. atlantica (OTUs 27- 

32), the extremes being connected by a cluster of specimens from W and C Algeria (OTUs 

18-25) that hold a intermediate position and include the types of A. granulata (OTU 18) 

and A. punctata varo microcephala (OTU 25). The types of A. tenuisecta Pomel, nom. 

illeg. (OTU 23), and A. pedunculata (OTU 3) are nested within the Moroccan and Spanish 

representatives, along with A. tuberculata varo discoidea (OTU 26), A. turolensis (OTU 

33), A. laeviuscula (OTU 16), and A. punctata varo maroccana (OTU 17). 

PC 3 (Fig. 21) contributes little to a further understanding of taxon relations within the 

Anthemis punctata and the A. pedunculata groups. The geographical cline within 

A. pedunculata, as revealed by PC 2, is lost in PC 3. 

For the Moroccan representatives of Anthemis pedunculata s.l. there seems to exist a 

week correlation between morphology and chromosome number: diploid representatives 

(OTUs 6, 8, 9) are mainly found to have rather low scores on PC 1, the tetraploids (OTUs 

10-15) tend to have higher scores on that axis, mainly due to size differences between the 

two cytotypes. However, the position of OTU 7, a diploid specimen from NE Morocco, 

shows that exceptions to this pattem occur. 

Table 24. Base sequences of primers used in the RAPD analysis of the 

Anthemis pedunculata - A. punctata complex. 

Primer Base sequence Primer Base sequence 

A-Ol 5 ' -CAGGCCCTTC-3 , A-07 S ' -GAAACGGGTG-3 ' 

A-02 S'-TGCCGAGCTG-3' A-08 S ' -GTGACGTAGG-3 ' 

A-03 S '-AGTCAGCCAC-3' A-09 5 ' -GGGTAACGCC-3 , 

A-04 S ' -AATCGGGCTG-3 ' A-IO 5 ' -GTGATCGCAG-3 , 

A-05 S ' -AGGGGTCTTG-3 ' A-Il S ' -CAATCGCCGT-3 ' 

A-06 S ' -GGTCCCTGAC-3 ' A-12 S ' -TCGGCGATAG-3 ' 

RAPD analysis 

Polymerase chain reactions (PCR) using 26 OTUs and 12 different primers resulted in 

106 informative amplification bands which are listed in Table 25. The number of informative 

bands per primer ranged from 1 (primer A-06) to 18 (primer A-Il). The dendrogram 

constructed using the Jaccard index as a distance measure and the neighbor-joining 

method for the agglomeration of OTUs resulted in the dendrogram shown in Fig. 22, A 

majority-rule consensus tree based on 100 replicates is presented in Fig. 23. Bootstrap 

values appear to be rather low throughout the consensus tree, indicating that the data set 

contains a considerable amount of random noise, The same conclusion is valid for the

Bocconea 9 - 1998 95 

Table 25: Distribution of variable amplification bands found in the RAPD analysis 

of representatives of the Anthemis pedunculata - A. punctata complex. 

Primer fragmenl OTU Primer fragment 

length 

length 

OTU 

(bp) 30014201010144443333444222 (bp) 30014201010144443333444222 

34109661728345601026187018 34109661728345601026187018 

A-OI 1425 01000000110010011110000100 A-08 954 11111111111107711101077011 

1360 00000000000011101000000100 

908 00001000000017700000077000 

1149 01000000000010000010000000 

814 00000000000017710111077000 

1101 00000000100000000010000000 

713 00000001000007700000077001 

1055 01010001000000000000000000 

642 11111011111107700000177110 

1012 00001000000000000000100001 

619 01000000110007700100077101 

895 000010000000000QOOO0100000 

577 00001001000007700100077011 

844 00000000110000000000000000 

750 11110111111100001111011111 467 10000011000007700100077010 

669 00000000000000000001000001 405 00100000000007700000177000 

644 00000000000000000001100000 A-09 854 01010101111100701001101010 

632 00000000000000001010000000 635 00000011000001700100000001 

482 10111000000000010111100111 614 11111101110100711111100111 

435 00100010000100011011000000 

545 11110100111110711110010100 

A-02 1187 00100001111110011111107101 

509 01000100000000700000000000 

1005 01001110111100000000000000 

436 10101111000101701111011000 

985 00100000101100000111100010 

779 00100001010000001111110011 A-IO 1056 01000100000001000000000000 

765 00100001010000001111110011 1021 00000000010000000000000100 

751 00100001010010001111110011 957 10111011001100000100000000 

737 00100001010000001111111111 870 00000100000001000110000001 

601 00100000000011111101111000 826 11000001000000011000000000 

579 00001000000001000000010000 784 11000001000000011000001111 

558 00000000001000000000100000 668 11111111111111011111111111 

533 00000000000001000000010000 612 11110101010110011111011111 

509 10100101011110011011110111 

499 00000000000010000010001000 

482 11110111111111111111110111 

485 00110100000000000000000111 

432 00000100000000000000000001 

355 00100001011100011111110011 A-Il 1117 00100000111100010000000000 

338 00100001111100011111111111 1096 00000000001000000100000110 

A-03 645 01001000001000000000011001 1034 00100000010000000000000000 

565 00010000000011000000001000 1015 01000000000100010000000000 

451 11110111111711100100011011 959 10001000000000000000000011 

428 01110111101710000100111111 860 00000001000000000000011011 

A-04 1144 11110101111111011111107111 747 00010010010000010100000000 

1073 11110101111111011111107111 698 00001000000001000011010001 

1040 00001000000000100000010000 686 11100101010100100000000100 

761 00000000101000010001007111 652 00010110101000011011100010 

592 00000000000010000100010000 620 00000100100000000010001100 

470 00100001000000001000000100 590 00000010100000000100000100 

443 11111111111100000000000000 585 00000000000000000000000011 

426 11111111111100000000000101 

562 01000111000000000011010000 

410 11100000010000000000010000 

544 00001000100000000000100000 

A-05 1127 00000000000000000000010001 

501 00000010000110000100010000 

1093 00000000010000000100000000 

379 00000000001111000001010000 

1038 00000001010000010000017711 

316 00000000000000010001111000 

997 00110000000000000000000000 

667 10000011010000011111117111 A-12 1218 00000000000000001010000000 

631 10010010010000017111117111 1067 10010000000000000000000000 

517 00000000000000010000001111 655 00000000010000000000000011 

A-06 1034 00001000000000011111000000 612 00000000000010010010100000 

A-07 881 00000000000010011111100111 572 00011000000010000000000111 

826 00000000000000000011100000 561 10100101100000001111100111 

696 00000000000000011010000000 427 00100100000000000000000000 

671 00000000000100011110000000 321 00111011010100001100000000


Table 26. Variable loadings on PC 1 to PC 3 

of principal component analysis (PCA) of the 

Anthemis punctata - Anthemis pedunculata 

complex (character symbols are explained in 

Table 19 and Fig. 17). Dominating loadings 

on each principal component are shown in 

bold-face type. 

Variable PC 1 PC2 PC3 

leafl .815 .090 -.234 

leafi -.214 .172 .028 

ulsl .689 .005 - .408 

ulsi .201 -.003 - .403 

invol .886 .276 -.068 

rayno .667 .129 .265 

oibl .831 .264 .277 

oibi .427 - .118 .612 

oibs .085 .081 .609 

mibl .802 .301 .157 

mibi .367 - .430 .501 

mibs - .181 .469 .483 

iibl .876 .264 .058 

iibi .539 -.626 .331 

iibs - .105 .726 .024 

palel .835 .267 - .127 

pale i .406 -.755 .069 

pales -.207 .841 - .011 

palet .122 -.910 .076 

rayl .807 .047 -.045 

rayi .452 .006 .173 

discl .804 .234 -.022 

disci .190 -.013 - .185 

achei .433 - .229 -.533 

achew .385 -.374 - .476 

coron .125 .192 - .430 

dendrogram in Fig. 22, where most of the interior branches connecting nodes are considerably 

shorter than the branches between nodes and OTUs, indicating that a high amount 

of parallelism exists in the data set and that only few amplification bands are significant 

for the tree topology. However, there is a good correlation of groupings in dendrogram 

and consensus tree, and only few discrepancies are found, which indicate that the essential 

structures of the dendrogram are trustworthy despite the low bootstrap values. 

The arrangement of OTUs according to their genetic similarity as assessed by RAPD 

analysis shows some interesting groupings. Four major c\usters were found to exist in both 

the originaI dendrogram and the consensus tree of the bootstrap analysis. The first group is 

formed by the annuals of Anthemis confusa (OTUs 44-46), A. stipa rum subsp. stipa rum

Bocconea 9 - 1998 97 

(OTU 47) and subsp. sabulicola (OTU 48). This cluster is conspicuously set apart from 

the rest of the OTUs, which cluster with the comparatively high bootstrap value of 61 %. 

A second cluster is formed by Tunisian representatives: the annual, diploid A. ubensis 

(OTUs 40, 41), the perennial, tetraploid A. punctata (OTUs 35, 36), and a diploid (OTU 

32) and tetraploid (OTU 31) of A. pedunculata subsp. atlantica. The arrangement of 

OTUs within this cluster differs markedly between the dendrogram and the consensus tree, 

so that groupings within this cluster are very unstable. Only the position of OTU 32 as an 

outlayer of this cluster proves to be fairly reproducible. 

A third cluster comprises the diploid and tetraploid Moroccan and Spanish representati 

ves of Anthemis pedunculata s.l (OTUs 4-l3) and, nested among them, A. abylaea (OTU 

l) and A. zaianica (OTU 49). The internaI structure of the cluster is faithfully maintained 

in the consensus tree. A first subcluster is formed by the tetraploid A. abylaea (OTU l) 

and two tetraploid Moroccan representatives of A. pedunculata (OTUs 11-l2), while the 

other tetraploids are nested within the cluster comprising Moroccan diploids (OTUs 6-8), 

along with representatives of A. turolensis (OTU 33) and A. tuberculata (OTU 4) from 

Spain, and A. tuberculata varo discoidea (OTU 26) and A. zaianica (OTU 49). 

The fourth cluster is formed by three Algerian representatives of Anthemis pedunculata 

s.1. (OTUs 20, 21, 28). It is linked with the Tunisian cluster in Fig. 22, but in the consensus 

tree (Fig. 23) it is associated with the cluster of Moroccan OTUs. In view of the very 

low bootstrap value for the latter grouping (17 %), the position of this cluster is uncertain. 

Discussion 

The morphological analysis shows a main discontinuity within the data set that is 

caused by correlated characters describing sizes of leaves, involucres, ray and disc florets. 

That combination of characters permits to discriminate between Anthemis punctata s.1. on 

the one hand and A. pedunculata s.1. on the other hand. The A. punctata group, thus defined, 

comprises taxa that were ali treated at infraspecific level within that species in the 

past: A. punctata var. abylaea (OTUs 1-2), A. punctata varo kabylica (OTUs 37-38), 

A. punctata var. baborensis (OTU 34), and A. punctata subsp. cupaniana (OTU 39). 

Two taxa assigned to Anthemis punctata in the past, however, are found nested within 

A. pedunculata s.l., the first of these is A. punctata varo microcephala, for which Faure & 

Maire (in Maire 1931 b) stated that it has the habit of A. tuberculata, differing in its adaxially 

coronated achenes, a character which they considered typical of A. puncata. Maire 

(1933) later included A. punctata var. microcephala in the synonymy of A. granulata, 

which he treated as a variety of A. pedunculata subsp. tuberculata in Jahandiez & Maire 

(1934). The inclusion of A. punctata var. microcephala in A. pedunculata s.1. is backed by 

the morphological analysis presented here, its type (OTU 25) being very similar to that of 

A. granulata (OTU 18). Both types, collected in W Algeria, belong to the group of OTUs 

intermediate between the Moroccan/Spanish representatives of A. pedunculata s.1. and the 

E Algerian ones. 

Anthemis punctata varo maroccana (OTU 17) is the other taxon once assigned to 

A. punctata bul showing a closer overall similarity with representatives of A. pedunculata 

s.1. in the peA. It was said by Maire (l93la) to differ from the typical variety of 

A. punctata by its smooth achenes lacking a corona, from A. punctata var. baborensis by 

its pubescent leaves and conspicuously mucronate ultimate leaf segments, and from


A. tuberculata by its twice as large capitula. However, Maire (l931b) adds: "Habitus 

A. tuberculatae Boiss." Indeed, specimens from the W High Atlas mountains ascribable to 

A. punctata varo maroccana tend to combine characters typical for A. pedunculata s.1. (e.g. 

smallleaves, ecoronate achenes) with large flower heads typical of A. punctata s.1. However, 

capitulum size still falls within the variation range of A. pedunculata s.I., its 

tetraploid but also some diploid (OTU 7!) representatives in Morocco. Therefore, 

A. punctata varo maroccana is better considered a member of A. pedunculata s.1. whose 

rather large capitula were acquired independently from those of the E Algerian and N 

Tunisian A. punctata. Pending further cytological and genetic studies of these populations 

in the W High Atlas mountains, they are here treated as belonging to the tetraploid representatives 

of A. pedunculata. 

The N Moroccan populations described as A. punctata varo abylaea, however, appear to 

be well separated from A. pedunculata s.1. in our morphological analysis. The two studied 

2 ,5~--------------------------------------------------~ 

2,0 

1,5 

1 ,O 

PC2 

28 

1( .- _. 

\. 

\. 

43 

~ " 

- .- \:' 32 

......... ~; - . -...... ,. 

\.42 ...... 

'". .... 

'., 

" . 

'-. 

36 

39 

• 

,5 

0,0 

-,5 

-1,0 

-- -- 18 

\ o 

19 

o 

-- \. 1~1 

" ". 20 

"~ 

5~ _ -\-----~--,- - -" ,t2 '~ 

r---- \. 2f'.>,

Bocconea 9 - 1998 99 

representatives of this taxon (OTUs 1-2) are quite similar to representatives of A. punctata 

varo kabylica (OTUs 36-37) from the Djurdjura mountains in C Algeria. That resemblance 

even induced Font Quer & Pau to identify a collection of Font Quer from the W Rif 

mountains (Iter Maroccanum 1930: n° 665) as A. cupaniana varo kabylica. However, there 

are some differences between the two taxa. Whereas in A. punctata varo kabylica the involucral 

bracts are completely glabrous and the disc achenes usually have an adaxial corona, 

the involucral bracts are conspicuously hairy in A. punctata varo abylaea, where 

achenes lack a corona. The results of the RAPD analysis point in a different direction. 

OTU 1, representing A. abylaea in the molecular analysis, is found in a cluster with two 

tetraploid representatives of A. pedunculata s.l., one from a nearby population in the W 

Rif mountains (OTU Il), the other from the NE Moroccan Beni Snassen mountains (OTU 

12). Other tetraploid Moroccan representatives of A. pedunculata s.l. (OTUs lO, 13), 

however, are nested among the Moroccan and Spanish diploids (OTUs 4, 6, 7, 8, 26, 33). 

3~---------------------------------------------------. 

2 

PC3 

24 

40 .... ~ 

21 .... ,! ........ ! 

~,./).-.:.:/.;;,./ ! 

f 

j 

; 

38 

o 

-1 

-2 

1 • 

.""" ,-!/",,'J:.


The two tetraploid groups differ markedly in the number of amplification bands found in 

the RAPD analysis: OTUs lO and 13 have 27 and 29 informative bands, respectively, 

which falls within the range found in the diploids (OTU 4: 28, OTU 6: 23, OTU 7: 29, 

OTU 8: 27, OTU 26: 28, OTU 33: 27); the tetraploids clustering with A. abylaea have 

higher numbers of informative bands (OTU 1: 39, OTU 11 : 39, OTU 12: 38), which may 

indicate their likely allotetraploid origin - the other tetraploids then being considered 

autotetraploids. However, most of the bands found in the presumed allotetraploids also 

occur in the local diploids, and since in the present study no assumption of variation of 

amplification bands within and between plant populations was made, this "allopolyploidisation" 

may also result from the combination of two different local genotypes of the same 

species (when the plants would stili be autotetraploids) rather than from the combination 

of two substantially different genomes of two independent species (as in true allotetraploids). 

The occurrence of four amplification bands (primer A-02: 779 bp, 765 bp, 751 bp, 

737 bp) that are shared by the three presumed allotetraploids, Algerian representatives of 

A. pedunculata s.I., and Tunisian diploids (A . ubensis) and tetraploids (A . punctata) supports 

the allotetraploid hypothesis, but further evidence is required. 

Only diploid chromosome numbers were so far reported for Moroccan representatives 

of Anthemis pedunculata s.1. (Galland 1985, 1991), but tetraploids proved to be common 

throughout Morocco (Vogt & Oberprieler 1993 and present study), from the Rif mountains 

in the N to the Anti-Atlas mountains in the S. They mostly grow in the same area as 

diploids, often in mixed stands. Since morphological and genetic differences between 

diploids and tetraploids were found to be small, it is reasonable to assume the tetraploids 

to be autotetraploids derived directly from the loeal diploids. Treating diploids and 

autotetraploids as a different speeies on the grounds of reproductive isolation is undesirable 

as long as no new morphological characters evolve within and spread through some 

or all tetraploid populations. As no sue h eharacters were found, diploids and tetraploids 

are treated as cytotypes of a single speeies, A. abylaea being the single exception to that 

mie: plants of this tetraploid segregate differ morphologically from all other diploid and 

tetraploid representatives of A. pedunculata s.1. and are therefore treated as members of an 

independent species. 

The patterns observed in Moroccan populations of the Anthemis pedunculata - 

A. punctata complex are paralleled to some extent in E Algerian and N Tunisian populations, 

where the occurrence of diploid annuals (OTUs 40-43) that fall within the morphological 

range of perennial A. ,Jedunculata s.1. (OTUs 27-32) adds to the complexity. Due 

to their annual habit, these plants were often considered to be closely related to annuals 

from S Tunisia, and united with them in A. pedunculata var. decumbens. However, RAPD 

analysis shows that the N Tunisian annuals are more closely related to the N Tunisian 

perennials of A. pedunculata s.I. and A. punctata, while the S Tunisian annuals, treated in 

the present study as A. confusa (OTUs 44-46), are clustered with A. stiparum subsp. stiparum 

(OTU 47) and subsp. sabulicola (OTU 48) of A. sect. Anthemis. Morphological 

evidence shows that the N Tunisian annuals hold an intermediate position between 

A. pedunculata s.I. and A. secundiramea, an annual species growing along the E Algerian 

and N and E Tunisian coasts of the Mediterranean Sea. The N Tunisian annuals share with 

A. pedunculata s.1. the erect habit, dark tipped pales (found at least in some populations), 

and the occurrence in mountainous inland habitats, while characters typical for 

A. secundiramea are the annual habit, the much-branched capitulescence (stems in

Bocconea 9 - 1998 101 

A. pedunculata s.1. usually bear up to three capitula), the peduncles which become inflated 

at maturity (they remain slender in A. pedunculata s.I.), the conical receptacles 

(hernispherical in A. pedunculata s.I.), and the comparatively small achenes with a relatively 

long corona (A. pedunculata s.1. usually has longer achenes with very short coronas). 

A majority of these characters thus favour a relationship with A. secundiramea. The 

occurrence of a segregate inland taxon of the coastal A. secundiramea has a parallel in the 

case of A. muricata, a discoid relative of A. secundiramea that occurs only inland in C 

Sicily. The intermediate annuals of N Tunisia (OTUs 40-43) are thus better placed in the 

44 confusa (2x) 

48 sabulicola (2x) 


47 stiparwn (2x) 


40 ubensis (2x) 

31 atlantica (4x) 

35 punctata (4x) 




----------------------------~- 

20 pedunculata (2x) F: 

21 pedunculata(2x) ~ 

28 atlantica (2x) ;:t> 

I abylaea (4x) 

II pedunculata (4x) 

12 pedunculata (4x) 

33 turolensis (2x) 

IO pedunculata (4x) 

49 zaianica (2x) 



26 discoidea (2x) 

3::: 

O 

~ 

(ì 

(ì 

O 

+ 




Fig. 22. Anthemis pedunculata - A. punctata complex: neighbour-joining distance analysis of 

Jaccard distances from RAPD pattems given in Table 25.

102 Oberprie1er: Anthemis in N Africa 

vicinity of A. secundiramea than near A. pedunculata s.l., even though the RAPD ana1ysis, 

where they cluster with representatives of A. pedunculata S.l. and A. punctata from N 

Tunisia, suggests intermediacy between A. sect. Anthemis and sect. Hiorthia. The occurrence 

of aneup10id chromosome numbers in two popu1ations, indicating genomic imba1- 

ance, may a1so point to a possib1e hybrid origino Inclusion of the N Tunisian annua1s in 

A. secundiramea or A. pedunculata wou1d harm the taxonomic integrity of either, and 

_----- 20 pedunculata (2x) 

95 



17 

65 





26 discoidea (2x) 

49 zaianica (2x) 


IO pedunculata (4x) 

~ 

~ 

() 

() 

O 

+ 

61 

33 turolensis (2x) 

Il pedunculata (4x) 

31 

53 


l abylaea (4x) 






62 

64 

82 


----------------------------------- 




48 sabulicola (2x) 

47 stiparum (2x) 

Fig. 23. Anthemis pedunculata - A. punctata complex: majority-ruIe consensus dendrogram 

summarising 100 dendrograms from a bootstrap analysis of RAPD patterns given in Table 25 . - 

Single dendrograms were constructed by neighbour-joining distance analyses of Jaccard distances. 

Bootstrap values denote the number of trees with a particular grouping of OTUs.

Bocconea 9 - 1998 103 

recognition of a separate species, A. ubensis (described from E Algeria), is therefore proposed. 

However, additional field and experimental work is needed to fully understand the 

complex relationships of that species. 

In parallel to the case of Anthemis abylaea in Morocco, some of the Tunisian 

tetraploids (OTUs 34-36) are morphologically distinct enough (see Fig. 21) to warrant 

their recognition as an independent species, A. punctata. Same as A. abylaea, A. punctata 

shows a marked tendency to large flower heads and long basai leafs. It shows a striking 

overall similarity with a tetraploid from Sicily, described as A. cupaniana (OTU 39), that 

is often treated as a subspecies of A. punctata. In the multivarite analysis of morphological 

characters these two tetraploids are indeed very similar, but there are considerable differences 

in the shape of the membranous margin of involucral bracts, which in A. punctata is 

broad and bicoloured (darker brown inwardly), but in A. cupaniana is narrow and 

uniformJy light to dark brown. 

After the segregation of the E Algerian and N Tunisian annuals (Anthemis ubensis) and 

the large-headed A. punctata, what remains of A. pedunculata s.l. is rather homogenous, 

consisting of small-headed perennials with a continuous range from Morocco and Spain to 

N Tunisia. Within this range clinal variation exists in the shape of inner involucral bracts 

and pales, as shown by PC 2 of the PCA (Fig. 20) and a1ready discussed. The two extremes 

are connected by a series of intermediate plants throughout W and C Algeria, but 

may nonetheless deserve recognition as subspecies. The type of A. pedunculata (OTU 3) 

falls within the western group; the deviating populations in E Algeria and N Tunisia are 

named A. pedunculata subsp. atlantica, based on A. atlantica (OTU 27), described from 

the Aurès mountains near the Algerian-Tunisian border. Quantitatively, as seen in the PCA 

plot (Fig. 20), A. pedunculata subsp. atlantica is very similar to A. ubensis, which ho w 

ever differs in its annual growth, much-branched capitulescence, peduncles somewhat 

inflated at maturity, conical receptacles, and smaller achenes with usually longer coronas. 

Anthemis pedunculata subsp. pedunculata, comprising diploids and tetraploids and 

found growing throughout S Spain, Morocco, E and C Algeria, shows little quantitative 

variation that might call for taxonomic recognition. The sculpturing of achenes has been 

used to delimit A. granulata (with extremely tuberculate achenes) and A. laeviuscula (with 

quite smooth achenes), and al so in some floras (e.g. Quézel & Santa 1963) to separate 

A. pedunculata subsp. pedunculata and subsp. tuberculata, but appears to be quite variable 

even within populations, and highly dependent on the degree of maturity (the type 

specimen of A. laeviuscula has unripe achenes, plants with ripe achenes collected later at 

the locus classicus have normally tuberculate achene surfaces), providing scant support to 

a further subdivision of A. pedunculata subsp. pedunculata. Only discoid plants, found 

within the range of A. pedunculata subsp. pedunculata in S Spain and in the Rif and High 

Atlas mountains, may deserve formai recognition, as a variety (var. discoidea), although 

they are not geographically separated and may have originated in parallel in the different 

regions. 

Anthemis turolensis (OTU 33), both in morphological and RAPD analysis, is well 

nested among the OTUs representing A. pedunculata subsp. pedunculata. According to 

Benedi i Gonzalez & Molero i Briones (1985) and Benedi i Gonzalez (1987), who studied 

Spanish populations of HA. tuberculata subsp. tuberculata" (= A. pedunculata subsp. 

pedunculata) and subsp. turolensis, the two taxa differ in achene length, shape of pales, 

and the colour and length of leaves. Fernandes (1976) also used achene length to discrimi-


nate between HA. tuberculata subsp. tuberculata" and subsp. turolensis. However, the 

respective ranges of achene length overlap, being 1.3-2.1 rnrn in A. tuberculata subsp. 

turolensis and 1.8-2.5 rnrn in A. tuberculata subsp. tuberculata. Benedf i GonzaIez (1987) 

describes the pales of A. tuberculata subsp. tuberculata as linear-subulate and usually 

tinged with black at the apex, those of A. tuberculata subsp. turolensis as oblanceolate 

with concolourous tips. Leaves of A. tuberculata subsp. tuberculata were found to be dark 

green and 2-3.5(-6.0) cm long, those of A. tuberculata subsp. turolensis, light green and 

only 1.2-2 cm long. As HA. tuberculata subsp. tuberculata" grows in the S Spanish 

mountain ranges and subsp. turolensis is restricted to mountains in C Spain (Sistema 

Ibérico), the treatment of A. turolensis as a subspecies of A. pedunculata (= A. tuberculata) 

is logica!. 

Anthemis clausonis, restricted to the coasts W of Algiers, also belongs to the 

A. pedunculata group of taxa and is treated here as its subspecies, A. pedunculata subsp. 

clausonis. Due to lack of fresh plant material this taxon was omitted from the morphological 

and molecular analyses. With its narrowly elliptical to subulate pales it c1early falls 

within the range of A. pedunculata subsp. pedunculata but differs by its pale membranous 

margins of the involucral bracts, when in all other taxa of the A. pedunculata group they 

use to be light to dark brown or even black. As A. ubensis from E Algeria and N Tunisia, 

and unlike other members of the A. pedunculata group, A. clausonis has a fairly branched 

capitulescence, but it differs from A. ubensis by its perennial growth and the shape of 

involucral bracts and pales. By the pale membranous margins of involucral bracts and the 

much-branched habit, A. clausonis resembles the NE Moroccan A. mauritiana and the S · 

Spanish A. bourgaei of A. ser. Bourgaeinianae which, however, are annuals with readily 

caducous pales. 

Anthemis zaianica is here described as new to accornrnodate a population in the Zaian 

area in Morocco that also holds a somewhat intermediate position between the A. pedunculata 

group and A. ser. Bourgaeinianae. It was not inc1uded in the numerical analysis, but 

morphologically it fits in A. ser. Bourgaeinianae due to its annual habit and its readily 

caducous pales. In contrast, results of RAPD analysis piace a representative of A. zaianica 

among Moroccan and Spanish populations of A. pedunculata, well apart from the two 

specimens of A. stiparum. The dendrogram in Fig. 22 al so indicates, however, that 

A. zaianica deviates genetically from its nearest neighbour, OTU 6, since the branch 

length between A. zaianica (OTU 49) and the internai node with OTU 6 is much longer 

than is the rule for representatives of A. pedunculata, i.e. genetic deviation of A. zaianica 

from a joint "cornrnon ancestor" conspicuously exceeds that of OTU 6. One may hypothesise 

that A. zaianica is rather an annual segregate of A. pedunculata than a member of 

A. ser. Bourgaeinianae, and that annual growth and caducous pales have evolved in parallei 

in A. zaianica and in the mentioned series. That hypothesis may be extended to question 

the monophyly of A. ser. Bourgaeinianae altogether by assurning that its members all 

derive from A. pedunculata, having become annuals independently in different areas in 

habitats that favour annuallife forms: the dry high plateaus of Algeria (A. stipa rum subsp. 

stipa rum and subsp. intermedia, A. monilicostata), the border lands of the Sahara desert 

(A . stiparum subsp. sabulicola), or the rather arid coastal areas of NE Morocco (A. mauritiana) 

and SW Spain (A. bourgaei). For the latter species a relation to A. tuberculata 

(= A. pedunculata) was already suggested by Talavera, cited by Benedf i Gonzalez (1987: 

216). Since A. pedunculata is a rather short-lived perennial herb that may often flower

Bocconea 9 - 1998 105 

already in the first year, the switch to annual Iife form is easy, and indeed took piace in 

some diploid populations of A. pedunculata in Morocco growing as weeds in arable fields, 

while plants growing on adjacent limestone c1iffs were tetraploid and distinctly perennial 

(e.g. at Foum Kheneg, Middle Atlas, OTUs 8 and 13). Our morphological and molecular 

studies are not sufficient to prove or disprove the monophyly of A. ser. Bourgaeinianae 

(and consequently of both A. sect. Anthemis and sect. Hiorthia). The answer must await a 

comprehensive study on the phylogeny of the genus. Pending this, I consider the annual 

habit and the readily caducous pales of A. zaianica to be sufficient grounds for its inc1usion 

in A. ser. Bourgaeinianae. 

Variation within Anthemis pedunculata in C and E Algeria and N Tunisia is paralleled 

in A. punctata in the same area. In the E Algerian and N Tunisian mountains (where 

A. pedunculata is represented by subsp. atlantica) A. punctata is characterized by obovate 

inner involucral bracts and tricuspidate pales (A. punctata subsp. punctata). The largeheaded 

plants from the Djurdjura mountains (where A. pedunculata is represented by 

subsp. pedunculata) differ by having elliptical inner involucral bracts and gradually tapering 

pales (A . punctata subsp. kabylica). Perhaps the local small-headed plants did participate 

in the evolution of the large-headed plants in their area. The large-headed plants 

being supposedly tetraploid, this participation may consist in the involvement of smallheaded 

diploids in the formation of these tetraploids. As cytological and morphological 

results demonstrate for Moroccan populations of A. pedunculata s.l., the increase of 

flower heads is not necessarily caused by autotetraploidisation. Thus, the large-headed 

Algerian plants may well have an allotetraploid origin, with the local A. pedunculata only 

contributing for one half, which would stili explain the observed paUern of parallel morphological 

variation. Since neither cytological nor experimental or molecular investigations 

were feasible with Algerian plants due to the present inaccessibility of this country, 

we are unable to test this hypothesis. 

A diagrammatic summary of the taxonomic relationships within the Anthemis pedunculata 

- A. punctata complex, as inferred from morphology, chromosome numbers and 

? 

t 

turolensis 

clausonis 

, , , , , , 

SPAIN ,MaROCCO, C ALGERIA ,C ALGERIA, E ALGERIA ,TUNISIA, 

, , Babors , , , 

? 

SICILY 

4X 

2X 

Fig. 24. Diagram summarising the hypothesised taxonomic relationships in the Anthemis 

pedunculata - A. punctata compI ex.


molecular data is given in Fig. 24. On the diploid level one observes a strong morphological 

differentiation within A. pedunculata, expressed by the recognition of four allopatric 

subspecies: A. pedunculata subsp. pedunculata (Betic mountain range of S Spain, alI 

Moroccan mountains from the Rif to the Anti-Atlas, W and C Algeria). In the Algerian 

Babors mountains (Petite Kabylie), parts of the Aurès range, on the Monts de Constantine 

and in N Tunisia it is replaced by A. pedunculata subsp. atlantica, in the Sistema Iberico 

in C Spain by A. pedunculata subsp. turolensis, and in a small coastal area W of Algiers 

by A. pedunculata subsp. clausonis. In E Algeria and N Tunisia, A. ubensis, an annual, 

presumably hybrid species mediating between A. pedunculata and A. secundiramea, replaces 

A. pedunculata. Variation observed on the diploid level is to a large extent paralleled 

on the tetraploid level. Presumed autoploidisation has led and stilI leads to the formation 

of tetraploids throughout the range of A. pedunculata subsp. pedunculata and 

subsp. atlantica that morphologicalIy strongly resemble their diploid progenitors. An 

allotetrapioid origin is suggested for the morphologicalIy deviating taxa of the A. punctata 

group, involving the local diploids of A. pedunculata and resulting in A. punctata subsp. 

punctata in E Algeria and Tunisia, in A. punctata subsp. kabylica in the Djurdjura 

mountains of C Algeria, and in A. abylaea in the W Rif mountains in N Morocco. The 

second species involved in this assumed alloploidisation is not yet known, perhaps some 

diploid with large flower heads, like the Sicilian endemie A. ismelia, may have played this 

role. Alternatively, one may speculate that A. punctata results from the hybridisation of a 

large-headed tetraploid species, like A. cupaniana or A. eretica subsp. columnae, with 

local autotetraploids of A. pedunculata s.l. At present it is impossible to decide between 

the two scenarios, further molecular and cytogenetic work, especially in Algeria, being 

needed.

Christoph Oberprieler The Systematics of Anthemis L. - Herbmedit.org

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?