Taxonomy, slime molds, and the questions we ask.pdf

Mycologia, 94(6), 2002, pp. 968–979.

Taxonomy, slime molds, and the questions we ask

Andrew R. Swanson 1

Frederick W. Spiegel

Department of Biological Sciences, University of

Arkansas, Fayetteville, Arkansas 72701

James C. Cavender

Department of Environmental and Plant Biology, Ohio

University, Athens, Ohio 45701

size stability (Mayr and Ashlock 1991, Greuter et al

2000). With the classical identiﬁcation, naming, and

subsequent cataloging of new species, evolutionary

theory has historically taken a secondary role to the

traditional (and intuitive) view that similarity of form

(whether macroscopic, microscopic, or molecular)

should indicate relatedness between taxa. Morpho-

logical, physiological, behavioral, and genetic char-

acteristics are certainly important in the development

of a taxonomic system. Once in place, however, a tax-

onomic system intrinsically drives the nature of bio-

logical inquiry, typically by inspiring assumptions

about the evolutionary processes that contribute to

the diversity of form.

The classical taxonomic grouping of Oomycota

with fungi, for example, encouraged the assumption

that ﬁlamentous growth and absorptive nutrition

were synapomorphies unifying them with the Fungi.

We now know that oomycetes are phylogenetically al-

lied with the heterokont algae (Barr 1992), a lineage

quite removed from the Opisthokonts [Fungi and

Animals (Cavalier-Smith 1998)]. This phylogenetic

realization veriﬁes convergence of a key taxonomic

character–ﬁlamentous hyphae—and leads to the

awareness that traditional taxonomy has impeded for-

mulation of the most fundamental and pertinent

questions regarding hyphal origins in all groups of

mycelial organisms (e.g., how many times have hy-

phae originated?).

Similarly, the aggregation of single amoebae into a

multi-celled fruiting body was an important taxonom-

ic character that uniﬁed the acrasids (sensu Olive

1975) and dictyostelid cellular slime molds into a

larger Class Acrasiomycetes (Raper 1984). Olive’s

(1975) observations leading to studies by Page and

Blanton (1985), and Roger et al (1996) have dem-

onstrated that the acrasids are members of the Het-

erolobosea, a group phylogenetically distant from the

Eumycetozoans. Differences in morphology have

turned out to be more important than superﬁcial

similarities of aggregation, and again, convergence

has been recognized in what was traditionally viewed

as a unique evolutionary event.

Taxonomic treatments inﬂuence the way we form

our inquiries, and we often fail to ask the right ques-

tions about the evolutionary mechanisms involved if,

for instance, convergence of characters is never con-

Abstract: Taxonomic treatments often inﬂuence the

way we both ask and attempt to answer certain bio-

logical questions. The classical taxonomy of the dic-

tyostelid cellular slime molds (Dictyosteliales) in-

volves a convenient set of categories that were devel-

oped independent of phylogeny. In order to test

whether the characters supporting the classical tax-

onomy hold any phylogenetic signal, we subjected 19

described taxa belonging to two families (Acytosteli-

aceae and Dictyosteliaceae) and three genera ( Acy-

tostelium , Dictyostelium , and Polysphondylium ) to root-

ed cladistic analyses using PAUP* v 4.0b4a. Neither

family nor any of the three genera were found to

represent monophyletic groups. These results con-

ﬁrm that the classical taxonomy used to delineate

families and genera within these slime molds carries

very little phylogenetic signal. Taxonomic character

sets should be scrutinized phylogenetically in order

to determine what information they provide about

the relatedness of taxa within a group. Because tax-

onomy often drives the nature of biological inquiry,

caution should be exercised when drawing conclu-

sions regarding the evolution of developmental sys-

tems in Dictyostelium .

Key Words: Acytostelium , Dictyostelium , Eumyce-

tozoa, Phylogeny, Polysphondylium

INTRODUCTION

A phylogeny is a scientiﬁc hypothesis. A taxonomy by

itself is not. Yet we often treat taxonomies as hypoth-

eses, and the questions we pose regarding phyloge-

netic relatedness are often driven by the insinuations

taken from the underlying taxonomy. The rules of

both botanical and zoological nomenclature empha-

Accepted for publication May 22, 2002.

1 Corresponding author, Email: arswans@uark.edu

968

2002 by The Mycological Society of America, Lawrence, KS 66044-8897

S WANSON ET AL :T AXONOMY AND SLIME MOLDS

969

T ABLE I. Probable synapomorphies among the Dictyosteliales

Synapomorphy

Reference

Streaming aggregation of individual myxamoebae fol-

lowed by differentiation into a multicellular fruiting

body composed of stalk and spore mass

Shaffer 1964; Bonner 1967; Hohl et al. 1968; Olive 1975;

Bonner 1982; Raper 1984

Stalk tube synthesis & ultrastructure

Gezelius, 1959; Hohl et al., 1968; George et al., 1972

Spore ultrastructure

Gezelius, 1959; Hohl et al., 1968; Hohl & Hamamoto,

1969

Myxamoeba ultrastructure

Gezelius, 1959; Hohl et al., 1968

Nature of mitosis

Roos, 1975; Moens, 1976; Heath, 1980; Raper, 1984

Ultrastructure of microtubule centers (MTC’s)

Guhl & Roos, 1994

Nucleoli peripheral in the nucleus

Gezelius, 1959; Hohl et al, 1968

Pseudoplasmodium organization

MacWilliams & Bonner, 1979

sidered. This trap is well camouﬂaged, owing ﬁrst to

modern taxonomy’s presumed acceptance of an evo-

lutionary worldview, and second to the history of no-

menclature in each group of related organisms. Bi-

ologists must be mindful that our ideas about how

characters evolve can be highly inﬂuenced by tax-

onomy.

Subclass Dictyosteliidae, Order Dictyosteliales is

clearly a monophyletic assemblage (T ABLE I) within

the Eumycetozoa, a natural group that includes the

protostelid, dictyostelid, and myxogastrid slime

molds (Olive 1975, Dykstra 1977, Drouin et al 1995,

Spiegel et al 1995, Keeling and Doolittle 1996, Bal-

dauf and Doolittle 1997, Baldauf 1999). The Dictyos-

teliales have traditionally been divided into two fam-

ilies: the Acytosteliaceae (which includes Acytoste-

lium ), with an acellular, hollow stalk, and the Dic-

tyosteliaceae (which includes Dictyostelium and

Polysphondylium ), which have a cellular stalk (Olive

1975, Raper 1984). Oskar Brefeld (1869) was the ﬁrst

to isolate and describe a dictyostelid, Dictyostelium

mucoroides , whose generic name was chosen based on

the net-like appearance of the fruiting body’s stalk

cells (Raper 1984). Members of Dictyostelium possess

relatively large fruiting bodies that are typically un-

branched or irregularly branched (F IG .1 A ). Brefeld

later (1884) described a second species, Polysphon-

dylium violaceum , complete with a new generic des-

ignation based on the regularly-whorled branches of

the fruiting body’s cellular stalk (F IG .1 B ). These two

genera were included in the family Dictyosteliaceae.

Acytostelium leptosomum , described by Raper in 1956,

and later characterized fully by Raper and Quinlan

(1958), possessed tiny, delicate fruiting bodies with

acellular hollow stalks (F IG .1 C ), and was deemed

unique enough to be assigned to a third genus in its

own, new family Acytosteliaceae.

Much speculation has been made on the phylo-

genetic relationships within the dictyostelids, but

none of these studies has questioned 2 basic assump-

F IG .1.a. Dictyostelium mucoroides ,b. Polysphondylium violaceum ,c. Acytostelium leptosomum . Bar

1 mm.

970

M YCOLOGIA

F IG . 2. Proposed evolutionary relationships among the

Eumycetozoa (modiﬁed from Olive 1975).

sistent with the current taxonomy: (i) acellular stalk

is a plesiomorphic character state, while cellular stalk

is a synapomorphy that deﬁnes the family Dictyoste-

liaceae; (ii) evenly-spaced whorled branching is a syn-

apomorphy that deﬁnes the genus Polysphondylium .

We also discuss the inﬂuence our results may have on

the formulation of questions about the evolution of

key characters deﬁning the two families and three

genera of the group.

tions implied by the taxonomy: (i) the ﬁrst dictyos-

telid had acellular stalks, and cellular stalk evolved

only once; (ii) regular, whorled branching evolved

only once. Holmes (unpubl), in a preliminary phy-

logenetic study of 24 species of dictyostelids, placed

several of the smallest species (including D. minu-

tum ) at early branching points, suggesting their prim-

itive evolutionary position. Vadell and Cavender

(1991) presented a phylogeny of 31 dictyostelid taxa,

showing a monophyletic Polysphondylium emerging

from within a paraphyletic Dictyostelium . The clado-

grams of both Holmes and Vadell and Cavender sug-

gested that the genus Dictyostelium is paraphyletic.

However, both of these analyses used Acytostelium as

an outgroup, rather than including this dictyostelid

within the analysis group.

Outgroup selection is obviously an important mat-

ter, and for examining relationships within the Dic-

tyosteliales, the use of a Eumycetozoan sister taxon

is most appropriate. Olive and Stoianovitch (1960)

hypothesized a relationship between the protostelid

Protostelium mycophaga and the dictyostelid genus

Acytostelium based on the two groups’ very similar

non-ﬂagellated amoebae and acellular fruiting body

stalks. Molecular work has supported a close rela-

tionship between Protostelium and dictyostelids (Dut-

ta and Mandel 1972, Spiegel et al 1995), as well as

between the protostelid Planoprotostelium (a close

relative to Protostelium (Spiegel 1990)) and Dictyos-

telium (Baldauf and Doolittle 1997). Spiegel et al

(1979) suggested that the similarities during culmi-

nation among stalk tube-synthesizing cells of dic-

tyostelids and protostelids indicated a shared evo-

lutionary history as well. These studies have lent

considerable support to Olive’s (1975) hypothesis

that a protostelid-like ancestor gave rise to Dictyos-

telids (F IG . 2).

In this paper, we use formal phylogenetic analysis

to investigate whether the traditional taxonomic

characters impart any information about the evolu-

tionary relatedness of 19 members of the Dictyoste-

liales in order to determine if those characters sup-

port the current classiﬁcation of two families and

three genera. We test two hyphotheses that are con-

MATERIALS AND METHODS

A data matrix containing 18 characters was constructed for

1 protostelid outgroup ( Protostelium mycophaga ) and 19 in-

group taxa (T ABLE II). Characters were drawn from the tax-

onomic literature and chosen according to their universality

among members of the Dictyosteliales. Character coding

was made according to the taxonomic works of Raper

(1984), Hagiwara (1989), Olive (1975), and original pub-

lished species descriptions. The 19 dictyostelid taxa were

chosen to cover the range of morphological and develop-

mental diversity found in the roughly 65 described species

(Swanson et al 1999). The ﬁnal matrix was analyzed using

PAUP* v 4.0b5 for Macintosh (Swofford 1999) applying

branch and bound methods for maximum parsimony, with

characters deﬁned as unordered and with equal weights.

Unrooted strict and 80% majority rule consensus trees were

constructed.

RESULTS

Thirty-six equally parsimonious trees were generated,

each with 58 steps. The strict and 80% majority rule

consensus trees generated from these data (F IG .3 A ,

B ) do not support the hypotheses that either family

of the Dictyosteliales is monophyletic, or that any of

the three genera is monophyletic.

An important feature to note is that Dictyostelium

lacteum is always positioned basal to a clade that con-

tains both Acytostelium ellipticum and all of the dic-

tyostelids with cellular stalks.

Using strict consensus, there is no support for a

single clade that contains all of Polysphondylium , al-

though the 80% consensus tree lends some support

for a monophyletic group containing the white-

spored species of Polysphondylium (F IG .3 B ).

DISCUSSION

The trees generated from the present character set

do not support the classical arrangement of fami-

lies into Acytosteliaceae and Dictyosteliaceae, nor

do they support a monophyletic genus, Polysphon-

dylium .

The current taxonomy of dictyostelids implies

T ABLE II. Character data matrix for 19 dictyostelid taxa and 1 outgroup taxon

Characters b

Taxon a

Polar

gran-

ules

Spore

shape

Micro-

cysts

Macro-

cysts

Type of

acrasin

Sorocarp

branch-

ing

Relative

sorocarp

size

Slug

behav-

ior

Growth

habit

Sorus/

Spore

pigment

Stalk

base

shape

Stalk

tip

shape

Photo-

tropism

Aggrega-

tion

type

Stalk

cellulari-

Nucleo-

lus

position

Stalk

pigment

Whorled

branch-

ing

Daurs

Ddemi

Ddisc

Dlact

Dlate

Dmacr

Dmexi

Dminu

Dmuco

Dpoly

Dpurp

Drhiz

Drosa

Pﬁla

Ppall

Pviol

Aelli

Alept

Asubg

Pmyco

a Taxa: Daurs

5 Dictyostelium aureostipes , Ddemi

5 D. deminutivum , Ddisc

5 D. discoideum , Dlact

5 D. lacteum , Dlate

5 D. laterosorum , Dmacr

5 D. macrocephalum ,

Dmexi

D. mexicanum , Dminu

D. minutum , Dmuco

D. mucoroides , Dpoly

D. polycephalum , Dpurp

D. purpureum , Drhiz

D. rhizopodium , Drosa

rosarium , Pﬁla

5 Polysphondylium ﬁlamentosum , Ppall

5 P. pallidum , Pviol

5 P. violaceum , Aelli

5 Acytostelium ellipticum , Alept

5 A. laptosomum , Asubg

5 A. subglobosum ,

Pmyco

5 Protostelium mycophaga .

b Character State Codes: ?

unknown; Polar Granules absent

0, unconsolidated

1, consolidated

2; Spore Shape spherical

0, elliptical/oblong

Microcysts absent

0, present

1; Macrocysts absent

0, present

1; Acrasin absent

0, cAMP

1, pterin der.

2, glorin

3, folic acid der.

4; Sorocarp

branching absent

0, irregular

1, regular

2; Sorocarp size small (0.1–2.0 mm)

0, medium (2.1–4.5 mm)

1, large (4.6–10 mm)

2; Slug n/a

0, absent

1, migrating with stalk

2, stalkless migrating

3; Growth habit solitary

0, clustered/gregarious

1, coremiform

2; Sorus/spore pigment absent

0, orange

1, yellow

2, blue/brown/lavender

3; Stalk base simple

0, digitate

1, discoid

2; Stalk tip simple

0, compound

1; Phototropism absent

0, present

1; Aggregation absent

0, microsporum-type

1, minutum-type

2, mucoroides-type

3, violaceum-type

4; Stalk acellular

0, cellular

1; Amoebal nucleolus

central

0, peripheral

1; Stalk pigment absent

0, orange

1, yellow

2, blue/brown/lavender

3; Branches non-whorled

0, whorled

972

M YCOLOGIA

F IG . 3. A. Strict consensus of 36 most parsimonious trees. B. 80% Majority rule consensus of 36 most parsimonious trees;

numbers indicate percent of trees that support the topology. (see T ABLE II, footnote ‘a’ for species abbreviations)

that cellular stalk arose once, and is a synapomorphy

of the Dictyosteliaceae, and that whorled branching

arose once, and is a synapomorphy of the genus Po-

lysphondylium . These have been unquestioned as-

sumptions in all published speculation on the phy-

logeny of the dictyostelids. Four possible phyloge-

netic arrangements are consistent with the hypoth-

eses implied by the current taxonomy (F IG .4 A – D ).

At one extreme (F IG .4 A ), all three genera and both

families are monophyletic. At the other extreme, the

only monophyletic genus is Polysphondylium and the

only monophyletic family is the Dictyosteliaceae

(F IG .4 D ). In one intermediate tree, Acytostelium and

Polysphondylium are monophyletic, and both fami-

lies are monophyletic (F IG .4 B ). In the other inter-

mediate tree, the Acytosteliaceae and Acytostelium

are paraphyletic, the Dictyosteliaceae is monophy-

letic, and within the Dictyosteliaceae, both Dictyos-

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: