Hereditas 143: 84 90 (2006)
Escape from an evolutionary dead end: a triploid clone of Gyrodactylus
salaris is able to revert to sex and switch host (Platyhelminthes,
Monogenea, Gyrodactylidae)
MAREK S. ZIE˛TARA1,3, JUSSI KUUSELA2,3 and JAAKKO LUMME3
1
Gdańsk University Biological Station, Laboratory of Comparative Biochemistry, Gdańsk-Sobieszewo, Poland
National Veterinary and Food Research Institute, Oulu Regional Unit, Finland
3
Department of Biology, University of Oulu, Oulu, Finland
2
Zie˛tara, M. S., Kuusela J. and Lumme, J. 2006. Escape from an evolutionary dead-end: a triploid clone of Gyrodactylus
salaris is able to revert to sex and switch host (Platyhelminthes, Monogenea, Gyrodactylidae). * Hereditas 143: 84 90.
Lund, Sweden. eISSN 1601-5223. Received April 18, 2006. Accepted May 2, 2006
Diploid parthenogenesis, with rare sex, is considered as the basic mode of reproduction among the hermaphroditic and
viviparous Gyrodactylus. A particular strain of the monogenean parasite Gyrodactylus salaris (RBT clone) was recognized
by an invariable, unique mitochondrial DNA haplotype in rainbow trout (Oncorhynchus mykiss ) farms. The RBT clone was
shown to be triploid and asexual by analyzing a 493 bp sequence of a nuclear DNA marker. Three alleles were present as
heterozygous in all 237 individuals sampled in years 2001 2005 from five isolated Finnish farms. The triploid clone probably
originated from a diploid oocyte fertilized by a non-self hermaphrodite, most probably in a fish farm. Identical
mitochondrial COI gene (1606 bp) was also found in G. salaris parasites on landlocked salmon (Salmo salar ) in two
rivers draining to the lake Kuitozero, Russian Karelia. In the river Pisto, the clone was triploid, but the diagnostic ‘‘short’’
nuclear allele of the RBT clone was replaced by an allele typical for salmon specific parasites in the Lake Onega. The clone in
the river Kurzhma was diploid, having lost the ‘‘short’’ allele, but still heterozygous for the other two alleles of the RBT clone.
Evidently, the triploid parthenogenetic RBT clone had produced diploid oocytes, when (as a female) stimulated by a non-self
mate in the new environment. The genetic reorganization coincided with a switch to the salmon host. Participation of
triploids into the gene pool of the species is rarely reported in animals, and the triploidy is generally considered as an
irreversible dead-end of the evolution. Liberalism in ploidy level may significantly add to the evolutionary options available
for a parasite in ever-changing environments.
Marek Zie˛tara, Department of Biology, University of Oulu, POB 3000, FIN-90014 Oulu, Finland. E-mail: zietara@biotech.
univ.gda.pl
The role of sex in evolution is a rich source of theory
and speculation. There are less asexual than sexual
organisms, and a mixed strategy seems to be optimal.
Several animals exploit seasonal or otherwise adaptive
cyclic parthenogenesis, being then usually diploids.
However, many forms of parthenogenesis are correlated with elevated ploidy, and then the clones are
usually considered evolutionary dead-ends: the meiosis of triploids never succeeds, and the clone is
ultimately excluded from the gene pool and further
evolution (BELL 1982; SUOMALAINEN et al. 1987).
Our interest in the topic originates from the other
direction. While attempting to solve a question:
why are there so many speciation-associated host
switching events in the evolution of the freshwater
fish parasite genus Gyrodactylus, we arrived to the
question of the roles of parthenogenetic propagation
and sexual recombination. Special emphasis is put on
pathogenic salmon parasite Gyrodactylus salaris. Very
few Gyrodactylus species have been studied in
detail with respect to the reproductive mode. All
Gyrodactylus are viviparous, giving birth to a fullgrown progeny, which is already ‘‘pregnant’’. Most, if
not all species are hermaphroditic, but the first
progeny is expected to be produced asexually, as
a twin of the mother, and the subsequent few, by
automictic parthenogenesis (CABLE and HARRIS
2002).
Gyrodactylus salaris is a polytypic species, in
the sense that the species is difficult to define. The
internal transcribed spacer (ITS) regions of the
nuclear ribosomal genes are nearly uniform in parasites on different hosts, grayling (Thymallus thymallus ), salmon (Salmo salar ), Arctic char (Salvelinus
alpinus ) and rainbow trout (Oncorhynchus mykiss )
(CUNNINGHAM 1997; CABLE et al. 1999; ZIE˛TARA and
LUMME 2002; ZIE˛TARA et al. 2002; LINDENSTRØM
et al. 2003). By analyzing mitochondrial DNA
sequences (MEINILÄ et al. 2002), Gyrodactylus salaris
was divided into seven evolutionary lineages or clades
(HANSEN et al. 2003, 2004; MEINILÄ et al. 2004). Five
of the mtDNA clades were specific for grayling
(Thymallus thymallus ). One variable mitochondrial
clade of G. salaris was found to be specific to
Baltic salmon and native in the entire Baltic Basin.
Several identifiable strains belonging to this clade
Hereditas 143 (2006)
were causing the Atlantic Ocean and White Sea
gyrodactylosis epidemics (MEINILÄ et al. 2004).
The seventh mitochondrial clade of G. salaris was
different from the others, by being monomorphic with
respect to the mtDNA, even though it had diverged
from the assumed common ancestor as early as the
other lineages (MEINILÄ et al. 2004). All other
mtDNA clades were variable, reflecting the accumulation of mutations since the glacial displacement of the
hosts and the switch to salmon. The monomorphic
mitochondrial clade was common in rainbow trout
farms in Finland, Denmark and Sweden, but it was
also found in some populations of salmon: in Lierelva,
Drammen and Laerdalselven in Norway (HANSEN
et al. 2003), and in the Pistojoki river (lake Kuitozero,
Russian Karelia), where it was suggested to be
introduced via rainbow trout farms (MEINILÄ et al.
2004).
Here, we put forward a hypothesis about the origin
of this fish farm specific strain (from hereafter, RBT
clone) of G. salaris. The clone was observed to be
triploid, which is the first case among Gyrodactylus.
We also analyzed the derived parasite populations on
salmon in Russian Karelia, carrying identical mitochondrial haplotype, and demonstrated that the
triploid RBT clone has been able to reversion and
had produced diploid eggs. Asexual triploids are
usually dead ends in the evolution, but the RBT clone
has been capable to recombination. The observations
also contribute to our understanding of the mode of
speciation in this genus. The history of RBT clone
contains host switching events. The original switch to
rainbow trout included the triploidization, and the
two derived clones found on salmon were produced by
genomic reorganization via diploid eggs.
MATERIAL AND METHODS
The rainbow trout parasites studied were from five fish
farms in northern Finland and collected during a
study on the temporal dynamics of the parasite
infection in years 2001, 2003 and 2005. Because the
Gyrodactylus salaris parasites from all farms were
genetically identical (n/237), the exact locations and
names of the fish farming companies are omitted. For
a detailed account of the distribution of the parasite
G. salaris (and G. lavareti ) in salmonid farms in
northern Finland, the reader is referred to KOSKI
and MALMBERG (1995).
In Russian Karelia, the salmon parasites in two
spawning rivers of the landlocked salmon population
in Kuitozero were studied. In 2001, the parasites from
the river Pisto were analyzed by mtDNA and they
were observed to belong to the rainbow trout specific
Escape from an evolutionary dead-end
85
mitochondrial clade (MEINILÄ et al. 2004). A known
upstream fish farm in Finnish side of the border
(KOSKI and MALMBERG 1995) was concluded as the
source of infection. In 2005, the two spawning rivers
Kurzhma and Pisto were sampled and the parasites
analyzed by the new nuclear marker in addition to
mtDNA.
Comparative data of nuclear markers was obtained
from a Gyrodactylus salaris clone in the Baltic river
Tornio, where the parasite is native and nearly
harmless.
Molecular methods
Single specimens of G. salaris were digested in a
solution consisting of 0.45% Tween, 0.45% Igepal and
60 mg ml1 proteinase K in 1/concentrated PCR
buffer. Digestion was carried out in a PCR machine
for 25 min at 658C followed by inactivation of
Proteinase K for 10 min at 958C and cooling at 48C.
Two ml of such digestions were used as templates for
PCR amplification.
The partial sequence of the mitochondrial cytochrome c oxidase subunit I (COI) gene of G. salaris
was studied like in MEINILÄ et al. (2002, 2004).
However, the length of the deposited sequence is
now 1606 base pairs (accession number of the RBT
clone mtDNA AF540750, the Pisto salmon clone
AF479750 and the Kurzhma salmon clone
DQ517533). The fragment was amplified with newly
designed primers: forward FCox5 (5?-TACTTTGGATCATAAGCGCAT-3?) and reverse 16SR (5?CATTTAATCATGATGCAAAAGG-3?). The 20 ml
PCR reaction consisted of 1/PCR buffer, 0.2 mM
dNTPs, 2 mM MgCl2, 1 mM of each primer, 0.5 U of
Fermentas’ Taq DNA polymerase and 2 ml of template
prepared as described above. The PCR program was
run as follows: 948C for 3 min, 38 cycles (948C for 40 s,
508C for 30 s and 728C for 1.5 min), 728C for 7 min,
and final cooling to 48C.
A nuclear single copy DNA marker was studied
utilizing four sets of primers. In initial experiments we
used the degenerate primers designed to amplify the
G. salaris b-tubulin gene (COLLINS et al. 2004): BT1
(5?-RTYCTWGTBGAYCTHGAACC-3?) and BT2R
(5?-GRATCCABTCNACRAARTAA-3?). In addition
to the target sequence (not considered here), these
primers weakly amplified two extra fragments of 552
and 529 bp long nuclear DNA in farm strains of
G. salaris. To improve the amplification, new internal
primers were designed, InsF (5?-GATCTGCAATTCATCCTAAT-3?) and the reverse InsR (5?-TACAATTCGACCAAGGGTAG-3?). They amplify shorter
pieces (493 and 470 bp) of the same DNA fragments.
We also designed selective forward primers to amplify
86
M. S. Zie˛tara et al.
the ‘‘long’’ and ‘‘short’’ allele separately: FBl (5?GCCCTTCCTATAACTCAAAAGAT-3?) and FBs
(5?-GCCCTTCCTATAACTCAAAAGAA-3?). These
primers, combined with the reverse InsR primer,
amplify 435 and 412 bp fragments of ‘‘long’’ and
‘‘short’’ alleles respectively, which still span all the
nucleotide variation observed in the original fragment.
The 20 ml reactions consisted of 1 / PCR buffer,
0.2 mM dNTPs, 2 mM MgCl2, 1 mM of each primer,
1 U of Fermentas’ DNA Taq polymerase and 2 ml of
template prepared as above. The PCR program was
run as follows: 948C for 3 min, 42 cycles (948C for 40 s,
508C for 30 s and 728C for 1 min), 728C for 7 min, and
cooling to 48C.
Finally, for screening a large number of farm
parasite specimens we developed primers to amplify
a short segment of the DNA marker to be able
to separate the size difference between the ‘‘short’’
and ‘‘long’’ alleles after primary PCR. These primers
are: forward FB (5?-GCCCTTCCTATAACTCAAAA-3?) and reverse RB (5?-CTGAACGAGACCGAAAAATG), and the bimodal product is /200 bp
long (Fig. 1). For 20 ml PCR reaction, the ingredients
are as follows: 1 / PCR buffer, 0.2 mM dNTPs, 2 mM
MgCl2, 1 mM of each primer, 0.5 U of Fermentas’
DNA Taq polymerase and 2 ml of template prepared
as described above. The program: 948C for 3 min,
30 cycles (948C for 40 s, 508C for 30 s and 728C
for 30 s), 728C for 7 min, and cool it to 48C. The
PCR product is enough to visualize the long and
short alleles on 2% agarose gel (Fig. 1). This procedure thus identifies the triploid RBT clone of
G. salaris conveniently by one step PCR and agarose
electrophoresis.
To confirm the genotyping made by selective and
specific primers, all alleles were cloned after PCR by
proofreading Phusion polymerase by Novagen
pSTBlue-1 Perfectly Blunt† cloning kit following the
Hereditas 143 (2006)
manufacturer’s instructions. From five to ten positive
colonies of each nuclear haplotype were sequenced.
Sequencing was based on Big Dye Terminator Cycle
Sequencing kit protocol and the ABI 3730 DNA
Analyzer (Applied Biosystems, Hitachi, Japan).
RESULTS
A new nuclear single copy DNA marker
The primers developed for nuclear marker amplified
supposedly noncoding, but non-repetitive DNA, with
CG content as low as 33.9%. The segment showed no
BLAST hits longer than /15 bp. It was confirmed
that the PCR product was not from fish tissues.
The nuclear marker was named as anonymous DNA marker 1 (ADNAM1). For naming
the alleles, the variable sites were utilized (Table 1),
and the phenotypes can be named by applying the
system used for ambiguous nucleotides in gene
sequencing. Thus, the phenotype of RBT farm
clone is TMRCRWWT, the phenotype of Pisto
clone is TMRCRWAT, and that of Kurzhma,
TMRCAWAT. The haplotype (allele) sequences were
deposited in GenBank under accession numbers:
RBT farm clone (DQ436477-DQ436479) and Pisto
(DQ471291-DQ471293). The Kurzhma phenotype
was identical to Tornionjoki phenotype already deposited in GenBank (DQ468129-DQ468130). In Table
1, the haplotype names needed for GenBank are also
shown (the list of variable nucleotides is not a valid
name for a GenBank entry). These names (WS1, BS1,
BS4, BS8) were not used here for any other purposes.
The four alleles of the ADNAM1 locus found in
Gyrodactylus salaris in this study are described in the
Table 1. They were characterised by five variable
nucleotides sites, one deletion of 23 bp, and also by
the selective primers BT1 and BT2R of COLLINS et al.
Fig. 1A-B. ADNAM1 locus amplified with primers FB and RB, designed to
separate the ‘‘short’’ and ‘‘long’’ alleles, and to diagnose the RBT clone by one
step PCR. The size of the DNA bands in the 2% agarose gel are 191 bp and 168 bp.
(A) Lane 1: MW standard 191 bp. Lanes 2 10: direct PCR by FB and RB of the
diploid G. salaris from river Tornio. Lanes 11 20: triploid G. salaris from rainbow
trout farm, amplified first by selective "old" primers and then secondarily by FB and
RB. (B) Direct PCR of RBT clone by FB and RB primers. Lane 1: MW standard
191-168 bp, lanes 2 20: triploid G. salaris. The long band is about twice as strong
as the short.
Escape from an evolutionary dead-end
Hereditas 143 (2006)
87
Table 1. Variable sites of the four ADNAM1 alleles in rainbow trout farm type and in the clones in Kuitozero of
Gyrodactylus salaris. The numbering is based on the long sequences deposited in GenBank. The invariable
nucleotides in the allele names are reserved for predictable future expansion of the data.
Allele name
old
TAGCAAAT
TCACGTATold
TCACATATold
del23 -TCACGTTTold
56 78
214
245
302
316
317
old*
Accession no.
GenBank name
/
/
/
del
A
C
C
C
G
A
A
A
A
G
A
G
A
T
T
T
A
A
A
T
/
/
/
/
DQ436479
DQ471292
DQ436478
DQ436477
WS1
BS1
BS4
BS8
*old/indicates that the allele is amplified by primers of COLLINS et al. (2004).
(2004), which amplify only one of the ‘‘long’’ alleles in
the RBT clone, and the ‘‘short’’. This allowed us
to sequence the two ‘‘long’’ alleles separately, leading
to the haplotype (allele) description presented in
Table 1. We also cloned all the alleles several times.
We have named the alleles so that the system will be
consistent with further observations concerning the
predicted wider collection of strains of G. salaris. In
this system, the ‘‘long’’ alleles of the RBT clone
are TCACATATold and TAGCAAATold, and the
‘‘short’’ allele del23 -TCACGTTTold. The new allele
found in the river Pisto clone is TCACGTATold.
Rainbow trout specific Gyrodactylus salaris
By utilizing the FB and RB primers for PCR, and
electrophoresis in 2% agarose, we screened 237 Gyrodactylus salaris specimens from five different rainbow
trout farms in northern Finland. The samples were
from years 2001, 2004 and 2005.
When applied after primary amplification by BT1
and BT2R, the secondary PCR by FB RB pair
produced two bands 191 bp and 168 bp of equal
brightness (Fig. 1A) from all farm specimens.
When FB and RB primers were used for the primary PCR, rainbow trout farm parasites invariably
produced two electrophoretic bands, the slower one
(191 bp) appearing much brighter as the faster
(168 bp). All G. salaris specimens studied from five
different rainbow trout farms were identical in the
non-selective short fragment test (Fig. 1B). Parasites
from the river Tornio population always produced
only the longer segment (Fig. 1A).
We analyzed sequences from 22 farm specimens of
G. salaris. When sequenced from the 3?end until the
deletion, the direct sequencing of the triploid clone
produced a phenotype TMRCRWWT.
The two ‘‘long’’ alleles (Table 1) were also found in
the parasites on wild Baltic salmon from the Tornio
river. The directly sequenced nuclear phenotype of
the Tornio mitochondrial clade Tornio TA was
invariably TMRCAWAT (n/46), and the separated
alleles TCACATATold and TAGCAAATold were
also cloned.
The salmon parasites in the Lake Kuitozero, Russian
Karelia
In the spawning rivers Pisto and Kurzhma, of
the landlocked salmon of Lake Kuitozero, Russian
Karelia, Gyrodactylus salaris was first observed in
2001. The rivers were sampled again in 2005. In the
river Pisto, out of 40 salmon, 14 carried parasites
(prevalence 35%), between 6 to 39 individual worms
per fish. In Kurzhma, the infection by G. salaris was
even weaker (13 worms on 6 out of 38 fish), and the
salmon juveniles carried also some specimens of
Gyrodactylus aphyae (N /2) and G. lucii (N /2),
parasites of minnow and pike, respectively.
The 1606 bp long mitochondrial sequences of the
Gyrodactylus salaris specimens in both rivers were
exactly identical to that of the RBT clone (n/4).
However, the nuclear ADNAM1 of G. salaris in
Kurzhma (n /6) was not identical with RBT clone.
The phenotype in Kurzhma was TMRCAWAT, and
the genotype was ascertained to be TCACATATold/
TAGCAAATold. The ‘‘short’’ allele of the RBT
clone was not present. The ambiguous peaks of the
variable sites in direct sequencing were of equal
intensity, not suggesting cryptic triploidy. We have to
conclude that the clone was diploid.
In the river Pisto, the G. salaris phenotype read by
direct sequencing was TMRCRWAT (n /13), and the
genotype consisted of three alleles (TCACGTATold/
TCACATATold/TAGCAAATold). Two alleles of
this triploid clone were the same as in the RBT clone
and in Kurzhma, but the allele TCACGTATold was a
new one, not found in rainbow trout farms.
DISCUSSION
Systematically, G. salaris belongs to the subgenus
Limnonephrotus (MALMBERG 1970, ZIE˛TARA and
LUMME 2004), and to the species group wageneri,
characterized by frequent long distance host switches
88
M. S. Zie˛tara et al.
to new fish families (ZIe˛TARA and LUMME 2002). The
mechanism of this innovative host switching ability
deserves an explanation. The results of this paper may
offer a key to explaining some parts of this: the RBT
clone on a non-native host was observed to be triploid
hybrid and supposedly asexual. When the maternal
lineage had switched back to salmon in the Lake
Kuitozero, we could demonstrate a temporary reversion to sexuality, and one derived clone was diploid. Is
it possible that the whole wageneri group is a flock of
apparently asexual clones, which use sex when there is
an opportunity to real recombination, and adaptive
novelty?
The rainbow trout farm clone of Gyrodactylus salaris is
different?
The Gyrodactylus specialists MALMBERG and MALM(1986) and MO (1991) (reviewed recently in
CUNNINGHAM et al. 2003) have always maintained
that the rainbow trout farm type of G. salaris can be
separated from other strains by morphology or
morphometrics. As long as the type was recognized
only in rainbow trout farms, this suggestion remained
both true and unfalsifiable. The results of this
study support this idea, because the RBT clone
observed is monoclonal and invariable.
Here we demonstrated that the rainbow trout farm
specific RBT clone of G. salaris is a triploid monophyletic clone, carrying a specific mtDNA. Among
the diverse forms of reproduction among animals,
it is usual that triploids are asexuals due to unavoidable meiotic problems (BELL 1982; SUOMALAINEN
et al. 1987).
Currently, we do not know, where is the mother of
the RBT clone. We do not know the father either,
but it is probably to be found among the parthenogenetic diploid Baltic G. salaris strains. Both long
ADNAM1 alleles of the RBT clone were found in
Tornio strain of G. salaris (Kuusela et al., unpubl.).
The monomorphic mitochondrial haplotype specific
for the RBT clone was also found in parasites on
salmon populations in Drammen, Lierelva, and Laerdalselven in Norway (HANSEN et al. 2003). In the
results of CUNNINGHAM et al. (2003), the IGS of the
parasites from Laerdalselva and Lierelva clustered
with the haplotypes from the main G. salaris epidemic,
not showing the diversity of IGS molecular clones
observed in the parasites from rainbow trout farm.
This makes the Laerdalselva and Lierelva G. salaris
the prime suspects as the mother. Using the marker
developed here the maternity case could possibly be
solved if the real parental strains can be found and
included among the suspects. Because the Norwegian
BERG
Hereditas 143 (2006)
parasite strains were not available for this study, the
question remains open.
Origin of Gyrodactylus salaris in Kuitozero. A virgin
fertilized?
While we have to leave the hypothesis on the origin
of the RBT clone await testing of the suspected
mothers, we succeeded to detect an unsolved paternity case among the progeny of RBT clone. The
mitochondrial DNA of the Gyrodactylus salaris in
both spawning rivers of Kuitozero showed that the
parasites were maternally derived from the RBT
clone. This was reported already from the 2001
observations in Pisto (MEINILÄ et al. 2004). However,
the nuclear DNA analysis here showed that the
parasites in Pisto and Kurzhma were not specimens
of original farm clone. While the farm type was
triploid and carries alleles [TAGCAAATold/
TCACATATold/(del23 -TCACGTTTold], the direct
sequencing of Kuitozero parasites gave a good
sequence phenotype TMRCAWAT from both directions, indicating that the allele with the 23 bp deletion is not there. Further analysis by selective primers
and molecular cloning revealed that the Pisto parasite was triploid [TCACATATold/TCACGTATold/
TAGCAAATold], and Kurzhma most probably
diploid [TCACGTATold/TAGCAAATold].
Consequently, it is obvious that the RBT clone
is the mother of the Kuitozero parasites. We suggest
the following a posteriori scenario. The RBT
clone of G. salaris arrived to Kuitozero from an
upstream fish farm in Finnish side of the border
(KOSKI and MALMBERG 1995). In Kuitozero, it
mated with a hypothetical native (or introduced) G.
salaris strain TMRCRWAT (/TCACGTATold/
TAGCAAATold), giving rise to the triploid type
now found in Pisto. This suspected father genotype is
found in the Lake Onega and in the river Keret, White
Sea (Kuusela et al., unpubl.).
Triploids are usually expected to be fecund but
sterile. Among 237 specimens collected from five
disconnected farms during three years, no segregating
individuals were found. This sampling represents a
rather large number of generations: the generation
length of Gyrodactylus salaris is estimated to be 4 to 8
days, depending on season (CABLE and HARRIS 2002).
So, either the segregation never occurs, or all segregated progeny is lethal. While the RBT clone is
monoclonal, arising from a single mating event, it
has good reasons to avoid sexual reproduction, as long
as it is alone. The heterozygosity obviously is advantageous, and it can best be maintained by self-incompatibility (PECK and WAXMAN 2000). But the move from
farm to a new lake may have offered an alternative.
Hereditas 143 (2006)
Currently, we have not recorded suitable mates in
Kuitozero (by extremely limited faunistic sampling).
In Pisto, there is G. salaris in grayling (MEINILÄ
et al. 2004), but in upstream waters we have observed only homozygous TAGCAAATold phenotypes (unpubl.). The G. salaris strains in the lake
Onega and in the river Keret fulfill the characteristics
of the ‘‘suspected father’’, having the allele TCACGTAT old needed to transform the RBT clone to the
strain in Pisto (Kuusela et al., unpubl.). However,
Kuitozero is biogeographically on the White Sea basin,
and G. salaris strains specific to salmon have never
been observed as native in this basin (MEINILÄ
et al. 2004).
The possibility remains that the hybridization of
RBT clone with the paternal strain has occurred
elsewhere than in the lake Kuitozero.
The origin of diploid Kurzhma strain from the
triploid mother in the above situation is not unexpected. If the triploid female was induced to produce
diploid oocytes, to be fertilized by a compatible nonself hermaphrodite, then some of the diploid oocytes
may well have utilized the default developmental
program of Gyrodactylus, and developed further
without fertilization
In their comprehensive review of reproduction in
Gyrodactylus, CABLE and HARRIS (2002) mentioned
that there are no known polyploids in Gyrodactylus.
The RBT clone of G. salaris is now the first triploid
observed. Considering the general systematic distribution of polyploidy and parthenogenesis among the
animals (SUOMALAINEN et al. 1987), we assume that
there will be more diversity in the sexuality and
reproductive systems of Gyrodactylus than has been
described.
The escape from triploid asexuality to other forms
of reproduction has been suggested in plants, e.g.
in dandelions (Taraxacum ) (MENKEN et al. 1995),
but it is very rare in animals. In planarian Schmidtea
polychroa, D’SOUSA et al. (2004) observed alternation of triploidy and tetraploidy, including haploid
sperm.
Acknowledgements Pasi Anttila and Perttu Koski helped in
obtaining the G. salaris samples from fish farms. The
Kuitozero expedition was conducted with Alexei Veselov
and Dimitry Stepanov. Hannele Parkkinen-Oinas assisted by
DNA sequencing. Financial support from the Finnish
Academy is appreciated.
REFERENCES
Bell, G. 1982. The masterpiece of nature.. In: The
evolution and genetics of sexuality. Univ. of California
Press.
Escape from an evolutionary dead-end
89
Cable, J. and Harris, P. D. 2002. Gyrodactylid developmental
biology: historical review, current status and future
trends. Int. J. Parasitol. 32: 255 280.
Cable, J., Harris, P. D., Tinsley, R. C. et al. 1999.
Phylogenetic analysis of Gyrodactylus spp. (Platyhelminthes: Monogenea) using ribosomal DNA sequences.
Can. J. Zool. 77: 1439 1449.
Collins, C. M., Miller, K. A. and Cunningham, C. O. 2004.
Characterization of a (-tubulin gene from the monogenean parasite, Gyrodactylus salaris Malmberg, 1957.
Parasitol. Res. 92: 390 399.
Cunningham, C. O. 1997. Species variation within the
internal transcribed spacer (ITS) region of Gyrodactylus (Monogenea: Gyrodactylidae) ribosomal RNA
genes. J. Parasitol. 83: 215 219.
Cunningham, C. O., Collins, C. M., Malmberg, G. et al.
2003. Analysis of ribosomal RNA intragenic spacer
(IGS) sequences in species and population of Gyrodactylus (Platyhelminthes: Monogenea) from salmonid fish in
northern Europe. Dis. Aquat. Org. 57: 237 246.
D’Sousa, T. G., Storhas, M., Schulenburger, H. et al. 2004.
Occasional sex in an ‘asexual’ polyploid hermaphrodite.
Proc. R. Soc. Lond. B 271: 1001 1007.
Hansen, H., Bachmann, L. and Bakke, A. M. 2003.
Mitochondrial DNA variation of Gyrodactylus spp.
(Monogenea, Gyrodactylidae) populations infecting
Atlantic salmon, grayling and rainbow trout in Norway
and Sweden. Int. J. Parasitol. 33: 1471 1478.
Hansen, H., Olstad, K., Bakke, T. A. et al. 2004. Species,
subspecies or strains: the taxonomy of Gyrodactylus
salaris and G. thymalli (Monogenea) revisited. In:
Proc. IX Eur. Multicolloquium Parasitol. EMOP.
Medimond S.r.l., Valencia, Spain. p. 229 234.
Koski, P. and Malmberg, G. 1995. Occurrence of Gyrodactylus (Monogenea) on salmon and rainbow trout in
fish farms in Northern Finland. Bull. Scand. Soc.
Parasitol. 5: 76 88.
Lindenstrøm, T., Collins, C. M., Bresciani, J. et al. 2003.
Characterization of a Gyrodactylus salaris variant: infection biology, morphology and molecular genetics.
Parasitology 127: 165 177.
Malmberg, G. 1970. The excretory systems and the
marginal hooks as a basis for the systematics of
Gyrodactylus (Trematoda, Monogenea). Arkiv Zool.
23: 1 235.
Malmberg, G., and Malmberg, M. 1986. Gyrodactylus in
salmon and rainbow trout. In: Stenmark, A. and
Malmberg, G. (eds), Parasites and diseases in natural
waters and aquaculture in Nordic countries. Proc. ZooTax-Symp., Stockholm, Sweden. p. 199 203.
Meinilä, M., Kuusela, J., Zie˛tara, M. S. et al. 2002. Primers
for amplifying /820 bp of highly polymorphic mitochondrial COI gene of Gyrodactylus salaris. Hereditas
137: 72 74.
Meinilä, M., Kuusela, J., Zie˛tara, M. S. et al. 2004. Initial
steps of speciation by geographic isolation and host
switch in salmonid pathogen Gyrodactylus salaris
(Monogenea: Gyrodactylidae). Int. J. Parasitol. 34:
515 526.
Menken, S. B. J., Smit, E. and Den Nijs, H. C. M. 1995.
Genetical population structure in plants: gene flow
between diploid sexual and triploid asexual dandelions
(Taraxacum section Ruderalia ). Evolution 49: 1108 1118.
90
M. S. Zie˛tara et al.
Mo, T. A. 1991. Variation of opisthaptoral hard parts of
Gyrodactylus salaris Malmberg, 1957 (Monogenea:
Gyrodactylidae) on rainbow trout Oncorhynchus mykiss
(Walbaum, 1792) in a fish farm, with comments on
the spreading of the parasite in south-eastern Norway.
Syst. Parasitol. 20: 1 9.
Peck, J. R. and Waxman, D. 2000. What’s wrong with a little
sex? J. Evol. Biol. 13: 63 69.
Suomalainen, E., Saura, A. and Lokki, J. 1987. Cytology
and evolution in parthenogenesis. CRC Press.
Hereditas 143 (2006)
Zie˛tara, M. S. and Lumme, J. 2002. Speciation by host
switch and adaptive radiation in a fish parasite genus
Gyrodactylus (Monogenea: Gyrodactylidae). Evolution
56: 2445 2458.
Zie˛tara, M. S. and Lumme, J. 2004. Comparison of
molecular phylogeny and morphological systematics in
fish parasite genus Gyrodactylus Nordmann, 1832
(Monogenea, Gyrodactylidae). Zool. Pol. 49: 5 28.
Zie˛tara, M. S., Huyse, T., Lumme, J. et al. 2002. Deep
divergence among subgenera of Gyrodactylus inferred
from rDNA ITS region. Parasitology 124: 39 52.