Table Of Content

© Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at The Species Problem in the Myxomycetes I. CLARK Abstract Small to microscopic eukaryotic orga- sted that the taxonomist become familiar nisms, such as the myxomycetes, often with the developmental and reproductive present taxonomic difficulties for the biology of the myxomycetes in order to systematise since they generally have a develop a better understanding of the range combination of limited morphological and causes of morphological variations and traits, frequent asexuality, poor fossil the units of evolution in this group. record, and a widespread dispersal of the Hopefully, this information will help taxo- morphospecies. Also, a workable taxonomists modify the current typologically nomy requires that a single species concept defined species into a more natural system. be adopted throughout a group. Thus, in Some of the more difficult taxonomic pro- these organisms the options are either a blems, in this group, seem to stem from the very narrow concept which tries to iden- presence of high levels of geographically tify each morphologically or reproductive- restricted apomictic clonal lines. Since ly isolated segment (including asexual these small genetically isolated populations lines), or a broad concept based upon a are capable of independent evolution, they group of related morphotypes and repro- may develop a distinct morphology. This ductive groups. Both of these approaches would then be seen as a swarm of poly- are difficult to realize with the classical morphic local populations, which would typological species concept. Therefore, greatly complicate the taxonomy. Since other species concepts have been reviewed similar apomictic lines are found in a num- as alternatives to the typological concept, ber of vascular plant genera, a survey of but, for the present, these are probably how it is handled in these groups can help best seen as supplements which can serve to understand the consequences of apply- to help broaden and define the current ing different taxonomic concepts to this morphological species concept. It is sugge- problem. Stapfia 73, zugleich Kataloge des 00. Landesmuseums, Neue Folge Nr. 155 (2000), 39-53. 39 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at The Species Problem morphological discontinuities which are thought to reflect genetic and evolutionary dif- A species in the Darwinian natural system ferences. The Linnaean view considers species is defined as a discrete, reproductively defined to consist of individuals conforming to a con- collection of populations with a common evo- stant morphological form which varies only lutionary history (METTLER et al. 1988). Since within narrow boundaries. Consequently, a kinship in a coherent group by evolutionary single typical specimen (type) can be used to descent is the fundamental criterion for taxa in characterize the essential features of the typo- the natural system, the delimitation of a spe- logical morphospecies (CAIN 1957) and, there- cies often depends upon interpreting available fore, new species are often described from the evidence in terms of evolutionary relatedness study of a few individuals collected in a single and separation. However, the application of locality. The Darwinian view, however, holds this definition to taxonomy is difficult, since that widely separate populations, which appear direct evidence of reproductive interconnec- to be morphologically distinct species, may tions in descent is seldom present and often have character gradients in regions of contact limited to inferences derived from morphology. which are indicative of genetic relatedness. For This difficulty has been called the species pro- this reason, and the increasing differential blem: i.e., how do you determine diagnostic powers of SEM and other methods of exami- levels of similarity (relatedness) and differen- ning morphological characters it has become ces (separation) with limited information. difficult to determine if a typological unit is in Also, since at least some speciation is believed itself a species or a subordinate group (race) of to occur by means of a gradual divergence of a larger species. This morphological species populations, there can be no clear cut point in concept has proven to be rather inadequate in this continuous process between races, subspe- the myxomycetes, since the sporangia, upon cies and.species. This is an acute problem in which almost all taxonomy is based, has a rela- the myxomycetes and other eukaryotic micro- tively simple structure with a limited suite of organisms (SONNEBORN 1957), where the variations whose phenotypic plasticity is large- combination of relatively limited morphologi- ly unknown. This produces typological taxa cal traits, frequent asexuality, widespread based upon a limited number of traits of dispersal of the morphospecies, and a poor fos- unknown constancy and thus each taxonomist sil record (DOMKE 1952; WAGGONER & produces a classification based upon his own POINAR 1992) makes the determination of intuitive evaluation of these characters. An natural species especially challenging. additional problem is the fairly frequent production of aberrant fructifications due to the disturbances of the developmental process. Species Concepts Without a thorough understanding of phenotypic plasticity, such aberrations may be descri- Thus, taxonomists have used a number of bed as new species or even genera, as in the species concepts (conceptual level philosophi- case of Squamuloderma nullifila KOWALSKI cal criteria used to define the species category) which is now considered to be a Didymium in attempts to provide useful guidelines for the (FARR 1982). This coincides with an estimati- delimitation of taxonomically useful (practical on that approximately 50% of all described level) species. While numerous species con- myxomycete species are known from only the cepts have been proposed, the four most appli- type locality or less than five localities (LADO cable to the myxomycetes are the morphologi- 2000). cal, biological, cladistic, and phenetic con- The biological species concept (MAYR cepts. 1970) defines species as "groups of actually or The morphological species concept is the potentially interbreeding natural populations oldest and still the most commonly used con- which are reproductively isolated from all cept. In this concept, morphologically similar other such groups". Since reproductively defi- individuals are designated a species and are ned populations are the major thrust of the assumed to be separated from other species by 40 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at natural system, the biological species concept derived characters are then used to construct, should correspond fairly closely to this system. using a number of different algorithms, clado- However, two major criticisms are leveled at grams (hierarchical tree patterns) that are this concept. First, it applies only to sexually believed to reflect evolutionary relationships. reproducing organisms (asexual or apogamic While morphological data can, in theory, be population can not be classified in this used in cladistic analysis, it is often difficult to system). Second, the primacy of the interbree- distinguish between primitive and derived ding requirement over all other characteristics morphological traits. However, DNA sequence can produce seemingly unreasonable results data can be ordered fairly easily and thus have (geographically disjunct populations, which become an important tool in taxonomy, espe- may have fairly distinctive morphological dif- cially in determining relationships at the hig- ferences, can sometimes interbreed). This pro- her taxonomic levels. Since the generation of blem arises from the fact that reproductive DNA sequence data is relatively expensive and isolation is not always correlated with morpho- technically sophisticated, its use in myxomy- logical differentiation. In sympatric (geo- cete taxonomy has been extremely limited graphically overlapping) and contiguous popu- (JOHANSEN et al. 1992) and in the foreseeable lations, morphological distinctness is quickly future it will probably be restricted to determi- integrated if there are no reproductive barriers; ning order and family relationships. however, in allopatric (geographically sepa- The phenetic species concept defines a rate) populations, which normally do not share species on the basis of overall similarity de- a gene pool, morphological differences can rived from a quantitative analysis of many cha- accumulate without reproductive barriers. This racters (DUNN & EVERITT 1982). Pheneticists can lead to biological species designations believe that quantitative measurements which varying from those consisting of morphologi- are treated by a set of clearly defined rules, pro- cally distinct and geographically disjunct duce an objective taxonomy. These measure- populations (which are interfertile) to those in ments are used to construct, using a number of which the species are morphologically identi- different algorithms, phenograms (hierarchical cal, but reproductively isolated (sibling spe- tree patterns) that are believed to reflect true cies). Therefore, the application of the biologi- relationships if enough characters are utilized. cal species concept to the myxomycetes poses This process is relatively easy to use and can be some difficulties. First, many morphologically applied to any codified data but it has been defined species contain both sexual and apoga- mostly used for morphological and protein mic strains (CLARK 1995), which are obvious- (isozyme) characters. However, in practice the ly related but do not fit in this concept. selection of traits is not purely objective and Second, these reproductive studies can only be the use of different traits can sometimes pro- conducted in the laboratory, require considera- duce different classification schemes. This ble time and effort, and only about 10% of the concept is also in its infancy in regards to myxomycete taxa can be easily cultivated. myxomycete taxonomy, with a single paper Also, culture determinations of reproductive published by ELHACE et al. (2000) on the isolation are not always reliable, since some Didymium squamubsum (ALB. & SCHW.) FRIES populations will interbreed and form hybrids in species complex. This procedure seems to be culture that would be unviable in nature. especially difficult to apply to myxomycete The cladistic species concept defines spe- morphology since the limited number of traits cies as populations which share derived cha- appear to be highly variable due to the effects racters (ELDRIDGE & CRACRAFT 1980). of environmental conditions during develop- Cladists distinguish between primitive (the ment. original trait in the ancestral population) and Since none of these concepts, in itself, derived (a change from the original trait which seem capable of producing a natural system occurs in some of the descendants of the ance- taxonomy for the myxomycetes, a pragmatic stral population) characters and use only deri- approach using the analyses and limited infor- ved characters to determine relatedness. These mation derived from these other systems to 41 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at Supplement morphological studies seems to be from each of these apomictic individuals are all the obvious direction for future studies in this identical and form clonal lines which are gene- group. At a minimum, this approach could tically isolated from all of the other apomictic help examine and validate various morpholo- lines and also the sexual isolates. This genetic gical characters in terms of taxonomically isolation could result in sympatric populations meaningful variations. accumulating independent variations and becoming adapted to very specific environ- ments. This would result in the production of a The Life Cycle and Taxonomy swarm of related polymorphic microtaxa, some of which could display morphological differen- While almost all taxonomic distinctions ces at a level that would greatly complicate the are based upon sporangial morphology, there taxonomy of this group. However, these mor- are a number of other important stages in the phologically distinguishable apomictic lines myxomycete life cycle. The myxamoebal stage, could revert to sexual reproduction (COLLINS which is derived from the spore by germinati- et al. 1983) and again blur these differences. on and feeds on bacteria, produces clonal The plasmodium, a diploid coenocytic uninucleate amoebal cell populations by mito- amoeboid cell feeding on bacteria and organic tic division. However, this stage is generally material, is another major life cycle stage. quite uniform throughout the group and proba- Although there are a number of different plas- bly does not display any morphological variati- modial types (ALEXOPOULOS 1960; HASKINS & ons that are of taxonomic value at the species HlNCHEE 1974) which have proven useful in level. The myxamoebae are also capable of differentiating higher level taxa, there are few producing flagella (swarm cells) or resistant recognizable plasmodial variations found at the structures (cysts) under appropriate conditi- species level, e.g., Didymium iridis (DlTMAR) ons, but these forms are also quite uniform FRIES and D. ovoideum NANN.-BREM. which are throughout the group. However, the myxamoe- morphologically similar species having brown bae can also be the gametic stage and are thus and yellow plasmodia, respectively (CLARK & directly involved in sexual reproduction and STEPHENSON 1994). However, the plasmodium plasmodial formation. Typically, the myxamoe- can also be taxonomically useful as a source of ba is haploid and expresses a mating type readily available proteins and DNA. For exam- allele which prevents sexual fusion with a ple, isozymes isolated from plasmodia have myxamoeba having the same mating type but been used in a biosystematic study of the allowing fusion and zygote formation with ano- Didymium squamulosum complex (ELHAGE et ther myxamoeba having a different mating al. 2000). Under appropriate conditions, the type allele. Each species has a multiple allelic plasmodium differentiates into fruiting bodies, mating gene (COLLINS 1979) which encoura- of which there are several different develop- ges outbreeding (since there are a large number mental types that are also correlated with of alleles, most myxamoebae from different higher level taxa. individuals will also have different alleles and will therefore undergo sexual fusion). Morphological Information Individuals that are able to interbreed are members of the same gene pool (and presumably a species), while those that can not inter- Although remarkably complex for a pro- breed are in different biological species even if tist, compared to vascular plants the sporopho- they are morphological identical (sibling spe- re of the myxomycetes is a relatively simple cies). However, many of the myxomycete iso- structure with a limited number of compo- lates grown in culture have proven to be non- nents. However, while these components are heterothallic and presumably apomictic rather variable for the myxomycetes as a group, (CLARK 1995). These isolates apparently sup- their variability within a species is generally press meiosis and produce diploid or polyploid not well known nor fully analyzed. Sporophore myxamoebae, which can then convert to the characters can be divided into three types for plasmodial stage without mating. The progeny convenience of discussion: overall aspects 42 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at (type, form, and association), generalized cha- be very closely related. However, a knowledge racters (color and lime), and specific compo- of the basis of the color can provide useful nent variations (capillitium, peridium, spores, information in terms of taxonomic problems. etc.). For example, the melanins of the Physarales The overall aspects of the sporophore and Stemonitales have been found to be diffe- include the type (sporangia, pseudoaethalium, rent (LOGANATHAN et al. 1989), and the aethalium, plasmodiocarp), form (globose, pul- melanins found in the taxonomically difficult vinate, urniform, cylindrical, etc.), and asso- genus Diachea are most similar to the ciation between the fruiting bodies (scattered, Stemonitales melanins (KALYANASUNDARAM clustered, aggregated, etc.). These sporophore & ALI 1989). Also, a Chromatographie study of aspects can be very useful in taxonomy but due pterdines pigments in multiple collections of caution must be used since they can grade into four Hemitrichia species (BLACKWELL &. each other and it is not uncommon to find BUSARD 1978) was able to distinguish between collections in which stipitate sporangia grade the four species, each of which displayed a into sessile sporangia, sporangia into plasmo- number of intraisolate variations. In a similar diocarps, globose sporangia into pulvinate spo- manner, thin layer chromatography of pig- rangia, or gregarious sporangia into clustered ments has provided useful information for dif- sporangia. The only solution for this problem is ferentiating the complex of red Arcyria species the examination of enough material to under- (REBHAHN et al. 1999). Inherited color varia- stand the range of variations found in each tions, within a species, can also occur, as in the taxon. For example, in a study (CLARK et al. brown and cream plasmodial variants in 1999) of 32 isolates of Physarum compressum Didymium iridis (COLLINS & CLARK 1966, ALB. & SCHW., which is characterized as COLLINS &. ERLEBACHER 1969), which direct- having gregarious, stipitate, compressed fan- ly effect sporangial stalk color (CLARK & shaped sporangia, some isolates were nearly to MlRES 1999). Lime color, in the Physarales, is completely sessile (sessile sporangia occurred often used as a taxonomic trait and the basis in most isolates). Other isolates produced for these color differences has been examined reniform and occasionally globose shaped spo- (ALDR1CH 1982). Using a dispersive X-ray rangial heads and several isolates formed spectrometer, he found that lime color diffe- contorted clusters that were identical to the rences, in a number of physaraceous species, morphological description of Physarum nicara- correlated with the presence of certain inorga- guense MACBRIDE. In culture, plasmodia can nic ions (manganese, barium and zinc). form on the Petri dish lid and these "upside Therefore, it is quite possible that the presence down" fructifications are usually sessile and or absence of these ions in the environment of often plasmodiocarpous, whereas the same the developing sporangium may control the plasmodium will produce stalked sporocarps in color of the lime in that specimen. For exam- the rest of the culture. ple P. bilgramii HAGELST., a blue lime species, crossed with Physarum globuliferum (BULL.) The more general color and lime charac- PERS., a white lime species, and P. bilgramii, ters are also rather variable and their under- when grown in agar culture, has white lime. lying basis poorly understood. Sporophore and plasmodial pigments have been investigated in The presence or absence and type of lime a number of species and a complex variety of (calcareous deposits) in a species is also an organic chemical forms have been reported: important key character in the myxomycetes. including indoles (STELICH et al. 1980), tetra- Although calcium is present in the peridium of meric acid derivatives (CASSER et al. 1987), a number of non-physaraceous species carotenoids (CZECZUGA 1980), pteridines (SCHOKNECHT 1975; NELSON et al. 1977; ALDRICH 1982), it is generally found in a (BLACKWELL & BUSARD 1978), and melanins (LOGANATHAN et al. 1989). Thus, simple crystalline form only in the Physarales, al- color may be misleading since two yellow pig- though under certain conditions calcium ments may be due to chemically different com- oxalate and silicon crystals are found in the pounds, while a yellow and a red pigment may peridium of Perichaena and Dianema species 43 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at (SCHOKNECHT & KELLER 1977). In the sting of the slime sheath with scant or no depo- Physareacae, the lime is present on both the sits of wall materials (ROSS 1957; MlMS 1973). peridium and capillitium and generally con- The ornamentations on the inner peridial wall tains both calcium and phosphate, while in the has been examined by several investigators Didymiaceae the lime is present only on the (ELIASSON & SUNHEDE 1980; RAMMELOO peridium and generally contains only calcium 1974a) and may provide useful taxonomic cha- (SCHOKNECHT 1975; SCHOKNECHT & KELLER racters; however, as in all cases of morphology, 1989). While other papers (ALDRICH 1982; care must be taken to insure an adequate deter- NELSON et al. 1977) report different results in mination of the range of variations within terms of the phosphate, SCHOKNECHT believes each taxon. The use of peridial layers as a that this is due to their using whole sporangia taxonomic character also needs to be exami- as an experimental unit as opposed to isolated ned, since these layer may or may not separate crystals. The lime is deposited on and in the at maturity to produce a single, double or triple peridial matrix from the cytoplasm via chan- peridium. The peridium is generally composed nels and pores where the final morphology of of a outer slime sheath layer (presumably the the lime is determined in part by chemical only layer in the Stemonitales and some composition and in part by the pores and Trichiales) and a thicker inner layer derived matrix structure (SCHOKNECHT &. KELLER from the cytoplasm which, for example, sepa- 1989). The Physaraceae produce globular rates in most Perichaena but remains together forms of calcium carbonate with calcium phos- in P. vermicularis and P. luteolum (KOWALSKI) phate and the Didymiaceae produce either GlLERT. The latter has been transfered crystalline calcium carbonate (Didymium) or (GlLERT 1995) from Calonema since it is a globular (cryptocrystalline) calcium carbonate Perichaena in all other aspects. Also, in the (Diderma). Unfortunately, under moist condi- Physarales, free crusts of lime are sometimes tions recrystallization of the lime can occur, considered to be a peridial layer, which adds to which thus camouflages the true nature of the the definitional confusion of this structure. lime crystals. The capillitium is produced by an anasto- However, it is the specific component mosing system of tubular cytoplasmic vacuoles (peridium, capillitium, spores, etc.) variations into which material is secreted (MlMS 1969, which provide the bulk of the taxonomic cha- 1973; CHARVATS et al. 1974), although in the racters in the myxomycetes. The development Stemonitales part of the capillitium is produ- and use of the electron microscope has been a ced by branching of the columella (MlMS major advancement in the detection of fine 1973). Several studies (SCHOKNECHT & details, in the comparison of structures beyond SMALL 1972; ELLIS et al. 1973) have dealt with the resolution of the light microscope and in details of capillitial ornamentation, which are the determination of development of the fruit- often used in taxonomy. An interesting finding ing body structural components. (ELLIS et al. 1973) in the Trichiales was that there is a continuous variation from solid to In Perichaena vermicularis (SCHW.) ROST, hollow capillitia in this order, which brings the peridium has been shown (CHARVATS et al. into doubt the basis for dividing the order into 1973) to be produced by autolysis of the outer the Trichiaceae and Dianemaceae based on the layer of cytoplasm and the laying down of wall hollow vs. tubular nature of the capillitial materials, including the inner peridial orna- threads. mentations, under a thin layer, consisting of the slime sheath and excreted materials, to Spores are cleaved out of the sporangial produce the mature peridial wall. This materi- protoplasm by means of vesicle fusion which al has been shown, in Fuligo septica (L.) produces uninucleate segments of protoplasm WlGGERS, to be composed of protein and car- around which spore wall material, including bohydrates with extensive lime deposits the ornamentations, is deposited (MlMS 1972). (CHAPMAN et al. 1982). The Stemonitales These spore ornamentations are key taxono- appear to be different in that there is no auto- mic characters, although they are at the limits phagy and the peridium is a thin layer consi- of light microscope resolutions in some species. 44 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at Therefore, numerous SEM and TEM investiga- Therefore, this controversial group still needs tions of spore ornamentation have been con- more work in order to determine its relation- ducted (SCHEETZ &. ALEXOPOULOS 1971; ship to the rest of the myxomycetes. Stalk RAMMELOO 1974b, 1975; DEMAREE & length, color, presence of lime, and internal KOWALSKI 1975; HASKINS & MCGUINNESS structures are also used in myxomycete taxo- 1986), and new terminology has been sugge- nomy, and while these can be useful characte- sted (RAMMELOO 1974b). While these spore ristics (MATSUMOTO &. DEGUCHI 1999a, b), ornamentations are considered to be quite con- they are also often quite variable within a stant, they can vary somewhat in different iso- taxon and therefore a determination of the lates of the same species (GAITHER & COLLINS range of variations for a trait is essential when 1984). In fact, as is the case for most charac- they are used. ters, spore ornamentation, size and color may In the Stemonitales, the columella is an be all alike in one collection but slightly diffe- extension, which can be large and branched, of rent between collections (MARTIN & the intra-cytoplasmically secreted stalk (ROSS ALEXOPOULOS 1969, page 21; ELHAGE et al. 1957, 1973; MlMS 1973) and, in the rest of the 2000). Therefore, a number of isolates must be myxomycetes, it is a intrusion, separated by an studied to determine the morphological range extension of the peridium (BLACKWELL 1974), of spore characters for a particular species of the stalk into the sporangial space. Again, a which in certain cases may be quite wide. knowledge of the development of this structu- Stalk and hypothallus development appear re and the variations found within a taxon are to be similar to peridial development. In most necessary if it is to be used in classification. For myxomycetes, an outer slime sheath covers a example, in many Didymium spp. the columel- fibrillar tube which is filled with food vacuoles, la is not so much an intrusion into the sporan- lime, spore-like bodies or other materials gial space as it is a folding back of the sporan- (BLACKWELL 1974; MIMS &. ROGERS 1975) gial space around the stalk apex (WELDEN and it is produced by the constriction of the 1955); therefore, what is usually called the basal region of the presporangial primordium. columella in this genus is actually the inside of This stalk wall is continuous with the peridium the peridial wall surrounding the stalk. This in most cases, although in Arcyria it is conti- process produces an extremely variable structu- nuous only with the calyculus since the rest of re, such that the morphospecies D. iridis, D. the peridium consists of only the slime sheath bahiense GOTTSBERGER, and D. verrucosporum (MlMS &. ROGERS 1975). The hypothallus in WELDEN can all be found in a single biological these myxomycetes is also a continuation of species (CLARK & MIRES 1999). the slime sheath and, in some cases, the stalk wall. In the Stemonitales, the stalk is secreted Non-Morphological Information intra-protoplasmically by the presporangial primordium and is not continuous with the Reproductive Systems peridium (Ross 1957, 1973; MlMS 1973). In this group, the hypothallus is the secreted basal The extensive early studies on myxomyce- portion of the stalk and does not involve a te life cycles and sexuality have been reviewed slime sheath. Diachea leucopodia (BULL.) ROST, by COLLINS (1979). In general, the life cycle has been found to have a non-stemonitaceous was found to include haploid myxamobae type of stalk development (BLACKWELL 1974) which carried a heterothallic mating gene that and D. bulbilbsa (BERK. & BROOME) LISTER controlled syngamy and plasmodial formation. has been seen to develop from a bright yellow This gene displayed multiple alleles which fun- phaneroplasmodium indistinguishable from ctioned to induce outbreeding since any two the plasmodium of a typical Physarum species myxamoebae which differed at their mating (SCHNITTLER pers. obs.). These observations type allele could undergo fusion (myxamoebae are contradictory to the melanin studies which carrying like alleles could not fuse). However, indicate a stemonitaceous relationship some isolates of a species had myxamoebae (KALYANASUNDARAM & ALI 1989). which could form plasmodia without crossing 45 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at to a different mating type allele and were pre- The first report that could be interpreted as sumably diploid or polyploid apomicts. These indicating the presence of genetical isolated apomictic lines are apparently derived from sibling species in the myxomycetes concerned the heterothallic isolates by suppression of isolates of the Physarum flavicomum BERK, mor- meiosis during spore formation, since conversi- phospecies (HENNEY 1967). However, their ons between the two reproductive systems presence and nature were established and de- have been shown to occur in the laboratory fined in a series of papers on Didymium iridis (COLLINS 1980, COLLINS et al. 1983). This (COLLINS 1976; BETTERLEY & COLLINS 1983; basic life cycle information has been accumu- CLARK &. STEPHENSON 1990, CLARK et al. lated for many species (Table 1), and this mix 1991, CLARK &. LANDOLT 1993, CLARK 1995) of multiple allelic heterothallism and apomic- where it was shown that there were numerous, tic lines is apparently nearly standard through- mostly allopatric, sibling species that encom- out the group (CLARK 1995). Of the 12 spe- passed not only the D. iridis morphospecies but cies, in 1995, having their reproductive system also included all or parts of the D. bahiense, and reported for four or more isolates, four display- D. nigripes (LINK) FRIES morphospecies to pro- ed only heterothallism, three only apomixis duce a morphological and genetic complex and the remaining five had mixed systems. At (CLARK & MIRES 1999). Two heterothallic present (2000) 19 species display three hetero- biological species, from the Sonoran and thallic, five apomictic and 11 mixed systems. Mojave deserts, have also been reported for Echinostelium minutum de BARY. These biological species Table 1. Heterothallic and non-heterothallic reproductive systems. have somewhat larger spores 1995 2000 and a more reduced capillitium Species heterothallic non- heterothallic non- when compared to the many heterothallic heterothallic apomictic isolates found in Order Physarales more temperate areas (CLARK Badhamia gracilis 0* 1 0 5 Didymium annellus 0 0 0 g & HASK1NS 1998). Sibling spe- D. diffbrme 0 9 0 9 cies have also been found in D. iridis 27(7)" 40 23(9)< 55 the Didymium squamulosum D. megalosporum 0 0 8C 1 morphospecies; however, in D. nigripes 1 4 0" 0 this case there appears to be a D. ovoideum 14 0 14 8 D. squamulosum 0 7 8(5) 28 higher rate of sympatric rela- Fuligo seplica 4 6 4 6 tionships between the sibling Physarum cinereum 1 2 1 7 species in this morphologically P. compressum 0 2 1 32 P.flavicomum 5(2) 0 5(2) 0 variable taxon (ELHAGE et al. P. melleum 0 0 0 17 2000). These reproductive P. polycephalum 12 1 • 12 1 system findings would seen to P. pusillum 3 4 3 4 indicate that, in general, the P. rigidium 4 0 4 0 myxomycete morphospecies is P. ?straminipes 0 0 3 17 Order Trichiales a complex of genetically inde- Arcyria cinerea 0 | 0 0 6 pendent apomictic lines and Order Echinosteliales allopatric and sympatric sib- Echinostelium coelocephalum 4 0 17 0 ling species. This genetic inde- E. minutum 0 7 7(2) 35 Order Stemonitiales pendence also allows indepen- Stemonitis flavogenita 1 1 7 1 7 dent evolution of each of these groups which could result in ' Number of isolates reported as having a heterothallic or non-heterothallic reproductive system in the review by the accumulation of morpho- CLARK (1995) and at 2000 (not all published). b Numbers in parenthesis indicate the number of biological species which are encompassed in the heterothallic logical variations. Some of the isolates. apomictic lines, which are c For Didymium iridis, 5 isolates reported as heterothallic in 1995 have been transferred to D. megalosporum. generally confined to restricted ' The isolates reported as D. nigripes in 1995 are now considered to be D. iridis. geographical regions, could accumulate enough morpho- 46 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at logical variations to be recognized as separate Proteins taxa. However, this would result in the naming Protein studies have also been very limited of multiple microtaxa which would be extre- in the myxomycetes. The first report on mely confusing and add little or nothing to our myxomycete proteins (FRANKE &. BERRY 1972) knowledge of the group. The sibling species are dealt with a limited number of physaraceous more difficult since they appear to be larger species and indicated that isozymes were pro- entities often with allopatric geographic ran- bably taxonomically valid characters in this ges. It is also quite likely that in some mor- group. However, it was not until a number of phospecies complexes these biological species Didymium iridis morphospecies isolates were have accumulated minor morphological diffe- examined (BETTERLEY &. COLLINS 1983) that rences which could be recognized as different isozyme data relevant to species level taxo- taxa once the underlying genetic information nomy were found. They worked with 28 iso- was known. lates, divided by reproductive systems into 19 non-heterothallc (apomictic) isolates and DNA three reproductively isolated mating series (biological species) consisting of seven isolates The limited DNA sequence data on (from Central America), one isolate (from myxomycete species is taxonomically inter- Kentucky) and one isolate (from Georgia), res- esting. The first report (JOHANSEN et al. 1992) pectively. They found that the seven isolates in indicated that Didymium iridis and Physarum the biological species from Central America polycephalum SCHW., species classified in diffe- shared almost identical isozyme patterns but rent families of the order Physarales, had differed considerably from the isolates of the undergone a very ancient separation. This has other two allopatric biological species. These been confirmed by a recent meetings abstract two isolates, which may be sympatric biologi- (MILLER & KRISHNAN 1999) where six species cal species, also differed from each other. All of were studied: Didymium iridis, D. nigripes, the non-heterothallic isolates, except for those Lycogala epidendrum (L.) FRIES, Physarum dider- which had been collected from adjacent sites, moides (PERS.) ROST., P. polycephalum and also differed from each other and from all three Scemonids flavogentia JAHN. They found that biological species. Isozyme patterns for nine the Didymium, Physarum and Stemonitis species other Didymium morphospecies, one to five diverged at about the same time with Lycogala isolates of each, were also compared for the having the earliest divergence. This is some- same set of isozymes and in most cases the pat- what surprising, since the stemonitaceous tern fell outside of the range found in the D. species have been segregated from the rest iridis isolates. Therefore, this research indica- of the myxomycetes as the subclass ted that the Didymium iridis morphospecies is Stemonitomycetidae on the basis of plasmodi- divided up into a number of related but geneti- al form and sporangium developmental type cally isolated apomictic lines and biological (ROSS 1973). While the differences between species. A recent study of 33 isolates of D. the Physarum didermoides and P. polycephalum squamulosum indicates an even more complex sequences also indicated an ancient diver- set of relationships among the isolates in this gence, the Didymium nigripes and D. iridis morphospecies (ELHAGE et al. 2000). These sequences indicated a closer relationship. isolates consisted of 27 non-heterothallic and However, the Kerr culture used in this study three mating series of four (from Costa Rica), and designated as D. nigripes is generally regar- one (from Costa Rica) and one (from Puerto ded as a D. iridis strain by most researchers in Rico) isolates, respectively. A phenogram deri- the field (BETTERLEY & COLLINS 1983; CLARK ved from an analysis of the isozyme data divi- 1995). The major message from this limited ded the isolates into two branches which parti- DNA sequence data is that the myxomyetes ally correlated with two overlapping morpho- are a very old group and that even species pla- logical forms: a short stalked form generally ced in the same genus may have diverged from resembling the type description, except that it each other in ancient times. has minimal lime deposits in culture, and a 47 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at longer stalked form approaching the type Costa Rica, Puerto Rico, Indonesia and description of D. floccoides NANN.-BREM. & Thailand). While there was some morphologi- YAMAMOTO. Interestingly, these branches did cal variation between these isolates, especially not correlate with reproductive system diffe- in terms of the shape of the sporangial head, it rences or the geographical origin of the isola- was not correlated with any aspect of the isozy- tes. The branch with the more typical morpho- me tree or the geographic origin of the isolates. logy contained one sexual and 17 non-hetero- Apparently, this morphospecies consist of a thallic isolates. These isolates, from different large number of closely related apomictic lines regions of Costa Rica and Indonesia, had fairly that are probably derived from a relatively variable isozyme patterns which did not corre- small number of sexual isolates by means of late with their geographical origin, except for a meiotic suppression. number of non-heterothallic isolates from very restricted local regions. This branch also con- Taxonomic Comparisons with tained one of the isolates (from Costa Rica) Vascular Plant Apomicts which belonged to the four isolate mating series, i.e., it mated with isolates in the other branch. The branch with the more floccoides- Since apomixis also occurs in the more like morphology contained five sexual and 10 extensively studied vascular plants, an exami- non-heterothallic isolates. Five of the non- nation of how this phenomenon is treated in heterothallic isolates were from a single local some of these better known genera may help site and had identical isozyme patterns; how- illuminate the current taxonomic problems in ever, the other five non-heterothallic isolates the myxomycetes. Flowering plants have cer- and the five sexual isolates were scattered tain advantages (size, a large number of easily within the branch without regards to geogra- recognized traits, larger and more complete phic origin (various regions of Costa Rica, collections, and more scientists doing taxo- Puerto Rico and Indonesia) or reproductive nomy) over myxomycetes which has produced system. The three isolates, in this branch, of a better resolution of population and geogra- the four isolate (Costa Rica) mating series phic parameters. Thus the size and location of were intermixed with the sympatric (Costa apomictic clones in these genera are often well Rica) single isolate mating series and two non- known and thus biotypes representing stable heterothallic isolates, and the allopatric apomictic clones can be differentiated from (Puerto Rico) single isolate mating series was non-stabilized recently derived lines. also adjacent to them in the dendrogram. Generally, with this increased knowledge of Apparently, the D. squamulosum morphospe- apomictic vascular plants there has been a ten- cies consists of a complex of allopatric and dency toward recognition of these stabilized sympatric sexual sibling species and numerous apomictic clones as species. However, the more local apomictic lines de-rived from them by frequent the conversion from apomictic to suppression of meiosis. Also, while the correla- sexual reproduction and back again occurs tion of morphological differences and isozyme (and the resulting morphological variability patterns would indicate genetic isolation, the comes closer to a continuum), the more diffi- fact that mating can still occur between parts cult it becomes to distinguish between the apo- of the two morphological groups would suggest mictic clones. The still unresolved taxonomy that speciation is not yet complete in this com- of the Poa pratensis L. complex, or the concept plex. A meetings abstract (CLARK et al. 1999) of "main" and "intermediate" species in concerning isozyme patterns in one sexual and Hieracium reflects this problem (ZAHN 1987). 32 non-heterothallic isolates of the Physarum Another obstacle is the sheer number of apo- compressum morphospecies produced very dif- mictic clones, especially in groups which oc- ferent results. The isozyme patterns of these casionally switch back to sexual reproduction isolates produced a phenogram which indica- as seen in Rubus (brambles), Hieracium (dan- ted that, while rarely identical, they were all delions), and the Rannuculus auricomus L. closely related to each other, including the iso- aggregate (goldilocks) in Central Europe lates from various geographic areas (regions of (Table 2). In Rubus only six sexual species 48