• Bayesian phylogenetics;
  • breeding system evolution;
  • Conchostraca;
  • maximum likelihood phylogenetics;
  • mixed mating system


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Examinations of breeding system transitions have primarily concentrated on the transition from hermaphroditism to dioecy, likely because of the preponderance of this transition within flowering plants. Fewer studies have considered the reverse transition: dioecy to hermaphroditism. A fruitful approach to studying this latter transition can be sought by studying clades in which transitions between dioecy and hermaphroditism have occurred multiple times. Freshwater crustaceans in the family Limnadiidae comprise dioecious, hermaphroditic and androdioecious (males + hermaphrodites) species, and thus this family represents an excellent model system for the assessment of the evolutionary transitions between these related breeding systems. Herein we report a phylogenetic assessment of breeding system transitions within the family using a total evidence comparative approach. We find that dioecy is the ancestral breeding system for the Limnadiidae and that a minimum of two independent transitions from dioecy to hermaphroditism occurred within this family, leading to (1) a Holarctic, all-hermaphrodite species, Limnadia lenticularis and (2) mixtures of hermaphrodites and males in the genus Eulimnadia. Both hermaphroditic derivatives are essentially females with only a small amount of energy allocated to male function. Within Eulimnadia, we find several all-hermaphrodite populations/species that have been independently derived at least twice from androdioecious progenitors within this genus. We discuss two adaptive (based on the notion of ‘reproductive assurance’) and one nonadaptive explanations for the derivation of all-hermaphroditism from androdioecy. We propose that L. lenticularis likely represents an all-hermaphrodite species that was derived from an androdioecious ancestor, much like the all-hermaphrodite populations derived from androdioecy currently observed within the Eulimnadia. Finally, we note that the proposed hypotheses for the dioecy to hermaphroditism transition are unable to explain the derivation of a fully functional, outcrossing hermaphroditic species from a dioecious progenitor.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Elucidating the forces that select for a separation of the sexes (i.e. into pure males and pure females, termed dioecy) relative to a combination of the sexes (i.e. cosexuals or hermaphrodites) is imperative for understanding breeding system evolution (Charnov et al., 1976; Charlesworth & Charlesworth, 1978; Charlesworth, 1984; Schemske & Lande, 1985; Jarne & Charlesworth, 1993; Barrett, 2002; Wolf & Takebayashi, 2004). A useful approach to assessing these selective forces is to study clades in which transitions among breeding systems have occurred repeatedly (e.g. hermaphroditism to dioecy). Because numerous transitions from hermaphroditism to dioecy are evident in flowering plants (Weiblen et al., 2000; Barrett, 2002), a good deal of theory has been developed to explain the likely evolutionary progression of this transition (reviewed in Charlesworth, 2006). Direct evolution of dioecy from hermaphroditism is not predicted to occur, but rather one of two temporary breeding systems is thought to be a likely intermediate stage in this transition (Lloyd, 1975; Charlesworth & Charlesworth, 1978; Charlesworth, 1984): gynodioecy (mixtures of females and hermaphrodites) or androdioecy (mixtures of males and hermaphrodites). A gynodioecious intermediate is predicted to be more common than an androdioecious intermediate (Lloyd, 1975; Charlesworth, 1984), and indeed gynodioecy is much more common in flowering plants than is androdioecy (Charlesworth, 1984; Pannell, 2002; Delph & Wolf, 2005).

Because of the relative frequency of the transition from hermaphroditism to dioecy in flowering plants, the evolutionary steps in this transition have been predicted in some detail. Charlesworth & Charlesworth (1978) proposed a plausible genetic model for the evolution of dioecy from hermaphroditism which suggested that the most likely transition would include a gynodioecious intermediate. They proposed that a recessive male sterility gene could spread in a partially selfing hermaphroditic population experiencing moderate to high inbreeding depression, thus producing females and hermaphrodites (i.e. gynodioecy). They suggested that a second mutation of a dominant modifier that reduced female function in the hermaphrodites could then spread in the gynodioecious population. This second mutation would eventually reduce female function to zero, and thus transform the hermaphrodites into males, resulting in dioecy. The spread of this second mutation would be greatly facilitated if it was tightly linked to the first, recessive male sterility gene (Charlesworth & Charlesworth, 1978).

Although the transition from hermaphroditism to dioecy has been thoroughly explored, the reverse transition, from dioecy to hermaphroditism, has not received nearly the level of detailed attention. Ghiselin (1974) provided several verbal models (‘low-density’, ‘size-advantage’ and ‘gene-dispersal’) outlining possible benefits for deriving hermaphroditism from dioecy. Charnov (1982) also outlined the conditions favouring hermaphroditism over dioecy using the concept of ‘fitness sets’. However, neither author presented detailed outlines for how hermaphroditism could evolve from dioecy, and the notions of intermediate stages (e.g. androdioecy or gynodioecy) were never specifically considered.

The dearth of detailed discussions about a dioecy to hermaphroditism transition is not because such transitions are believed uncommon. Hermaphroditism is quite common in animals: when one excludes insects, up to one-third of animal species are hermaphroditic (Jarne & Charlesworth, 1993; Jarne & Auld, 2006). The distribution of hermaphroditism in animals is sporadic, with some higher taxa being primarily hermaphroditic (e.g. Platyhelminthes, pulmonate molluscs) and others having few hermaphroditic representatives (e.g. Echinoderms, Chordates; Ghiselin, 1974; Bell, 1982; Jarne & Charlesworth, 1993). Ghiselin (1974) has argued that the majority of these hermaphroditic animals are derived from dioecious ancestors (for an alternative perspective, see Eppley & Jesson, 2008; Lyer & Roughgarden, 2008), and thus these numerous species in disparate animal taxa suggest numerous dioecy to hermaphroditism evolutionary transitions. Therefore, understanding the details of the transition from dioecy to hermaphroditism should be quite important to those interested in the evolution of animal breeding systems.

One group of crustaceans, the Branchiopoda, displays a wide range of breeding systems (Sassaman, 1995; Dumont & Negrea, 2002): dioecy, androdioecy, hermaphroditism, parthenogenesis (i.e. asexual) and cyclic parthenogenesis (i.e. many rounds of parthenogenesis with a single episode of dioecy at the end of a growing season), and thus presents an opportunity to study many breeding system transitions within a single taxon. Because the basal clade in the Branchiopoda, the Anostraca (Negrea et al., 1999), is almost entirely dioecious, it appears that androdioecy, hermaphroditism, parthenogenesis and cyclic parthenogenesis all have evolved from dioecy (although not necessarily directly) in this group. In fact, all of these breeding systems are found in what were historically termed the ‘Conchostraca’ or ‘clam shrimp’ (the Conchostraca have been determined to be a polyphyletic group and thus it has now been split into the orders Laevicaudata and Diplostraca; Fryer, 1987; Spears & Abele, 2000; Braband et al., 2002). Sassaman (1995) outlined a scheme in which androdioecy, hermaphroditism and parthenogenesis could evolve (through a series of mutational steps) from a female-heterogametic, dioecious sex determining system (which Sassaman predicted to be the ancestral condition within the clam shrimp). Sassaman (1995) additionally predicted that cyclic parthenogenesis then evolved from parthenogenesis. Because of the breeding system diversity within clam shrimp, and because of our recent advances in understanding their biology and ecology, we believe this group presents an excellent opportunity to study the evolution of various breeding systems from a presumably dioecious ancestor.

Within the clam shrimp, one family, the Limnadiidae (Spinicaudata: Diplostraca), has three of the five above mentioned breeding systems: dioecy, hermaphroditism and androdioecy (Sassaman & Weeks, 1993; Sassaman, 1995; Weeks et al., 2008). The Limnadiidae contains five extant genera: Eulimnadia, Imnadia, Metalimnadia, Limnadia and Limnadopsis (Baird, 1849; Straskraba, 1964). Of these, Eulimnadia is the most speciose (containing over 40 species that inhabit every continent except Antarctica; Brtek, 1997) and is the best studied genus from a reproductive biology perspective (reviewed in Weeks et al., 2006a). In the current study, we will outline the breeding system transitions inferred from a DNA sequence/morphology-based phylogeny of the Limnadiidae. Although the ancestral breeding system for the Limnadiidae has been assumed to be dioecy (Sassaman, 1995) and a preliminary phylogeny was erected for the family (Hoeh et al., 2006), no ancestral character state reconstruction has been conducted to confirm or refute Sassaman’s assertion. Our analyses indicate that dioecy is indeed the ancestral state for the Limnadiidae and that both androdioecy and hermaphroditism are derived states within this family. We combine these insights on breeding system transitions with previously published information about these crustaceans to consider hypotheses regarding the processes underlying transitions from dioecy to androdioecy and hermaphroditism in the Limnadiidae.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Specimen collection/rearing

We examined 173 individuals from 42 species/lineages, 10 genera and three families; these samples were collected from six continents (Table 1). Specimens were either adults preserved in 95% ethyl alcohol or were reared from eggs in the laboratory. Samples were either collected by us or sent to us by colleagues. For each of the populations that were reared from eggs, we collected soil from natural, dried field sites. We made soil collections by sampling at many spots across the dried pools and then homogenizing the soil in plastic bags. Approximately 500 mL of this field-collected soil was placed in the bottom of a 37-L aquarium and hydrated with deionized water. The aquarium was maintained under ‘standard conditions’ (Weeks et al., 1997, 1999, 2001) of 25–28 °C, low aeration, constant light, and fed a mixture of baker’s yeast and ground Tetramin™ flake fish food (Tetra Werke, Melle, Germany) (2.5 g of each suspended in 500 mL of water). Shrimp were reared to sexual maturity (based on the presence of eggs in the brood chamber for females/hermaphrodites and presence of claspers in males) and then preserved in 95% ethanol or frozen in a −80 °C freezer for morphological and molecular analyses, respectively.

Table 1.   Specimen information.
FamilyGenusSpeciesID#28SEF1αCOIBSysCollection location
  1. GenBank accession numbers are shown for 28S, elongation factor 1-alpha (EF1α) and cytochrome c oxidase I (COI). ID#’s in bold were quantified for morphological characters.

  2. BSys, breeding system (A, androdioecy; D, dioecy; H, hermaphroditic; X, asexual).

CyzicidaeCyzicusgynecia (Mattox 1949)NS30AY851402  XUnited States: PA
gynecia (Mattox 1949)NS31AY851403FJ499036 XUnited States: PA
gynecia (Mattox 1949)NS36AY851404FJ499039 XUnited States: PA
gynecia (Mattox 1949)NS37AY851405FJ499040 XUnited States: PA
lutraria (Brady 1886)ZMUC CRU-9946EF189639EF189665EF189592DN.S.W., Australia
gifuensis (Ishikawa, 1895)ZMUC CRU-9947EF189640  DJapan
sp.W181FJ499303 FJ499176DWestern Australia
sp.W183FJ499304 FJ499177DWestern Australia
sp.W333FJ499305FJ499120 DWestern Australia
sp.W340FJ499306  DN. Terr., Australia
sp.W345FJ499307  DN. Terr., Australia
sp.W346FJ499308FJ499121FJ499221DSouth Australia
sp.W347FJ499309FJ499122FJ499222DSouth Australia
Eocyzicusdigueti (Richard 1895)NS52AY851406FJ499042FJ499133DBaja California
digueti (Richard 1895)NS53AY851407FJ499043FJ499134DBaja California
digueti (Richard 1895)W219FJ499300FJ499090FJ499188DUnited States: NM
digueti (Richard 1895)W220FJ499301FJ499091FJ499189DUnited States: NM
sp.W298FJ499302FJ499112FJ499213DSouth Australia
LeptestheridaeLeptestheriacompleximanus (Packard 1877)NS14AY851391FJ499032FJ499124DUnited States: NM
compleximanus (Packard 1877)NS15AY851392FJ499033FJ499125DUnited States: NM
compleximanus (Packard 1877)NS20AY851393 FJ499126DUnited States: NM
compleximanus (Packard 1877)NS32AY851395FJ499037FJ499129DUnited States: NM
compleximanus (Packard 1877)NS33AY851396FJ499038FJ499130DUnited States: NM
compleximanus (Packard 1877)NS39AY851398FJ499041FJ499131DUnited States: NM
compleximanus (Packard 1877)W214FJ499296FJ499085FJ499184DUnited States: NM
compleximanus (Packard 1877)W215FJ499297FJ499086FJ499185DUnited States: NM
dahalacensis (Rüppel, 1837)NS68AY851408FJ499044FJ499135DAustria
dahalacensis (Rüppel, 1837)NS69AY851409FJ499045FJ499136DAustria
dahalacensis (Rüppel, 1837)ZMUC CRU-9945EF189648EF189670AF526291DAustria
kawachiensis Uéno, 1927ZMUC CRU-9944EF189649  DJapan
sp.W217FJ499298FJ499088 DUnited States: NM
sp.W218FJ499299FJ499089FJ499187DUnited States: NM
LimnadiidaeEulimnadiaafricana (Brauer, 1877)W261DQ198215 FJ499195ABotswana
africana (Brauer, 1877)W285FJ499232FJ499104FJ499202 South Africa
africana (Brauer, 1877)W320FJ499233 FJ499220ABotswana
agassizii Packard, 1874W272FJ499242 FJ499198HUnited States: MA
agassizii Packard, 1874W278FJ499241 FJ499201HUnited States: MA
brasiliensis Sars, 1902W225DQ198203  ABrazil
brasiliensis Sars, 1902W228FJ499245  ABrazil
brasiliensis Sars, 1902W229DQ198204FJ499093 ABrazil
brasiliensis Sars, 1902W230FJ499246  ABrazil
braueriana Ishikawa, 1895NS40AY851425  AJapan
braueriana Ishikawa, 1895NS41AY851426 FJ499132AJapan
braueriana Ishikawa, 1895ZMUC CRU-9949EF189644EF189667EF189593 Japan
colombiensis Roessler 1989NS105AY851414FJ499048 HVenezuela
cylindrova Belk, 1989NS11AY851418  ABaja California
cylindrova Belk, 1989NS16AY851422  ABaja California
cylindrova Belk, 1989NS17AY851419  ABaja California
cylindrova Belk, 1989NS65AY851432  AGalapagos
cylindrova Belk, 1989NS79AY851440 FJ499138 Japan
cylindrova Belk, 1989NS80AY851442 FJ499139 Japan
cylindrova Belk, 1989NS103DQ198177   Venezuela
cylindrova Belk, 1989NS104AY851413   Venezuela
cylindrova Belk, 1989W147DQ198189 FJ499167 Martinique, FWI
cylindrova Belk, 1989W149DQ198188 FJ499168HDesirade, FWI
cylindrova Belk, 1989W204DQ198197  AJapan
cylindrova Belk, 1989W205DQ198198  AJapan
cylindrova Belk, 1989W269FJ499240FJ499101 AGalapagos
dahli Sars, 1896W101DQ198175 FJ499142HWestern Australia
dahli Sars, 1896W102DQ198176 FJ499143AWestern Australia
dahli Sars, 1896W103DQ198177  AWestern Australia
dahli Sars, 1896W106DQ198180 FJ499144AWestern Australia
dahli Sars, 1896W107DQ198181  HWestern Australia
dahli Sars, 1896W112DQ198182  HWestern Australia
dahli Sars, 1896W113DQ198183 FJ499148HWestern Australia
dahli Sars, 1896W115DQ198184 FJ499149HWestern Australia
dahli Sars, 1896W231DQ198205  AWestern Australia
dahli Sars, 1896W236DQ198207  AWestern Australia
dahli Sars, 1896W238DQ198208FJ499094 AWestern Australia
dahli Sars, 1896W240DQ198209FJ499095 AWestern Australia
dahli Sars, 1896W242DQ198210FJ499096 AWestern Australia
dahli Sars, 1896W246DQ198211  AWestern Australia
dahli Sars, 1896W296FJ499228FJ499111FJ499211HWestern Australia
dahli Sars, 1896W297FJ499229 FJ499212HWestern Australia
diversa Mattox, 1937NS8AY851441  AUnited States: AZ
diversa Mattox, 1937NS22AY851420  AUnited States: AZ
diversa Mattox, 1937NS23AY851421  AUnited States: AZ
diversa Mattox, 1937W132AY851455FJ499064 AUnited States: IN
diversa Mattox, 1937W223DQ198202  AUnited States: IL
diversa Mattox, 1937W258DQ198213  AUnited States: NE
diversa Mattox, 1937W259DQ198214  AUnited States: NE
diversa Mattox, 1937W276FJ499237 FJ499200 United States: FL
diversa Mattox, 1937W312FJ499234FJ499116FJ499216AUnited States: IN
diversa Mattox, 1937W317FJ499235FJ499119FJ499218AUnited States: MS
diversa Mattox, 1937W318FJ499236 FJ499219AUnited States: MS
follisimilis Pereira & Garcia 2001W321FJ499238  AUnited States: NM
follisimilis Pereira & Garcia 2001W322FJ499239  AUnited States: NM
magdaliensis Roessler 1990NS58AY851430   United States: MA
magdaliensis Roessler 1990NS59AY851431   United States: MA
magdaliensis Roessler 1990NS99AY851445FJ499047  Venezuela
michaeli Nayar & Nair 1968W348FJ499243FJ499123 HThailand
michaeli Nayar & Nair 1968W349FJ499244  HThailand
texana Packard 1871W280FJ499230FJ499102 AUnited States: NM
texana Packard 1871W281FJ499231FJ499103 AUnited States: NM
sp. 1W170DQ198190FJ499073 AUnited States: GA
sp. 1W209DQ198200  AUnited States: GA
sp. 1W252DQ198212  AUnited States: GA
sp. 1W253FJ499226  AUnited States: GA
sp. 2W293FJ499223FJ499109FJ499208 N. Terr., Australia
sp. 2W294FJ499224FJ499110FJ499209 N. Terr., Australia
sp. 2W315 FJ499117 AN. Terr., Australia
sp. 2W316FJ499225FJ499118FJ499217AN. Terr., Australia
sp. 3W274FJ499227 FJ499199 Japan
Imnadiayeyetta Hertzog 1935NS110FJ499254  DAustria
yeyetta Hertzog 1935W125AY851449FJ499059FJ499156DAustria
yeyetta Hertzog 1935W72FJ499255FJ499050FJ499141DAustria
yeyetta Hertzog 1935W128AY851446FJ499061FJ499159DAustria
yeyetta Hertzog 1935W129AY851450 FJ499160DAustria
yeyetta Hertzog 1935W130AY851447FJ499062FJ499161DAustria
yeyetta Hertzog 1935W131AY851448FJ499063FJ499162DAustria
yeyetta Hertzog 1935 EF189646EF189668AF526289 Austria
Limnadiabadia Wolf 1911W124FJ499256FJ499058FJ499155 Western Australia
badia Wolf 1911 W135 FJ499065FJ499163Western Australia
badia Wolf 1911W136FJ499257FJ499066FJ499164 Western Australia
badia Wolf 1911W144FJ499258FJ499068FJ499166DWestern Australia
badia Wolf 1911W158FJ499259 FJ499170 Western Australia
badia Wolf 1911W159FJ499267FJ499070FJ499171 Western Australia
badia Wolf 1911W161FJ499260FJ499071FJ499172DWestern Australia
badia Wolf 1911W250FJ499261 FJ499191 Western Australia
badia Wolf 1911W251FJ499262 FJ499192 Western Australia
cygnorum (Dakin 1914)W193FJ499271FJ499075  South Australia
cygnorum (Dakin 1914)W194FJ499272FJ499076  South Australia
lenticularis Linnaeus 1761NS24AY851399FJ499034FJ499127HUnited States: FL
lenticularis Linnaeus 1761NS25AY851400FJ499035FJ499128HUnited States: FL
lenticularis Linnaeus 1761W66AY851401  HUnited States: FL
lenticularis Linnaeus 1761W154FJ499279FJ499069FJ499169 Italy
lenticularis Linnaeus 1761W210FJ499282FJ499081 HAustria
lenticularis Linnaeus 1761W211FJ499283FJ499082 HAustria
lenticularis Linnaeus 1761W212FJ499284FJ499083FJ499183HAustria
lenticularis Linnaeus 1761W213FJ499285FJ499084 HAustria
lenticularis Linnaeus 1761W216FJ499286FJ499087FJ499186HJapan
lenticularis Linnaeus 1761W254FJ499280FJ499097FJ499193HAustria
lenticularis Linnaeus 1761W255FJ499281FJ499098FJ499194HAustria
lenticularis Linnaeus 1761ZMUC CRU-9948EF189651EF189671  Austria
sordida King 1855W110FJ499273FJ499052FJ499146DWestern Australia
sordida King 1855W111FJ499274FJ499053FJ499147DWestern Australia
sordida King 1855W118FJ499263FJ499055FJ499151DWestern Australia
sordida King 1855W119FJ499264FJ499056FJ499152DWestern Australia
sordida King 1855W120FJ499265 FJ499153DWestern Australia
sordida King 1855W121FJ499266FJ499057FJ499154DWestern Australia
sordida King 1855W137 FJ499067FJ499165DWestern Australia
sordida King 1855W197FJ499277 FJ499178DWestern Australia
sordida King 1855W198FJ499278FJ499077FJ499179DWestern Australia
sordida King 1855W299FJ499275FJ499113 DN. Terr., Australia
sordida King 1855W300FJ499276FJ499114 DN. Terr., Australia
stanleyana King 1855W179FJ499269 FJ499174DN.S.W., Australia
stanleyana King 1855W180FJ499270FJ499074FJ499175DN.S.W., Australia
urukhai Webb & Bell 1979W169FJ499268FJ499072FJ499173 N.S.W., Australia
Limnadopsisbirchii (Baird 1860) EF189652 AF526290 
parvispinus Henry 1924W108AY851453FJ499051FJ499145DWestern Australia
parvispinus Henry 1924W109AY851451  DWestern Australia
parvispinus Henry 1924W116AY851454FJ499054FJ499150 Western Australia
parvispinus Henry 1924W126AY851452 FJ499157 Western Australia
parvispinus Henry 1924W127 FJ499060FJ499158 Western Australia
tatei Spencer & Hall 1896W201FJ499287FJ499079FJ499181DN. Terr., Australia
tatei Spencer & Hall 1896W202FJ499288FJ499080FJ499182DN. Terr., Australia
tatei Spencer & Hall 1896W290FJ499289FJ499107FJ499205 N. Terr., Australia
sp. 1W305FJ499292FJ499115FJ499215 Western Australia
sp. 2W222FJ499290FJ499092FJ499190DWestern Australia
sp. 3W303FJ499291 FJ499214 Western Australia
Undescribed limnadopsoid speciessp.W291FJ499293FJ499108FJ499206 N. Terr., Australia
sp.W292FJ499294 FJ499207 N. Terr., Australia
sp.W295FJ499295 FJ499210 N. Terr., Australia
Undescribed eulimnadoidsp. 1NS74AY851439FJ499046FJ499137 Mauritius
sp. 2W199FJ499248FJ499078FJ499180 South Africa
sp. 2W284FJ499249   South Africa
sp. 2W286FJ499250   South Africa
sp. 2W287FJ499251   South Africa
sp. 2W288FJ499252FJ499105FJ499203 South Africa
sp. 2W289FJ499253FJ499106FJ499204 South Africa

Morphological analyses

The ethanol-fixed specimens were examined using a Wild M8 dissection stereomicroscope. To separate males from females/hermaphrodites, each specimen was examined for presence of eggs and elongated epipodites (females/hermaphrodites) or claspers (males). Because there are no recent keys for this family, species diagnostic characters were identified using descriptions from peer reviewed scientific literature, original descriptions, older keys and direct comparisons with previously identified material in public and private collections. Characters/character states were defined, scored and included in the phylogenetic analyses and their specifics are presented in Appendix S1.

Breeding system assignment

Breeding system determinations for 47 of the 54 limnadiid clam shrimp populations were identified in a recent study by Weeks et al. (2008). Breeding system determinations for four of the remaining seven populations were inferred using criteria outlined in that study, as follows. Weeks et al. (2008) concluded that within the Limnadiidae ‘using simple sex ratios to infer breeding system can be valid if sex ratios are 1 : 1 or strongly female-biased’. Populations that contain 100% egg-bearing individuals are considered all-hermaphroditic while those that have male frequencies at 45% or above are considered dioecious (Weeks et al., 2008). One of the seven populations noted above (i.e. that were not studied by Weeks et al. (2008)) had 0% males (represented by W149; Eulimnadia cylindrova from Desirade) and was thus considered hermaphroditic in the current study. Three of these seven populations were considered dioecious using the above noted 45% male criterion: (1) W161 from a population of L. badia collected from Western Australia – 46% males; (2) W198 from a population of L. sordida collected from Western Australia – 55% males; and (3) W299 from a population of L. sordida from collected Northern Territory, Australia – 56% males.

The remaining three populations (represented by W320, E. africana from Botswana; W225, E. brasiliensis from Brazil; and W246, E. dahli from Western Australia) all had natural sex ratios of 23–25% males and thus could not be classified using the above noted sex ratio criteria outlined by Weeks et al. (2008). All three populations had 3–8 hermaphrodites that produced male and hermaphroditic offspring in a 3:1 ratio. To date, all cases in which isolated hermaphrodites produced offspring with ∼25% males have been found to be androdioecious (Sassaman, 1988; Sassaman & Weeks, 1993; Weeks et al., 2006c, 2008). Therefore, we categorized these three remaining populations as androdioecious.

Breeding systems for most of the nonlimnadiid species included in our analyses were drawn from Sassaman (1995). The remainder was drawn from several other sources (Mattox 1950; Sassaman 1990; Tinti and Scanabissi 1996).

DNA sequencing

Total DNA was isolated from individual clam shrimp using the QIAGEN DNeasy Plant Kit (QIAGEN, Germantown, MD, USA). Portions of the nucleus-encoded 28S rDNA, the elongation factor 1-alpha (EF1α) and the mitochondrion-encoded cytochrome c oxidase I (COI) genes were polymerase chain reaction (PCR) amplified using the following primer pairs: 28S: D1F/D6R (Park & O’Foighil, 2000); EF1α: M44-1/3′EF1 (Braband et al., 2002); COI: 5′Cox1CrustForward 5′-TCHACHAAYCAYAARGAYATYGGNAC-3′, MidCox1CrustForward 5′-TNCCNGTNYTDGCNGGNGCHATYAC-3′, 3′Cox1LimnReverse 5′-TCDDYRTARCTRTGYTCWGCNGGRGG-3′. EF1α and 28S were chosen because of their phylogenetic utility in previous studies (EF1α: Braband et al., 2002; 28S: Hoeh et al., 2006), and COI because of its utility in many studies. Each PCR reaction consisted of 5 μL of 10× Qiagen PCR buffer, 1 μL of dNTPs (0.2 mm each), 2.5 μL of each primer (0.5 μm), between 1 and 5 μL of template DNA, 0.2 μL of Qiagen Taq polymerase (1 U), and enough H2O to bring the total volume to 50 μL. PCR reactions were carried out in PTC-100 and PTC-200 thermal cyclers (Bio-Rad Laboratories, Hercules, CA, USA). The thermal cycler programs consisted of an initial incubation at 85 °C for 1 min, followed by 45 cycles of 94 °C for 0.5 min, annealing at 40 °C for 28S rDNA, 53 °C for EF1α and 46 °C for COI for 1 min, and extending at 72 °C for 1.25 min, followed by a final extension of 72 °C for 10 min. PCR products were purified using 1.5% NuSieve (GTG agarose; FMC Bioproducts, Rockland, ME, USA) low melting point gels. Sequencing-template purification was performed using the Wizard PCR preps DNA purification system (Promega, Madison, WI, USA). The mitochondrial and nuclear amplicons were characterized by cycle sequencing using the PCR amplification primers. The protocols for cycle sequencing of the amplicons are as presented in Folmer et al. (1994) and they include cycle-sequencing of both strands of each purified template using labelled primers. The separation of cycle-sequencing-reaction products was performed in 3.7% and 5.5% polyacrylamide gels on LI-COR (LI-COR Biosciences, Inc., Omaha, NE, USA) 4200L-2 and 4200S-2 automated DNA sequencers, respectively. The resulting sequences were aligned initially using AlignIR (v2.0; LI-COR Biosciences, Inc.) with subsequent refinement performed manually using MacClade v. 4.05 (Maddison & Maddison, 2002). All sequences generated for this project have been deposited in the GenBank database (see Table 1 for accession nos). The alignment of the COI and EF1α sequences utilized herein was straightforward since no indels have been detected at these loci in the clam shrimp sequences we have generated to date. However, the 28S rDNA sequences contained multiple indels and such areas of ambiguous alignment were deleted prior to phylogenetic analyses. The aligned 28S matrix is available from the authors.

Phylogenetic analyses

Phylogenetic analyses were conducted on a concatenated 3480-character data set that included the three afore-mentioned genes (3453 characters: 28S = 962 bp, EF1α = 1039 bp, COI = 1452 bp) plus 27 morphological characters (Appendix S1) using Bayesian inference (BI) via Mr. Bayes (v. 3.1.2; Huelsenbeck & Ronquist, 2001; Ronquist & Huelsenbeck, 2003). The data set contained 167 terminals for which we generated sequences, plus an additional six terminals whose sequences were obtained from GenBank (Table 1). Two independent simultaneous analyses were performed using the GTR + G + I substitution model (Rodriguez et al., 1990). Searches were conducted for 13.224 million generations with six search chains each, the molecular data were partitioned by gene region and by codon position (two gene regions × three codon positions for the COI and EF1α partitions and a single partition for 28S rDNA) yielding a total of eight partitions, and saving a total of 52 896 trees (one tree saved every 500 generations in each of the two analyses). To allow each partition to have its own set of parameter estimates, revmat, tratio, statefreq, shape and pinvar were all unlinked during the analysis. The analyses were terminated when the standard deviation of split frequencies fell below 0.02. The 10 448 postburnin trees (determined by examination of the log probability of observing the data × generation plot) were used to calculate the majority rule consensus tree. To obtain the most accurate branch length estimates possible, the option prset ratepr = variable was employed as per the recommendations of Marshall et al. (2006). A best maximum likelihood (ML) tree (using default settings except for the following: autoterminate run 1 000 000 generations postlast improved topology, lnL increase for significantly better topology = 0.0001 and score improvement threshold = 0.0005) and a 1000-replicate ML majority-rule bootstrap (Felsenstein, 1985) tree (using default settings except for the following: lnL increase for significantly better topology = 0.001 and score improvement threshold = 0.005), based on analyses of the concatenated three-gene matrix with no data partitioning, were generated using GARLI (Zwickl, 2006). All phylogenetic analyses included representatives of (1) each extant limnadiid genus, (2) the Leptestheriidae and (3) the Cyzicidae (all families are Branchiopoda: Spinicaudata) and designated representatives of the Cyzicidae as the outgroup (as per figures 7 and 8 in Richter et al., 2007).

The estimation of ancestral breeding system character states (Table 1), based on the Bayesian topology with the highest overall posterior probability, was carried out using the ML algorithm in Mesquite (v.2.5; Maddison & Maddison, 2008). The 173 terminal best BI tree was reduced to 79 terminals by first pruning out the terminals for which the breeding system character states were unknown and then by reducing duplicate non-Eulimnadia lineages to single representative individuals. The ML optimization utilized the Markov k-state one parameter model (Lewis, 2001) and incorporated branch length and parameter estimates from the Bayesian analyses. The use of a likelihood ratio test to calculate P-values for ancestral states is not possible because hypotheses regarding the likelihoods of each possible state at a given node are non-nested. Therefore, to make decisions regarding the significance of ancestral character states, Pagel (1999; following Edwards, 1972) recommended that ancestral character state estimates with a log likelihood two or more units lower than the best state estimate [decision threshold (T) set to T = 2] be rejected. Generally viewed as a conservative cutoff, this threshold has been used by numerous recent authors (e.g. Moczek et al., 2006; Fernandez & Morris, 2007; Murphy et al., 2007; Koepfli et al., 2008). For the data presented herein, this protocol ensures that all of the character states judged to be significant have proportional likelihoods (PL) at least 10 times greater than that of any other state.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

The 173 terminal best BI tree (that with the highest posterior probability (PP) from our two independent analyses), with branch lengths, PPs (×100) and ML bootstrap information (1000 replicates) displayed, indicates strong support for limnadiid monophyly as well as for the monophyly of most traditional spinicaudate genera, such as Eulimnadia, Metalimnadia, Imnadia, Limnadopsis, Leptestheria, Cyzicus and Eocyzicus (Fig. 1). Additionally, two well supported, undescribed limnadiid clades, likely warranting generic rank, have been detected in South Africa (undescribed eulimnadioid lineage ZA, Fig. 1a) and Australia (undescribed limnadopsoid lineage AU, Fig. 1b). In contrast, representatives of the genus Limnadia occur in two distinct, well supported locations in the tree in Fig. 1: (1) in a clade (with terminals distributed in the Holarctic) sister to the genus Imnadia (Fig. 1b: node inline image) and (2) in a clade (with terminals distributed in Australia) sister to the genus Limnadopsis (Fig. 1b: node inline image). Taxonomic issues, such as the polyphyletic nature of the genus Limnadia and the undescribed limnadiid lineages, will be dealt with in separate manuscripts (D.C. Rogers et al., unpublished data) and are not germane to the discussion of breeding system evolution in the Limnadiidae that follows below. Strongly supported intergeneric relationships displayed in Fig. 1 include the sister taxon relationships of Eulimnadia Metalimnadia (Fig. 1a: node inline image) and ‘Australian Limnadia’ + Limnadopsis (Fig. 1b: node inline image). The above-described evolutionary relationships are also supported by the best ML tree (not shown).


Figure 1.  Bayesian tree of highest posterior probability showing the apical (1a) and basal (1b) halves of the tree from a combined evidence analysis of 28S, elongation factor 1-alpha (Ef1α), cytochrome c oxidase I (COI) and morphology. Bayesian PP ≥ 95 and maximum likelihood (ML) bootstrap percentages ≥70 are denoted with asterisks above and below the branches, respectively. Codes after taxon names indicate individual specimen numbers (see Table 1) and two-letter country designations: Australia (AU); Austria (AT); Brazil (BR); Ecuador (EC); Guadeloupe (GP); Italy (IT); Martinique (MQ); Mauritius (MU); Mexico (MX); Japan (JP); South Africa (ZA); Thailand (TH); United States (US); Venezuela (VZ). Highlighted nodes are as follows: (1a: node inline image) – intergeneric relationship of Eulimnadia + Metalimnadia; (1a: nodes inline image and inline image) – major lineages within Eulimnadia that contain one or more androdioecy-to-hermaphroditism transition; (1b: node inline image) – Holarctic Limnadia; (1b node inline image) – Australian Limnadia; and (1b: node inline image) – intergeneric relationship of Australian Limnadia + Limnadopsis.

Some species determinations within the Limnadiidae are likely problematic because of the lack of species monophyly sometimes displayed in Fig. 1 (e.g. E. diversa, E. follisimilis, E. cylindrova and L. sordida). Species and even generic determinations have been confusing in Eulimnadia and Limnadia for over a century, especially for Australian taxa (Sayce 1903; Henry 1924; Daday 1925; Straskraba 1964; Webb and Bell 1979; Belk 1989; Richter and Timms 2005). The specifics of these taxonomic issues will be the topic of a companion paper (Rogers et al. in preparation) and herein we will primarily concentrate on the inferred evolutionary transitions of the breeding systems within the Limnadiidae.

Figure 2 displays the ML estimation of breeding system ancestral states onto a 79 terminal topology that maintains the relative evolutionary relationships portrayed in the best 173 terminal BI tree (Fig. 1). Singular character state estimates for 57 of the 58 interior nodes in this topology were deemed significant by Mesquite. The internal nodes in Fig. 2 denote ‘PL’ for each of the four reproductive character states. Nodes that are primarily one colour usually signify a > 90% probability that the ancestral character was the type signified by the respective colours. There were only two nodes in which the PL of the most likely ancestral character state was < 90%: (1) the ancestral node for Eulimnadia Metalimnadia sp. (PLandrodioecy = 0.56; PLdioecy = 0.40; both of these states being significantly better than the other two possible states, but not significantly better than one another) and (2) the node defining the split between Cyzicus sp. and C. gynecia (PLdioecy = 0.87). Even though the PL for the majority state at the latter was < 0.9, this state was judged by ML to be the single, significantly best state for this node, and the PL for this state was more than 13 times greater than the PL for any other state.


Figure 2.  Maximum likelihood optimization of breeding system on a pruned topology from Fig. 1 analysed with Mesquite using the Markov k-state one parameter model. Taxa pruned from Fig. 1 includes those from populations whose breeding system are undetermined, as well as duplicate non-Eulimnadia lineages. Significance of ancestral character state estimates determined by one character state having a log likelihood two or more units higher than all others. All nodes are significant for a single character state except a single node, denoted with an asterisk (*), which has two states (androdioecy and dioecy) significantly better than the others. Codes after taxon names indicate individual specimen numbers (see Table 1) and two-letter country designations: Australia (AU); Austria (AT); Brazil (BR); Ecuador (EC); Guadeloupe (GP); Italy (IT); Mexico (MX); Japan (JP); Thailand (TH); United States (US); Venezuela (VZ). Highlighted nodes are as follows: node inline image– dioecy is the inferred ancestral state for the Limnadiidae; node inline image– transition to all-hermaphroditism in the holartic Limnadia; node inline image– transition to hermaphrodites + males (androdioecy) in the Eulimnadia; nodes inline image and inline image major lineages within Eulimnadia that contain one or more androdioecy-to-hermaphroditism transitions.

Download figure to PowerPoint

The breeding system ancestral states analysis indicates that dioecy was the breeding system of the limnadiid ancestral lineage (PL = 0.94; Fig. 2: node inline image). Furthermore, independent gains of hermaphroditic reproduction occurred in the ancestral lineage of Holarctic Limnadia (i.e. the Limnadia clade sister to Imnadia; Fig. 2: node inline image; Fig. 3: arrow A) and Eulimnadia (Fig. 2: node inline image; Fig. 3: arrow B). In the Holarctic Limnadia, the hermaphrodites replaced both males and females while in Eulimnadia, hermaphrodites replaced only females initially (yielding androdioecy) with later male loss in some populations (yielding all-hermaphroditism; Fig. 3: arrows B [RIGHTWARDS ARROW] C). Thus, within the typically androdioecious genus Eulimnadia, our ML optimization estimated that a shift from androdioecy to hermaphroditism has independently occurred seven times (Fig. 2; Fig. 3: arrow C). However, it should be noted that many of the nodes within Eulimnadia received low statistical support (BI PP < 0.95 and ML bootstrap percentage (BSP) < 70) as indicated by the relative paucity of asterisks on Fig. 1a. This topological instability can be accounted for when estimating the minimum number of breeding system shifts in Eulimnadia. Within Eulimnadia, there is a major subclade that received high Bayesian nodal support (Fig. 1a: node inline image, Fig. 2: node inline image) and contains four of the seven estimated independent transitions from androdioecy to hermaphroditism mentioned above. We could more conservatively estimate that this major subclade contains a single, independent transition by recognizing that the hermaphroditic lineages therein could actually form a clade. The same could be argued for the other three transitions occurring in the other major Eulimnadia subclade (Fig. 1a: node inline image, Fig. 2: node inline image). Thus, a conservative estimate of the minimum number of transitions from androdioecy to all-hermaphroditism within Eulimnadia would be two independent transitions. However, considering that there are some relatively long branch lengths separating some of the taxa within these subclades (e.g. the total branch length between E. michaeli and any one E. dahli), the actual number of androdioecy-to-hermaphroditism transitions within Eulimnadia likely lies between two and seven. The current ancestral states analysis suggests one single transition to asexuality from dioecy in the all-female Cyzicus gynecia (Fig. 2; Fig. 3: arrow D).


Figure 3.  Evolutionary transitions inferred from the analysis in Fig. 2. Arrow A: transition occurred in the ancestor to Limnadia lenticularis; arrow B: transition occurred in the ancestor to Eulimnadia; arrow C: transition occurred in the ancestor to some Eulimnadia species; arrow D: transition occurred in the ancestor to Cyzicus gynecia. The dashed arrow A denotes that although a possible direct pathway from dioecy to hermaphroditism may have occurred, an androdioecious intermediate is a more likely scenario (i.e. the B [RIGHTWARDS ARROW] C transition; see Discussion).

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

To understand the evolution of hermaphroditism in animals, we need to discern the number and types of transitions from ancestral states, and determine the selective processes (and potential constraints) that shape these transitions. For the former, mapping breeding system onto a robust phylogeny to infer evolutionary transitions is most useful (Kiontke et al., 2004; Sargent & Otto, 2004; Lopez-Vaamonde et al., 2005; Surget-Groba et al., 2006; Rossi et al., 2007). Herein we have conducted such a phylogenetic comparison and below we will interpret these transitions by considering the selective regimes and the potential constraints that affect these transitions.

Breeding system transitions within the Limnadiidae

It has long been assumed that the ancestral breeding system for the Limnadiidae was dioecy (Sassaman, 1995). Sassaman (1995) proposed a genetic model specifically for the clam shrimp by which parthenogenesis and androdioecy have directly evolved from dioecy and that selfing hermaphroditism and cyclic parthenogenesis were then derived from androdioecy and parthenogenesis, respectively. However, to date no one has conducted an ancestral character state reconstruction to confirm any of these assertions.

Using the character state optimization outlined in Fig. 2, we infer that the ancestral breeding system for the Limnadiidae is indeed dioecy (Fig. 2: inline image), as was suggested by Sassaman (1995). We further infer that there have been two separate derivations of hermaphroditism from dioecy: one in the progenitor to the all-hermaphroditic L. lenticularis (Fig. 2: inline image; Fig. 3: arrow A) and one in the progenitor to the hermaphroditic + male (i.e. androdioecious) Eulimnadia (Fig. 2: inline image; Fig. 3: arrow B). If a sister relationship existed between Limnadia lenticularis and Eulimnadia, the assertion that there were two independent derivations of hermaphroditism from dioecy would be questionable. However, there are two robustly supported nodes in Fig. 1a that reject this possibility: (1) (Metalimnadia +Eulimnadia) (Fig. 1a: node inline image) and (2) {undescribed eulimnadioid sp. 1 +  [undescribed eulimnadioid sp. 2 +  (Metalimnadia + Eulimnadia)]} (Fig. 1a: node inline image). Therefore, the inference of two independent derivations of hermaphroditism is robustly supported.

In the Eulimnadia, the hermaphroditic variants have outcompeted the females but have largely coexisted with males to form androdioecious populations (Fig. 2), which coincides with the assertions of Sassaman (Fig. 3: arrow C). In Limnadia lenticularis, our data suggest a direct derivation of all-hermaphroditism from dioecy (Fig. 3: dashed arrow A). There are no clear androdioecious close relatives to L. lenticularis (Fig. 2) and thus no evidence that this all-hermaphrodite species derived from an androdioecious progenitor. Nevertheless, there is good reason to suspect that such a progenitor may have initially evolved and has since gone extinct. We outline these arguments (largely drawn from Sassaman, 1995) below.

To understand the evolution of hermaphroditism in the Limnadiidae, Sassaman (1995) suggested that we use the genetic sex determining system first elucidated in Eulimnadia texana (Sassaman & Weeks, 1993) and assume it is shared among Eulimnadia more generally (Sassaman, 1995; Weeks et al., 2008). In this genetic system, males are homogametic (ZZ) while hermaphrodites are of two genetic types: ZW (termed ‘amphigenic’) and WW (‘monogenic’). Selfing ZW hermaphrodites produce one-quarter males while selfing WW hermaphrodites produce all hermaphrodites (Sassaman & Weeks, 1993). Sassaman suggested that the derivation of all-hermaphroditic limnadiid lineages is a simple product of selection for the WW hermaphrodites from within this mix of the three mating types (Fig. 3: arrow C).

We see evidence of Sassaman’s supposition within the Eulimnadia (Fig. 2). Each of the major subclades within Eulimnadia (Fig. 2: nodes inline image and inline image) has experienced at least one derivation of all-hermaphroditism from androdioecy, and if our best estimate of phylogeny is correct (Fig. 1), as many as seven independent derivations of hermaphroditism have occurred among the Eulimnadia populations we sampled (Fig. 2). Additionally, two other Eulimnadia species have all-hermaphrodite populations from which data have not been analysed herein (E. diversa and E. feriensis Dakin 1914), and these all-hermaphrodite populations are much less common than their androdioecious conspecific counterparts (Sassaman, 1989; Weeks et al., 2008). Thus, in the current and previous studies, it appears that all-hermaphrodite populations have been repeatedly derived from androdioecious populations, and we may expect this to have occurred in the development of all-hermaphroditism in L. lenticularis also (see ‘Re-evaluation’ section below).

Adaptive mechanisms promoting the evolution of hermaphroditism

Sassaman’s (1995) model is primarily genetically based, and thus does not provide expected criteria under which one breeding system should be selected over another. However, there are two published mechanisms by which all-hermaphrodite populations may be expected to be derived from androdioecious progenitors. First, Chasnov (in press) suggested that outcrossing may be selected against in hermaphrodites which have < 50% inbreeding depression among selfed offspring. Chasnov & Chow (2002) additionally predicted that such hermaphrodites should be selected to reduce or eliminate outcrossing with males, leading to all-selfing, hermaphroditic populations. Such reduced outcrossing has apparently been selected in the androdioecious Caenorhabditis elegans (Chasnov & Chow, 2002; Chasnov et al., 2007). If this phenomenon were occurring in Eulimnadia, we would then expect a lower propensity to mate and a general observation of lower inbreeding depression in the all-hermaphrodite compared with the androdioecious populations. At this point, we do not have the data needed to test this hypothesis, but this ‘reduced outcrossing propensity’ model could clearly explain the derivation of all-hermaphrodite populations from androdioecious progenitors in Eulimnadia.

A second hypothesis has been suggested by Pannell (1997, 2002): hermaphrodites are better early colonists and thus commonly are found in all-hermaphroditic, younger populations. Males are then later able to colonize these younger pools to re-establish androdioecy as the populations become larger and better established. There is strong evidence that this metapopulation hypothesis explains the mix of androdioecious and all-hermaphrodite populations of the plant Mercurialis annua (Obbard et al., 2006; Dorken & Pannell, 2008; Pannell et al., 2008). If this mechanism operates in Eulimnadia, we would then expect all-hermaphrodite populations to be younger, have lower genetic diversity, and have higher among-population genetic differentiation (i.e. higher FST) than androdioecious populations (Pannell, 2002; Obbard et al., 2006). Again, we do not yet have sufficient data to test these predictions, but clearly this hypothesis could well explain the observed patterns of sex ratio variation among populations in the genus Eulimnadia.

Both of the above models assume that hermaphroditism is selected within a dioecious species because of the advantages of ‘reproductive assurance’ (Baker, 1955) when population sizes are commonly low, such as in species that regularly colonize new habitats. Short-lived, ephemeral ponds are the typical habitat for these clam shrimp (Dumont & Negrea, 2002; Weeks & Bernhardt, 2004), and thus reproductive assurance is completely feasible as an important aspect of the life history of these branchiopod crustaceans.

If reproductive assurance is the primary force selecting hermaphroditism, as postulated, then the hermaphrodites should be primarily ‘female-biased’ because such low-density situations would disallow much fitness gain through male function (Pannell, 1997). In other words, the hermaphrodites should be primarily allocating reproductive investment to egg production and only produce enough sperm to ensure fertilization of their own eggs. This prediction is upheld in Eulimnadia as well as L. lenticularis hermaphrodites: hermaphrodites allocate only a small portion of their gonads to sperm production (Zaffagnini, 1969; Zucker et al., 1997; Scanabissi & Mondini, 2002; Weeks et al., 2005). Such female-biased allocation is also noted in androdioecious nematodes (Ward & Carrel, 1979) and fish (Harrington, 1963). Thus, the life history prediction of these two models that hermaphrodites will be female-biased is upheld in the well-studied androdioecious animal species noted to date.

Potential constraints on the evolution of hermaphroditism from dioecy

An alternate argument has been forwarded for the observation of female-biased hermaphroditism in these shrimp and the other androdioecious animals noted above: the development of a functional hermaphrodite from a sexually dimorphic ancestor may be constrained to be one that functions primarily as one sex, that sex being female (Weeks et al., 2006a). If there are many physiological, morphological and/or behavioural traits that differ between males and females (i.e. the species is strongly sexually dimorphic), the odds of producing a hermaphrodite that fully captures all of the necessary phenotypes of both sexes to function equally well in both sexual roles might be prohibitively low. For example, clam shrimp males have male gonads, ‘claspers’ (used to attach to females during sperm transfer), elongate carapaces and male-specific behaviours (e.g. searching behaviour, faster swimming, etc.; Scanabissi Sabelli & Tommasini, 1994; Knoll, 1995; Olesen et al., 1996; Medland et al., 2000). Females have female gonads, ovoid carapaces, a ‘brood chamber’ to store eggs, extensions of their epipodites for egg attachment and female-specific behaviours (e.g. slow swimming, hole digging for egg laying, etc.; Scanabissi Sabelli & Tommasini, 1990; Dumont & Negrea, 2002; Zucker et al., 2002). If each of these traits is encoded by one or more genes, the odds of mutations or re-arrangements of these genes to form a phenotype that successfully combines all traits from both sexes is miniscule. More commonly, a ‘hermaphrodite’ would likely be a dysfunctional combination of some subset of the sexual phenotypes of both sexes. For example, we have observed one case of an E. texana‘intersex’ that had male claspers, male mating behaviour, and apparently functional ovotestes (Weeks et al., 2006b). However, this intersex did not have a brood chamber nor epipodites for egg attachment; therefore all of its eggs were found in distorted clumps and all eggs proved to be inviable. Additionally, the individual had a normal E. texana hermaphrodite’s ovotestes, which is highly skewed toward egg production (Zucker et al., 1997), and thus could not produce enough sperm to effectively fertilize hermaphrodites. Thus, although this intersex was ‘closer’ to being fully competent in male and female roles than the common female-biased, self-compatible hermaphrodites (i.e. it had the claspers needed for pairing, had the appropriate mate searching behaviour, and produced fully yolked and shelled eggs), it still did not have all the needed character traits to be competent in either sexual role and therefore was sterile. Thus, a more parsimonious expectation for the formation of a functional hermaphrodite would be one that is primarily one sex but that had co-opted one or at most a few of traits of the opposite sex (e.g. through mutation or crossing over; Weeks et al., 2006b). If this were true, the most likely arrangement to be selectively advantageous would be a female that could produce sperm but had no other male traits (Weeks et al., 2006a). This would be more functional than a male that produced eggs, since egg production commonly needs extra traits to produce viable offspring, such as the brood chamber and hole-digging behaviour in the clam shrimp example noted above.

Thus, although the independent derivations of female-biased hermaphroditism within the Limnadiidae noted herein (i.e. in Limnadia lenticularis and Eulimnadia) is consistent with two models based on reproductive assurance (Pannell, 1997; Chasnov, in press), it can also be explained by a constraint argument based on the most parsimonious method to produce a hermaphrodite from a sexually dimorphic, dioecious progenitor (Weeks et al., 2006a). Further data collection that can confirm/reject the additional predictions of the two selective models in nematodes, killifish and clam shrimp should resolve which of these explanations is most viable.

Re-evaluation of Sassaman’s model of the evolution of hermaphroditism in the Limnadiidae

We can use the above discussion to construct an argument that is consistent with Sassaman’s (1995) hypothesis for the development of hermaphroditism within the Limnadiidae. Let us assume that self-compatible hermaphroditism is selected from dioecy because of the benefits of ‘reproductive assurance’ in sperm-limited environments (Pannell, 1997; Wolf & Takebayashi, 2004; Chasnov, in press). A female-biased hermaphrodite is either specifically selected (Pannell, 1997; Chasnov, in press) or is the only viable mechanism to produce a functioning hermaphrodite in the Limnadiidae (Weeks et al., 2006a). Such a female-biased, hermaphroditic variant arose twice within the Limnadiidae (Fig. 2). In Eulimnadia, this hermaphroditic variant then spread to displace females but was maintained with males, either because the correct balance of migration and colonization rates was achieved (Pannell, 1997, 2002) or because this migration/colonization process is combined with a constraint on the elimination of males because of the unique sex determining mechanism in this genus (Pannell, 2008). In L. lenticularis, the female-biased hermaphroditic variant spread to displace both females and males, either because very high levels of extinction and low migration rates caused most populations to be in a constant state of low abundance and recent establishment (Pannell, 1997) or because inbreeding depression among selfed offspring was below the threshold 50% level favouring selfing over outcrossing (Chasnov, in press). Chasnov argued that the latter scenario would be a two-step process, which would first manifest as hermaphrodites displacing females to form androdioecy and then later spreading to displace males once inbreeding depression is purged to the point where inbred offspring experience < 50% inbreeding depression. If this two-step process is valid, then the direct evolution of hermaphroditism from dioecy (Fig. 3: arrow A) did not occur but rather an androdioecious intermediate developed for some period of time and was later replaced by the all-hermaphrodite WW lineages, as predicted by Sassaman’s (1995) model (Fig. 3: arrows B and C). Additionally, an argument can be made that some of the current populations/species of Eulimnadia may be undergoing Chasnov’s second stage (i.e. elimination of males) that L. lenticularis underwent at some point in the more distant past.

Parthenogenesis derived from dioecy?

One last reproductive transition obvious in Fig. 2 is the derivation of parthenogenesis from dioecy in Cyzicus gynecia (Fig. 3: arrow D). Sassaman (1995) predicted that C. gynecia evolved directly from a dioecious ancestor, likely C. mexicana, by a mutation suppressing meiosis. Our data are certainly consistent with this prediction, although we cannot assess the underlying genetics of the reported asexuality in C. gynecia. Indeed, to date, no one has determined whether C. gynecia is truly parthenogenetic rather than being self-compatible hermaphrodites; determination of parthenogenesis has been only on the basis of an observed lack of males (Sassaman, 1995). Thus, it would be constructive to assess the genetics and anatomy of C. gynecia‘females’ to check for levels of heterozygosity (parthenogenesis is commonly associated with high heterozygosity while selfing hermaphrodites are commonly completely homozygous; Bell, 1982) and the presence/absence of testicular tissue to determine the true mode of reproduction. Additionally, a population genetic comparison with C. mexicana (as suggested in Sassaman, 1995) and other Cyzicus species would allow a test of Sassaman’s prediction that C. gynecia was recently derived from C. mexicana.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

In conclusion, our data indicate that self-compatible hermaphroditism arose from dioecy independently twice within the Limnadiidae, likely because of the benefits of reproductive assurance in low-density environments. We suggest that the predictions of Sassaman (1995), that androdioecy and parthenogenesis are directly derived from dioecy (Fig. 3: arrows B and D, respectively) and that selfing hermaphroditism is secondarily derived from androdioecy (Fig. 3: arrow C), are true, although we cannot refute the possibility that the all-hermaphrodite L. lenticularis was directly derived from dioecy (Fig. 3: arrow A). In the limnadiid lineages examined to date, hermaphrodites are always ‘female-biased’ (i.e. produce few sperm and cannot outcross through male function). This type of hermaphrodite is consistent with other androdioecious systems in which males coexist with female-biased hermaphrodites (e.g. nematodes and killifish) and may be explained either using adaptive models which predict such female-biased hermaphroditism (Pannell, 1997, 2002; Chasnov, in press) or by a constraint argument based on the most parsimonious mechanism by which self-compatible hermaphroditism can be derived from a sexually dimorphic, dioecious ancestor (Weeks et al., 2006a). Future studies should concentrate on testing the predictions of the two adaptive models combined with a comparative assessment of the validity of the constraint hypothesis. Additionally, although these models do predict a transitional pathway to produce hermaphrodites from dioecy, they are not sufficient to explain how fully functional hermaphrodites (i.e. that are competent in both male and female roles) can evolve from a dioecious ancestor. Because the majority of animal hermaphrodites appear to be derived from dioecious ancestors (Ghiselin, 1969, 1974; Jarne & Charlesworth, 1993; but see Eppley & Jesson, 2008; Lyer & Roughgarden, 2008 for an alternative interpretation), we need to expand our models to include an explanation of the derivation of fully functional, outcrossing hermaphrodites from dioecious progenitors.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

The authors thank: M. Hamer, L. Brendonck, N. Rabet, M. Grygier, B. Timms, A. Ohtaka, B. Lang, M. Hill, U. Balaraman, G. Pereira, S. Leslie, L. Sanoamuang, A. Ooyagi, J. Hoover, A. Ferreira, E. Eder, S. Richter, S. Wu, M. Cesari, F. Scanabissi, J. Garcia, D. Smith, A. Maeda-Martinez, Merlijn Jocqué and T. Spears for soil samples and/or preserved clam shrimp; C. Sassaman for help with species identifications; A. Crow, C. Komar, R. Posgai and B. Wallace for help with rearing clam shrimp in the wet lab; and R. Mitchell and N. Rabet for thoughtful comments on a previous version of this paper. Eric Chapman is supported by the Kentucky Agricultural Experiment Station State Project KY008043. This material is based upon work supported by the National Science Foundation under Grant No. DEB-0235301.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information
  • Baird, W. 1849. Monograph of the family Limnadiidae, a family of entomostracous crustacea. Proc. Zool. Soc. Lond. 17: 8490.
  • Baker, H.G. 1955. Self-compatibility and establishment after “long-distance” dispersal. Evolution 9: 347349.
  • Barrett, S.C.H. 2002. The evolution of plant sexual diversity. Nat. Rev. Gen. 3: 274284.
  • Belk, D. 1989. Identification of species in the conchostracan genus Eulimnadia by egg-shell morphology. J. Crust. Biol. 9: 115125.
  • Bell, G. 1982. The Masterpiece of Nature. University of California Press, Berkeley, CA.
  • Braband, A., Richter, S., Hiesel, R. & Scholtz, G. 2002. Phylogenetic relationships within the Phyllopoda (Crustacea, Branchiopoda) based on mitochondrial and nuclear markers. Mol. Phylogenet. Evol. 25: 229244.
  • Brtek, J. 1997. Checklist of the valid and invalid names of the “large branchiopods” (Anostraca, Notostraca, Spinicaudata and Laevicaudata), with a survey of the taxonomy of all Branchiopoda. Zb. Slov. Nar. Muz. Prir. Vedy 43: 265.
  • Charlesworth, D. 1984. Androdioecy and the evolution of dioecy. Biol. J. Linn. Soc. 22: 333348.
  • Charlesworth, D. 2006. Evolution of plant breeding systems. Curr. Biol. 16: R726R735.
  • Charlesworth, D. & Charlesworth, B. 1978. A model for the evolution of dioecy and gynodioecy. Am. Nat. 112: 975997.
  • Charnov, E.L. 1982. The Theory of Sex Allocation. Princeton University Press, Princeton, NJ.
  • Charnov, E.L., Maynard Smith, J. & Bull, J.J. 1976. Why be an hermaphrodite? Nature 263: 125126.
  • Chasnov, J.R. in press. Sexual conflict over mating in androdioecious nematode worms and clam shrimp. Evolution.
  • Chasnov, J.R. & Chow, K.L. 2002. Why are there males in the hermaphroditic species Caenorhabditis elegans? Genetics 160: 983994.
  • Chasnov, J.R., So, W.K., Chan, C.M. & Chow, K.L. 2007. The species, sex, and stage specificity of a Caenorhabditis sex pheromone. Proc. Natl Acad. Sci. USA 104: 67306735.
  • Daday, E. 1925. Monographie systematique des Phyllopodes Conchostraces. Troisieme partie. Annales des Sciences Naturelles, Zoologies 10e serie 8: 143184.
  • Delph, L.F. & Wolf, D.E. 2005. Evolutionary consequences of gender plasticity in genetically dimorphic breeding systems. New Phytol. 166: 119128.
  • Dorken, M.E. & Pannell, J.R. 2008. Density-dependent regulation of the sex ratio in an annual plant. Am. Nat. 171: 824830.
  • Dumont, H.J. & Negrea, S.V. 2002. Introduction to the Class Branchiopoda. Backhuys, Leiden.
  • Edwards, A.W.F. 1972. Likelihood. Cambridge University Press, Cambridge.
  • Eppley, S.M. & Jesson, L.K. 2008. Moving to mate: the evolution of separate and combined sexes in multicellular organisms. J. Evol. Biol. 21: 727736.
  • Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783791.
  • Fernandez, A.A. & Morris, M.R. 2007. Sexual selection and trichromatic color vision in primates: statistical support for the preexisting-bias hypothesis. Am. Nat. 170: 1020.
  • Folmer, O., Black, M., Hoeh, W.R., Lutz, R. & Vrijenhoek, R.C. 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Mar. Biol. Biotechnol. 3: 294299.
  • Fryer, G. 1987. A new classification of the branchiopod crustacea. Zool. J. Linn. Soc. 91: 357383.
  • Ghiselin, M.T. 1969. The evolution of hermaphroditism among animals. Q. Rev. Biol. 44: 189208.
  • Ghiselin, M.T. 1974. The Economy of Nature and the Evolution of Sex. University of California Press, Berkeley, CA.
  • Harrington, R.W. 1963. Twenty-four hour rhythms of internal self-fertilization and of oviposition by hermaphrodites of Rivulus marmoratus. Physiol. Zool. 36: 325341.
  • Henry, M. 1924. A monograph of the freshwater Entomostraca of New South Wales. Proc. Linn. Soc. NSW 49: 120137.
  • Hoeh, W.R., Smallwood, N.D., Senyo, D.M., Chapman, E.G. & Weeks, S.C. 2006. Evaluating the monophyly of Eulimnadia and the Limnadiinae (Branchiopoda: Spinicaudata) using DNA sequences. J. Crustacean Biol. 26: 182192.
  • Huelsenbeck, J.P. & Ronquist, F. 2001. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17: 754755.
  • Jarne, P. & Auld, J.R. 2006. Animals mix it up too: the distribution of self-fertilization among hermaphroditic animals. Evolution 60: 18161824.
  • Jarne, P. & Charlesworth, D. 1993. The evolution of the selfing rate in functionally hermaphrodite plants and animals. Annu. Rev. Ecol. Syst. 24: 441466.
  • Kiontke, K., Gavin, N.P., Raynes, Y., Roehrig, C., Piano, F. & Fitch, D.H.A. 2004. Caenorhabditis phylogeny predicts convergence of hermaphroditism and extensive intron loss. Proc. Natl Acad. Sci. USA 101: 90039008.
  • Knoll, L. 1995. Mating-behavior and time budget of an androdioecious crustacean, Eulimnadia texana (Crustacea, Conchostraca). Hydrobiologia 298: 7381.
  • Koepfli, K.P., Deere, K.A., Slater, G.J., Begg, C., Begg, K., Grassman, L., Lucherini, M., Veron, G. & Wayne, R.K. 2008. Multigene phylogeny of the Mustelidae: resolving relationships, tempo and biogeographic history of a mammalian adaptive radiation. BMC Biol. Vol. 6, article 10.
  • Lewis, P.O. 2001. A likelihood approach to estimating phylogeny from discrete morphological character data. Syst. Biol. 50: 913925.
  • Lloyd, D.G. 1975. The maintenance of gynodioecy and androdioecy in angiosperms. Genetica 45: 325339.
  • Lopez-Vaamonde, C., Godfray, H.C.J., West, S.A., Hansson, C. & Cook, J.M. 2005. The evolution of host use and unusual reproductive strategies in Achrysocharoides parasitoid wasps. J. Evol. Biol. 18: 10291041.
  • Lyer, P. & Roughgarden, J. 2008. Dioecy as a specialization promoting sperm delivery. Evol. Ecol. Res. 10: 867892.
  • Maddison, W.P. & Maddison, D.R. 2002. MacClade: Analysis of Phylogeny and Character Evolution, Version 4.05. Sinauer Assoc., Inc., Sunderland, MA.
  • Maddison, W.P. & Maddison, D.R. (2008) Mesquite: A Modular System for Evolutionary Analysis, Version 2.5.
  • Marshall, D.C., Simon, C. & Buckley, T.R. 2006. Accurate branch length estimation in partitioned Bayesian analyses requires accommodation of among-partition rate variation and attention to branch length priors. Syst. Biol. 55: 9931003.
  • Mattox, N.T. 1950. Notes on the life history and description of a new species of conchostracan phyllopod, Caenestheriella gynecia. Trans. Amer. Micro. Soc. 69: 5053.
  • Medland, V.L., Zucker, N. & Weeks, S.C. 2000. Implications for the maintenance of androdioecy in the freshwater shrimp, Eulimnadia texana Packard: encounters between males and hermaphrodites are not random. Ethology 106: 839848.
  • Moczek, A.P., Cruickshank, T.E. & Shelby, A. 2006. When ontogeny reveals what phylogeny hides: gain and loss of horns during development and evolution of horned beetles. Evolution 60: 23292341.
  • Murphy, N.P., Carey, D., Castro, L.R., Dowton, M. & Austin, A.D. 2007. Phylogeny of the platygastroid wasps (Hymenoptera) based on sequences from the 18S rRNA, 28S rRNA and cytochrome oxidase I genes: implications for the evolution of the ovipositor system and host relationships. Biol. J. Linn. Soc. 91: 653669.
  • Negrea, S., Botnariuc, N. & Dumont, H.J. 1999. Phylogeny, evolution and classification of the Branchiopoda (Crustacea). Hydrobiologia 412: 191212.
  • Obbard, D.J., Harris, S.A. & Pannell, J.R. 2006. Sexual systems and population genetic structure in an annual plant: testing the metapopulation model. Am. Nat. 167: 354366.
  • Olesen, J., Martin, J.W. & Roessler, E.W. 1996. External morphology of the male of Cyclestheria hislopi (Baird, 1859) (Crustacea, Branchiopoda, Spinicaudata), with a comparison of male claspers among the Conchostraca and Cladocera and its bearing on phylogeny of the ‘bivalved’ Branchiopoda. Zool. Scr. 25: 291316.
  • Pagel, M. 1999. The maximum likelihood approach to reconstructing ancestral character states of discrete characters on phylogenies. Syst. Biol. 48: 612622.
  • Pannell, J.R. 1997. The maintenance of gynodioecy and androdioecy in a metapopulation. Evolution 51: 1020.
  • Pannell, J.R. 2002. The evolution and maintenance of androdioecy. Annu. Rev. Ecol. Syst. 33: 397425.
  • Pannell, J.R. 2008. Consequences of inbreeding depression due to sex-linked loci for the maintenance of males and outcrossing in branchiopod crustaceans. Genet. Res. 90: 7384.
  • Pannell, J.R., Dorken, M.E., Pujol, B. & Berjano, R. 2008. Gender variation and transitions between sexual systems in Mercurialis annua (Euphorbiaceae). Int. J. Plant Sci. 169: 129139.
  • Park, J.-K. & O’Foighil, O. 2000. Sphaeriid and corbiculid clams represent separate heterodont bivalve radiations into freshwater environments. Mol. Phylogenet. Evol. 14: 7588.
  • Richter, S. & Timms, B.V. 2005. A list of the recent clam shrimps (Crustacea: Laevicaudata, Spinicaudata, Cyclestherida) of Australia, including a description of a new species of Eocyzicus. Rec. Aust. Mus. 57: 341354.
  • Richter, S., Olesen, J. & Wheeler, W.C. 2007. Phylogeny of Branchiopoda (Crustacea) based on a combined analysis of morphological data and six molecular loci. Cladistics 23: 301336.
  • Rodriguez, F., Oliver, J.L., Marin, A. & Medina, J.R. 1990. The general stochastic model of nucleotide substitution. J. Theor. Biol. 142: 485501.
  • Ronquist, F. & Huelsenbeck, J.P. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 15721574.
  • Rossi, V., Gandolfi, A., Baraldi, F., Bellavere, C. & Menozzi, P. 2007. Phylogenetic relationships of coexisting Heterocypris (Crustacea, Ostracoda) lineages with different reproductive modes from Lampedusa Island (Italy). Mol. Phylogenet. Evol. 44: 12731283.
  • Sargent, R.D. & Otto, S.P. 2004. A phylogenetic analysis of pollination mode and the evolution of dichogamy in angiosperms. Evol. Ecol. Res. 6: 11831199.
  • Sassaman, C. 1988. Clutch sex-ratio polymorphism in the clam shrimp Eulimnadia antlei. Am. Zool. 28: A135.
  • Sassaman, C. 1989. Inbreeding and sex-ratio variation in female-biased populations of a clam shrimp, Eulimnadia texana. Bull. Mar. Sci. 45: 425432.
  • Sassaman, C. 1990. Inheritance of a sex-linked locus in the clam shrimp Eocyzicus concavus (Mackin). Am. Zool. 30: A108.
  • Sassaman, C. 1995. Sex determination and evolution of unisexuality in the Conchostraca. Hydrobiologia 298: 4565.
  • Sassaman, C. & Weeks, S.C. 1993. The genetic mechanism of sex determination in the conchostracan shrimp Eulimnadia texana. Am. Nat. 141: 314328.
  • Sayce, O.A. 1903. The Phyllopoda of Australia, including descriptions of some new genera and species. Proc. R. Soc. Vic. 15: 224261.
  • Scanabissi, F. & Mondini, C. 2002. A survey of the reproductive biology in Italian branchiopods. Hydrobiologia 486: 263272.
  • Scanabissi Sabelli, F.S. & Tommasini, S. 1990. Origin and early development of female germ-cells in Eoleptestheria ticinensis Balsamo-Crivelli, 1859 (Crustacea, Branchiopoda, Conchostraca). Mol. Reprod. Dev. 26: 4752.
  • Scanabissi Sabelli, F.S. & Tommasini, S. 1994. Functional morphology and ultrastructure of the male reproductive system in the Leptestheriidae (Branchiopoda, Conchostraca). Crustaceana 67: 362370.
  • Schemske, D.W. & Lande, R. 1985. The evolution of self-fertilization and inbreeding depression in plants. 2. Empirical observations. Evolution 39: 4152.
  • Spears, T. & Abele, L.G. 2000. Branchiopod monophyly and interordinal phylogeny inferred from 18S ribosomal DNA. J. Crustacean Biol. 20: 124.
  • Straskraba, M. 1964. Taxonomic studies on Czechoslovak chonchostraca, 1. Family Limnadiidae. Crustaceana 9: 263273.
  • Surget-Groba, Y., Heulin, B., Guillaume, C.P., Puky, M., Semenov, D., Orlova, V., Kupriyanova, L., Ghira, I. & Smajda, B. 2006. Multiple origins of viviparity, or reversal from viviparity to oviparity? The European common lizard (Zootoca vivipara, Lacertidae) and the evolution of parity. Biol. J. Linn. Soc. 87: 111.
  • Tinti, F. & Scanabissi, F. 1996. Reproduction and genetic variation in clam shrimps (Crustacea, Branchiopoda, Conchostraca). Can. J. Zool. 74: 824832.
  • Ward, S. & Carrel, J.S. 1979. Fertilization and sperm competition in the nematode Caenorhabditis elegans. Dev. Biol. 73: 304321.
  • Webb, J.A. & Bell, G.D. 1979. A new species of Limnadia (Crustacea: Conchostraca) from the granite belt in southern Queensland and Northern New South Wales. Proc. Linn. Soc. NSW 103: 237246.
  • Weeks, S.C. & Bernhardt, R.L. 2004. Maintenance of androdioecy in the freshwater shrimp, Eulimnadia texana: field estimates of inbreeding depression and relative male survival. Evol. Ecol. Res. 6: 227242.
  • Weeks, S.C., Marcus, V. & Alvarez, S. 1997. Notes on the life history of the clam shrimp, Eulimnadia texana. Hydrobiologia 359: 191197.
  • Weeks, S.C., Marcus, V. & Crosser, B.R. 1999. Inbreeding depression in a self-compatible, androdioecious crustacean, Eulimnadia texana. Evolution 53: 472483.
  • Weeks, S.C., Crosser, B.R. & Gray, M.M. 2001. Relative fitness of two hermaphroditic mating types in the androdioecious clam shrimp, Eulimnadia texana. J. Evol. Biol. 14: 8394.
  • Weeks, S.C., Posgai, R.T., Cesari, M. & Scanabissi, F. 2005. Androdioecy inferred in the clam shrimp Eulimnadia agassizii (Spinicaudata: Limnadiidae). J. Crustacean Biol. 25: 323328.
  • Weeks, S.C., Benvenuto, C. & Reed, S.K. 2006a. When males and hermaphrodites coexist: a review of androdioecy in animals. Integr. Comp. Biol. 46: 449464.
  • Weeks, S.C., Reed, S.K., Cesari, M. & Scanabissi, F. 2006b. Production of intersexes and the evolution of androdioecy in the clam shrimp Eulimnadia texana (Crustacea, Branchiopoda, Spinicaudata). Invertebr. Reprod. Dev. 49: 113119.
  • Weeks, S.C., Sanderson, T.F., Reed, S.K., Zofkova, M., Knott, B., Balaraman, U., Pereira, G., Senyo, D.M. & Hoeh, W.R. 2006c. Ancient androdioecy in the freshwater crustacean Eulimnadia. Proc. R. Soc. B Biol. Sci. 273: 725734.
  • Weeks, S.C., Sanderson, T.F., Zofkova, M. & Knott, B. 2008. Breeding systems in the clam shrimp family Limnadiidae (Branchiopoda, Spinicaudata). Invertebr. Biol. 127: 336349.
  • Weiblen, G.D., Oyama, R.K. & Donoghue, M.J. 2000. Phylogenetic analysis of dioecy in monocotyledons. Am. Nat. 155: 4658.
  • Wolf, D.E. & Takebayashi, N. 2004. Pollen limitation and the evolution of androdioecy from dioecy. Am. Nat. 163: 122137.
  • Zaffagnini, F. 1969. Rudimentary hermaphroditism and automictic parthenogenesis in Limnadia lenticularis (Phyllopoda, Conchostraca). Experientia 25: 650651.
  • Zucker, N., Cunningham, M. & Adams, H.P. 1997. Anatomical evidence for androdioecy in the clam shrimp Eulimnadia texana. Hydrobiologia 359: 171175.
  • Zucker, N., Aguilar, G.A., Weeks, S.C. & McCandless, L.G. 2002. Impact of males on variation in the reproductive cycle in an androdioecious desert shrimp. Invertebr. Biol. 121: 6672.
  • Zwickl, D.J. (2006) Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion. PhD Thesis, The University of Texas, Austin, TX, USA

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Appendix S1Characters and character states of the specimens coded for and present in the phylogenetic analyses.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

JEB_1813_sm_AppendixS1.doc41KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.