Morphologic and Genetic Variation in Triops (Branchiopoda: Notostraca) from Ephemeral Waters of the...

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, researchlibraries, and research funders in the common goal of maximizing access to critical research.

Morphologic and Genetic Variation in Triops (Branchiopoda: Notostraca) fromEphemeral Waters of the Northern Chihuahuan Desert of North AmericaAuthor(s): Kenneth S. Macdonald III, Rossana Sallenave, and David E. CowleySource: Journal of Crustacean Biology, 31(3):468-484. 2011.Published By: The Crustacean SocietyDOI: http://dx.doi.org/10.1651/10-3411.1URL: http://www.bioone.org/doi/full/10.1651/10-3411.1

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, andenvironmental sciences. BioOne provides a sustainable online platform for over 170 journals and books publishedby nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance ofBioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiriesor rights and permissions requests should be directed to the individual publisher as copyright holder.

http://dx.doi.org/10.1651/10-3411.1

http://www.bioone.org/doi/full/10.1651/10-3411.1

http://www.bioone.org

http://www.bioone.org/page/terms_of_use

MORPHOLOGIC AND GENETIC VARIATION IN TRIOPS (BRANCHIOPODA: NOTOSTRACA) FROM

EPHEMERAL WATERS OF THE NORTHERN CHIHUAHUAN DESERT OF NORTH AMERICA

Kenneth S. Macdonald III, Rossana Sallenave, and David E. Cowley

(KSM, correspondence, [email protected]; DEC, [email protected]) Department of Fish, Wildlife and Conservation Ecology,

New Mexico State University, Box 30003, MSC 4901, Las Cruces, NM 88003-8003, U.S.A.; Molecular Biology Program,

New Mexico State University, Las Cruces, New Mexico, U.S.A.;

(RS, [email protected]) Department of Fish, Wildlife and Conservation Ecology, New Mexico State University,

Box 30003, MSC 4901, Las Cruces, NM 88003-8003, U.S.A.; Department of Extension Animal Sciences and Natural Resources,

New Mexico State University, Las Cruces, New Mexico, U.S.A.

A B S T R A C T

Tadpole shrimp are known to be important animals in the ecology of ephemeral wetlands. In the northern Chihuahuan Desert of North

America, the tadpole shrimp fauna is composed of possibly three species in the genus Triops, which have variously been referred to as

species, subspecies, and intraspecific variation. Our results support the presence of three morphologically distinct forms, which will be

referred to herein as T. newberryi, T. longicaudatus ‘‘short,’’ and T. longicaudatus ‘‘long.’’ We report analyses of Triops spp. sampled in

summer 2008 from 14 natural playas and man-made flood retention ponds. Data were recorded on meristic counts and quantitative

measurements of morphological features. We also sequenced portions of the mitochondrial COI and ND1 genes for 72 shrimp, including

individuals from all three morphological forms and for multiple ponds for each form where possible. The three forms were

morphologically distinct for multiple characters and molecular analyses showed large differences in DNA nucleotide sequence and the

presence of multiple unique amino acid substitutions in each form. Finally, prior literature suggests the three forms exhibit different

reproductive systems, with populations of T. longicaudatus ‘‘long’’ thought to be gonochoric (equal sex ratios and obligate outcrossing),

T. longicaudatus ‘‘short’’ having only self-fertilizing hermaphrodites, and T. newberryi being androdioecious, having both self-fertilizing

hermaphrodites and males. While these three forms may be sufficiently distinct in morphology, mitochondrial DNA, and reproductive

life history to warrant elevation to species level, additional geographical sampling and an examination of the various type specimens are

necessary for a formal taxonomic revision.

KEY WORDS: androdioecy, ephemeral wetlands, hermaphroditism, Notostraca, playas, Triops

DOI: 10.1651/10-3411.1

INTRODUCTION

Ephemeral aquatic habitats are the main wetland type inmany arid and semi-arid regions and they are a primarysource of water for many plant and animal species forcertain periods of the year. The northern reaches of theChihuahuan Desert of North America, like many other aridregions world wide, have numerous ephemeral man-madeponds and natural ‘‘playa’’ lakes that fill during wetterclimatic cycles and then ebb with drier conditions. Becausethese playas store water in parts of the country that receivelittle rainwater and often have no permanent rivers orstreams, they are areas of high biodiversity and importanthabitat for migratory birds, amphibians, reptiles, andmammals. A remarkable assemblage of crustacean inver-tebrates has evolved to exploit the climatically drivenappearance and disappearance of water in these ephemeralwetlands.

Notostraca, commonly referred to as tadpole shrimp, areone of the more important animals in ephemeral wetlandsand even the surrounding landscape, where they stronglyaffect faunal composition (Haukos and Smith, 1994; Yee etal., 2005). The notostracan genus Triops occurs in

ephemeral wetlands in arid areas world wide, bridgingperiods of drought by laying desiccation-resistant restingeggs (or cysts) that can lay dormant and viable for years inunflooded soil (Brendonck, 1996; Brendonck and deMeester, 2003; Brendonck et al., 2008). Soon after re-immersion in water, a portion of the eggs hatch, and withina few weeks adult tadpole shrimp deposit hundreds tothousands of eggs before the ephemeral water-bodies dryagain (Brendonck, 1996). How these crustaceans aretransported between hydrologically isolated ephemeralwaters is unknown. While there has been much conjecture(Longhurst, 1955; Bilton et al., 2001; Figuerola et al.,2005), research on the topic has focused primarily on otherbranchiopods. Brendonck and Riddoch (1999) andVanschoenwinkel et al. (2008) demonstrated that a varietyof desiccated crustacean (but not notostracan) eggs leaveephemeral rock-pools by force of wind, and branchiopodeggs, but not notostracan eggs, have also been found in thefeces of wild waterfowl (Proctor, 1965; Figuerola et al.,2003; Green et al., 2008; Brochet et al., 2010), in the fur ofnutria (Waterkeyn et al., 2010a), and on the boots andvehicle tires of researchers (Waterkeyn et al., 2010b).

JOURNAL OF CRUSTACEAN BIOLOGY, 31(3): 468-484, 2011

468

However, systematic studies examining methods of notos-tracan dispersal have yet to be published.

The tadpole shrimps have a long evolutionary historyand easily recognizable Triopsidae are known fromPermian fossil beds (Ruedemann, 1922; Gand et al.,2008). Contemporary specimens of tadpole shrimps exhibitsufficient morphological variation within an overallconservative Bauplan to make the taxonomy of the genusuncertain. Classification of species in Triops is complicatedby the existence of a range of apparent reproductivesystems in nearly morphologically identical populations,from unisexual populations (no males) to bisexual popu-lations with sex ratios ranging from strongly female-biasedto equality (Sassaman, 1991), further confusing traditionalspecies identification.

The number of species of Triops recognized in NorthAmerica over the years has ranged from one, with thedescription of T. longicaudatus (as Apus longicaudatus;LeConte, 1846), to six, adding T. aequalis, T. lucasanus, andT. newberryi (also described as Apus spp.; Packard, 1871),and T. biggsi and T. oryzaphagus (Rosenberg, 1947). In the1950s, two reviews (Linder, 1952; Longhurst, 1955c),concluded that the genus Triops in North America wascomprised of a single, highly variable species, T. longi-caudatus. More recently, a survey of the genus Triops in thesouthwestern United States by Sassaman et al. (1997)concluded that there were three lineages that differed inthe number of legless rings and in population sex ratios. Oneof the lineages (with approximately eight mean legless ringsfor individuals bearing ovisacs) was tentatively elevated bySassaman et al. (1997) to species level and referred to as T.newberryi, while the other two were considered forms of T.longicaudatus. Triops longicaudatus ‘‘long’’ individuals hada greater number of legless abdominal rings (approximately10 mean legless rings; Sassaman et al., 1997), andpopulations exhibited generally equal sex ratios, while T.longicaudatus ‘‘short’’ individuals had fewer legless rings(approximately six mean legless rings; Sassaman et al.,1997) and had only ovisac-bearing individuals. Distribu-tional patterns of the two forms were inconsistent (see Fig. 1,page 127 of Sassaman et al., 1997), with equal-sex-ratiopopulations of T. longicaudatus tending towards the easternstates, southern California sites exclusively inhabited by T.

newberryi, and the rest of the distribution seeminglyhaphazardly inhabited by T. newberryi or the short-formpopulations of T. longicaudatus, or rarely, mixed popula-tions of the two latter forms. However, the survey ofSassaman et al. (1997) focused mostly in California (22collection sites) with three or fewer sites each in Arizona,Colorado, Kansas, Nevada, New Mexico, Utah and Mexico,so knowledge of distributions outside California is incom-plete. The purpose of this paper is to describe variation inmorphology, life history, and mitochondrial DNA forpopulations of Triops in ephemeral waters of the northernChihuahuan Desert in southern New Mexico (USA).

MATERIALS AND METHODS

Sampling Locations

A total of 30 playas and flood retention ponds in southern New Mexicowere visited between July and September 2008. Most playas weresubjected to minimal human alteration aside from an excavation at one endto enhance the water holding capacity of some playas to serve as an open-end stock tank for livestock. In addition to the naturally occurring playas,our sampling also included a number of man-made flood retention pondsthat intercept storm runoff into the Rio Grande. In southern New Mexico,overflow waters collected in these retention ponds drain into canals thatdischarge into the Rio Grande; the retained water may persist within theseponds for several weeks to months, depending on multiple factors such asthe amount of rainwater collected, the size of the flood basin, drainageefficiency, air temperature and relative humidity. When full, thesetemporary ponds are ecologically similar to the naturally occurringephemeral playa ecosystems, and are colonized by invertebrates andtadpoles, and used by amphibians, reptiles, waterfowl and other birdspecies. Flooding in the northern Chihuahuan Desert normally occursduring the rainy season in July through September, but some years thewetlands remain dry. Fortunately, the summer of 2008 was a relatively wetyear for southern New Mexico (National Resources Conservation Service,accessed January 2011: http://www.wcc.nrcs.usda.gov/cgibin/precip.pl?state5new_mexico). A list of 14 sampling locations containing Triopsspp. is provided in Table 1 and their locations are shown in Fig. 1.

In most cases the ponds were full of water at the time they were visited,and when present, live tadpole shrimp were sampled using 3 mm meshseines and 1 mm mesh dip nets. Tadpole shrimp were preservedimmediately in 95% ethanol, with 30 individuals at each location collectedfor genetic analysis preserved singly in individual 20 ml vials and returnedto the laboratory where they were stored at 220 uC. An additional 50 to100 individuals were collected from each site (when possible) andpreserved together for morphological measurements and estimation ofpopulation sex ratio. Water temperature, pH and conductivity weremeasured at each site using a Hach HQ portable meter (Hach Company

Table 1. Location and physical characteristics of collection sites. All sites were sampled from July through September 2008. FP designates floodretention ponds, PL designates natural and enhanced playas used as cattle stock ponds. CDRRC indicates Chihuahuan Desert Rangeland Research Center,and JER is the Jornada Experimental Range.

Site Location Type Lat/Long Date collected Salinity (%) Temp (uC) pH

FP02 Las Cruces Flood Retention Pond 32u17938.70N, 106u43927.00W 7/14/2008 0.10 33.4 7.98FP03 Las Cruces Flood Retention Pond 32u19913.70N, 106u44931.00W 7/14/2008 0.08 27.4 8.15FP04 Las Cruces Flood Retention Pond 32u20912.80N, 106u45928.20W 7/14/2008 0.07 32.3 8.79FP06 Garfield Flood Retention Pond 32u45937.20N, 107u14948.60W 9/19/2008 0.10 23.0 8.04PL05 CDRRC Enhanced Playa 32u32925.30N, 106u53900.90W 7/17/2008 0.06 33.2 8.27PL06 CDRRC Enhanced Playa 32u32946.80N, 106u49919.00W 7/17/2008 0.06 33.2 8.19PL07 CDRRC Enhanced Playa 32u31954.10N, 106u47923.60W 7/22/2008 0.14 33.6 8.07PL08 CDRRC Enhanced Playa 32u31938.00N, 106u46909.70W 7/22/2008 0.08 28.2 8.05PL09 CDRRC Natural Playa 32u32941.20N, 106u54941.50W 8/2/2008 0.10 24.2 7.69PL10 N of Las Cruces Enhanced Playa 32u28904.00N, 106u42953.30W 7/22/2008 0.10 23.7 7.95PL11 JER Enhanced Playa 32u30931.60N, 106u44933.20W 7/22/2008 0.12 27.6 7.95PL12 JER Enhanced Playa 32u33929.40N, 106u42900.40W 8/15/2008 0.09 24.4 8.10PL17 JER Enhanced Playa 32u41959.70N, 106u48904.70W – – – –PL23 SE of Deming Natural Playa 32u07900.90N, 107u21928.20W 9/12/2008 1.37 29.6 9.17

MACDONALD ET AL.: VARIATION IN NORTHERN CHIHUAHUAN TRIOPS 469

2006). All individual Triops specimens were identified to one of threepossible forms recognized by Sassaman et al. (1997) using the number oflegless rings and the apparent sexual composition of the population fromwhich the individual came (see below).

Morphological Measurements

All collected individual tadpole shrimp were examined to determine sexand to count the number of legless rings. Sex was determined by thepresence/absence of ovisacs modified from the eleventh pair of abdominalappendages (Linder, 1952). Some ovisac-bearing individuals may alsohave testicular tissue and may be hermaphrodites (see Discussion).However, because we did not examine ovisac-bearing individuals forpresence of testicular tissue, and since no external characters have yet beenfound to easily distinguish females from hermaphrodites, we will continueto call all individuals with ovisacs ‘‘ovisac-bearing individuals’’ through-out the manuscript. We also examined individuals for the presence/absenceof the mystax, a ‘‘conspicuous, protuberant, dark-brownish sclerodermalstripe located on the anteroventral flange of the carapace’’ (Obregon-Barboza et al., 2007) that was found on male North American Triops spp.,but absent in ovisac-bearing individuals. We have followed previousliterature in classifying individuals as males if they lacked ovisacs.However, we acknowledge that to date there has been no definitive geneticevidence proving that putative males contribute alleles to offspring inreproduction for the three forms considered in the present paper. Furthergenetic research is needed to clarify the reproductive systems on allspecies of Triops.

The number of legless abdominal rings was counted for all collectedindividuals from the anterior-most segment without any attached ventralappendages to the posterior-most segment, excluding the telson. Partialsegments, if they contained at least one spine, were counted as K rings,and added to the total, whereas spineless partial segments were notcounted. A subset of collected individuals (at least 30 ovisac-bearingindividuals and 30 males from each sampling site, where possible), werefurther measured for carapace length using a Leica Wild M5 stereoscopefitted with an ocular micrometer. Carapace length was measured along thedorsal midline of the carapace from the anterior-most point to the posteriormargin. Posterior marginal teeth were not included in the measurement.For this subset of individuals, we also counted the number of teeth on theposterior margin of the carapace (see Fig. 2). Teeth were counted only ifthey arose directly from the margin. They were counted in both directionsfrom the dorsal midline to the largest tooth on the posterior-lateral corner;additional smaller teeth were often found beyond this large tooth, but forthe sake of consistency were not counted. Three ponds containedpopulations consisting of two forms. For these populations, all collectedindividuals were measured for carapace length and posterior margincarapace teeth were counted. Presence of eggs in sacs, and egg color, were

also noted for all ovisac-bearing individuals measured for carapace length.Finally, for a haphazardly chosen subset of measured egg-carryingindividuals (minimum five from each collection site), all eggs wereremoved from the sac and counted. The diameter of ten eggs wasmeasured, and individual egg volumes were calculated using thevolumetric equation for a sphere: v 5 (4/3)pr3. While many eggs werenot spherical (the majority of non-spherical eggs were torus-shaped,possibly due to dehydration by the ethanol preservative), spherical eggswere haphazardly chosen to measure. Although counting of abdominalrings and all relevant measurements were performed on ovisac-bearingindividuals and males, previous species delineation has focused on ovisac-bearing individuals, and some populations lack males, so all statisticalanalyses testing interspecific differences only compared ovisac-bearingindividuals of the three forms. However, within T. longicaudatus ‘‘long’’and T. newberryi, we compared morphology between males and ovisac-bearing individuals to test for possible sexual dimorphisms in addition topresence/absence of ovisacs.

Statistical Analysis

We tested whether ovisac-bearing individuals of the three forms differedfor three morphological characters: number of teeth on the dorsal marginof the carapace, individual egg volume, and clutch size (number of eggsheld in the ovisac). Differences among morphological forms were testedusing a fixed effects ANOVA (GLM, SAS Institute, Inc. 1989). ANOVAswere run with pond nested within morphological form. Clutch size wascorrelated with body size, with equal slopes among the three forms,therefore carapace length was used as a covariate in the analysis of clutchsize differences among morphological forms. Separate analyses ofvariance were also conducted to test for differences in carapace length,dorsal margin teeth, individual egg volume, and clutch size between formsthat were co-inhabiting ponds. These analyses were modeled withmorphological form nested within pond. Carapace length was used as acovariate in the analysis of variance for clutch size.

We conducted additional analyses of variance to test for differences innumber of legless rings, carapace length, and dorsal margin teeth betweenmales and ovisac-bearing individuals. These analyses were for a modelwith sex nested within pond.

Molecular Techniques

We sequenced the LCO-1490/HCO-2198 (Folmer et al., 1994) portion ofthe mitochondrial cytochrome c oxidase subunit I (COI) and a project-specific portion of the mitochondrial NADH dehydrogenase 1 (ND1) genefor a selected sample of collected animals (Table 2). We sampled 26putative T. longicaudatus ‘‘short’’ from five ponds, 24 putative T.longicaudatus ‘‘long’’ from one pond, and 21 putative T. newberryi from11 ponds, including three ponds that also contained T. longicaudatus‘‘short.’’ We also included sequences from a single individual T.longicaudatus ‘‘short’’ hatched from desiccated soil collected approxi-mately 100 km west of our sample sites (Lordsburg Playa, NM). Finally,our analysis included 12 additional sequences for Triops spp. obtainedfrom GenBank. Both COI and ND1 sequences were obtained from aputative T. longicaudatus (Accession No. NC006079)) and from T.cancriformis (Accession No. NC004465). COI sequences were obtainedfor two additional T. cancriformis (EF675903, DQ369317), two uniden-tified Triops spp. from unknown localities (DQ310625, DQ310623), threeT. australiensis (EF189677, DQ889135, DQ310624), and three Lepidurusspp. as the outgroup (DQ310622, EF189669, AF209067).

DNA was isolated from all specimens using DNeasy (Qiagen, Valencia,California) tissue preparation kits. Primers for amplification were eithermodified from universal primers (COI - LCO-T 59-YTCAACAAATCA-TAAAGATATTGG, HCO-T 59-TAAACTTCMGGGTGACCRAAAAA-TCA; Folmer et al., 1994) or designed by the authors specifically fornotostracans (ND1F-T 59-ATTGCGAAAGGGYCCTAAT, ND1R-T 59-CCCTCYGCAAARTCAAAAGG). Typical 25 ml PCR reactions con-tained 5 ml GoTaq 53 Flexi PCR buffer (Promega, Fitchburg, WI), 2 mMMgCl2, 0.2 mM dNTP mixture (Promega), 5 mM of each primer, 0.1 UGoTaq DNA polymerase (Promega), and 1 ml template DNA solution.Reactions were cycled on a Bio-Rad thermocycler (Bio-Rad, Hercules,CA), starting with a 4 min denaturing step at 95 uC, followed by 35 cyclesof the following reaction: 95 uC for 15 sec, 50 uC for 15 sec, and 72 uC for30 sec. Reactions finished with a single 72 uC, 5 min elongation step. PCR

Fig. 1. Map of southern New Mexico, with labels of Triops spp.-containing playas (PL) and flood retention ponds (FP) sampled in summer2008. The shaded area represents the northern reaches of the ChihuahuanDesert, and is based on a map from the Chihuahuan Desert ResearchInstitute (Fort Davis, Texas; www.cdri.org/Desert/map.JPG).

470 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 31, NO. 3, 2011

products were cleaned using ExoSAP-IT (USB Corporation, Cleveland,OH) following the manufacturer’s protocols. Sequencing of PCR productswas performed in both directions by NMSU’s MOLBIO MolecularAnalysis Services (http://research.nmsu.edu/molbio/MOLB_2_Sequence/NMSU {{Index}}.htm).

Resultant sequences were visually pre-aligned and trimmed using theprogram CodonCode Aligner v3.0.1 (CodonCode Corporation, Dedham,MA) and aligned by eye using Mesquite (Maddison and Maddison, 2010).Uncorrected pairwise distances (p) were calculated in the program

MEGA4 (Tamura et al., 2007). These distances are calculated as thepercentage of sites differing between two individuals, and were calculatedfor all possible pairs of individuals. We constructed trees using Bayesianmethods with the program MrBayes v3.12 (Ronquist and Huelsenbeck,2003) and a GTR + G + I nucleotide substitution model, determined to bethe best according to the AIC criterion in Modeltest v3.7 (Posada andCrandall, 1998), with independent model parameters for the two genes,and allowing for different substitution rates for the three codon positions.We ran two simultaneous Markov Chain Monte Carlo (MCMC) searches

Fig. 2. Comparison of morphological features of the three forms of Triops collected in New Mexico in summer 2008. Carapace of (A) T. newberryi, (B) T.longicaudatus ‘‘long,’’ and (C) T. longicaudatus ‘‘short,’’ showing difference in color and mottling. Close-up of carapace of (D) T. newberryi, (E) T.longicaudatus ‘‘long,’’ and (E) T. longicaudatus ‘‘short,’’ showing miniscule teeth on surface and larger teeth on posterior margin. Posterior margin teethare labeled.


Table 2. Individuals used in molecular analysis. FP designates flood retention pond, PL designates playas and stock ponds. ‘‘–’’ denotes individuals fromwhich sequences for that gene were not obtained.

Morphological form Pond Sample name

Accession #

COI ND1

T. longicaudatus ‘‘long’’ PL 23 20III09_14 HQ908544 HQ908614T. longicaudatus ‘‘long’’ PL 23 20III09_5 HQ908545 HQ908615T. longicaudatus ‘‘long’’ PL 23 24VI10_11 HQ908546 HQ908616T. longicaudatus ‘‘long’’ PL 23 24VI10_13 HQ908547 HQ908617T. longicaudatus ‘‘long’’ PL 23 24VI10_15 HQ908548 HQ908618T. longicaudatus ‘‘long’’ PL 23 24VI10_18 HQ908549 HQ908619T. longicaudatus ‘‘long’’ PL 23 24VI10_2 HQ908550 HQ908620T. longicaudatus ‘‘long’’ PL 23 24VI10_5 HQ908551 HQ908621T. longicaudatus ‘‘long’’ PL 23 24VI10_6 HQ908552 HQ908622T. longicaudatus ‘‘long’’ PL 23 20III09_13 HQ908553 –T. longicaudatus ‘‘long’’ PL 23 24VI10_7 HQ908554 –T. longicaudatus ‘‘long’’ PL 23 24VI10_8 HQ908555 HQ908623T. longicaudatus ‘‘long’’ PL 23 26VI10_1 HQ908556 HQ908624T. longicaudatus ‘‘long’’ PL 23 26VI10_12 HQ908557 HQ908625T. longicaudatus ‘‘long’’ PL 23 26VI10_13 HQ908558 HQ908626T. longicaudatus ‘‘long’’ PL 23 26VI10_3 HQ908559 HQ908627T. longicaudatus ‘‘long’’ PL 23 26VI10_4 HQ908560 HQ908628T. longicaudatus ‘‘long’’ PL 23 26VI10_5 HQ908561 HQ908629T. longicaudatus ‘‘long’’ PL 23 26VI10_7 HQ908562 HQ908630T. longicaudatus ‘‘long’’ PL 23 26VI10_8 HQ908563 HQ908631T. longicaudatus ‘‘long’’ PL 23 26VI10_9 HQ908564 HQ908632T. longicaudatus ‘‘long’’ PL 23 2VI10_13 HQ908565 HQ908633T. longicaudatus ‘‘long’’ PL 23 2VI10_16 HQ908566 HQ908634T. longicaudatus ‘‘long’’ PL 23 2XII09_5 HQ908567 HQ908635T. longicaudatus ‘‘short’’ PL 05 26VI10_2 HQ908530 HQ908600T. longicaudatus ‘‘short’’ PL 05 19VII10_7 HQ908538 HQ908608T. longicaudatus ‘‘short’’ PL 05 19VII10_9 HQ908539 HQ908609T. longicaudatus ‘‘short’’ PL 05 19VII10_8 HQ908540 HQ908610T. longicaudatus ‘‘short’’ PL 05 9IV09_18 HQ908524 HQ908594T. longicaudatus ‘‘short’’ PL 06 20III09_4 HQ908522 HQ908593T. longicaudatus ‘‘short’’ PL 06 24VI10_10 HQ908527 HQ908597T. longicaudatus ‘‘short’’ PL 06 24VI10_3 HQ908528 HQ908598T. longicaudatus ‘‘short’’ PL 06 26VI10_11 HQ908529 HQ908599T. longicaudatus ‘‘short’’ PL 06 19VII10_10 HQ908533 HQ908603T. longicaudatus ‘‘short’’ PL 07 24VI10_4 HQ908536 HQ908606T. longicaudatus ‘‘short’’ PL 07 20III09_12 HQ908520 HQ908591T. longicaudatus ‘‘short’’ PL 07 20III09_16 HQ908523 –T. longicaudatus ‘‘short’’ PL 08 20III09_6 HQ908519 HQ908590T. longicaudatus ‘‘short’’ PL 07 20III09_9 HQ908521 HQ908592T. longicaudatus ‘‘short’’ PL 08 24VI10_17 HQ908526 HQ908596T. longicaudatus ‘‘short’’ PL 08 19VII10_11 HQ908534 HQ908604T. longicaudatus ‘‘short’’ PL 08 24VI10_12 HQ908537 HQ908607T. longicaudatus ‘‘short’’ PL 08 23VI10_1 HQ908541 HQ908611T. longicaudatus ‘‘short’’ PL 08 8VI10_1 HQ908542 HQ908612T. longicaudatus ‘‘short’’ PL 09 24VI10_14 HQ908517 HQ908588T. longicaudatus ‘‘short’’ PL 09 24VI10_16 HQ908525 HQ908595T. longicaudatus ‘‘short’’ PL 09 14VI10_1 HQ908531 HQ908601T. longicaudatus ‘‘short’’ PL 09 14VI10_10 HQ908532 HQ908602T. longicaudatus ‘‘short’’ PL 09 19VII10_12 HQ908535 HQ908605T. longicaudatus ‘‘short’’ PL 09 20III09_18 HQ908518 HQ908589T. longicaudatus ‘‘short’’ Lordsburg Playa 9IV09_13 HQ908543 HQ908613T. newberryi FP 02-1 20III09_17 HQ908515 HQ908586T. newberryi FP 02-2 9IV09_1 HQ908516 HQ908587T. newberryi FP 03-1 20III09_15 HQ908513 HQ908584T. newberryi FP 03-2 9IV09_2 HQ908514 HQ908585T. newberryi FP 04-1 9IV09_3 HQ908511 HQ908582T. newberryi FP 04-2 9IV09_7 HQ908512 HQ908583T. newberryi FP 06-1 9IV09_8 HQ908509 HQ908580T. newberryi FP 06-2 9IV09_4 HQ908510 HQ908581T. newberryi PL 05-1 9IV09_17 HQ908505 HQ908576T. newberryi PL 05-2 26VI10_14 HQ908507 HQ908578T. newberryi PL 05-3 26VI10_15 HQ908508 HQ908579T. newberryi PL 06 20III09_10 HQ908504 HQ908575T. newberryi PL 07-1 20III09_8 HQ908503 HQ908574T. newberryi PL 07-2 24VI10_1 HQ908506 HQ908577T. newberryi PL 10-1 20III09_7 HQ908501 HQ908572T. newberryi PL 10-2 9IV09_9 HQ908502 HQ908573T. newberryi PL 11 9IV09_10 HQ908500 HQ908571

Sample name


with 4 chains each for 20 million generations, sampling both runs every1000 generations. We constructed likelihood trees using Garli 1.0 (Zwickl,2006) with a GTR + G + I nucleotide substitution model, and allowing fordifferent rates of evolution for the three codon positions. Twelveindependent tree searches were run for 5,000,000 generations each.Likelihood bootstrap support values were also obtained using Garli 1.0,running 100 pseudoreplicates, with each pseudoreplicate consisting of fourindependent tree searches run for 5,000,000 generations. We constructedparsimony trees in TNT (Goloboff et al., 2008) using a heuristic searchwith tree bisection and reconnection (TBR) branch-swapping and 1000random-addition searches. We also calculated bootstrap support values,running 1000 pseudoreplicates (each with 10 random-addition searches).

RESULTS

From July to September 2008 we collected 1,161specimens of Triops spp. from 14 playas and floodretention ponds in the vicinity of Las Cruces, NM (Table 1;Fig. 1). Sex and number of legless rings were determinedfor all collected individuals, while carapace length, numberof teeth on the posterior margin of the carapace andpresence/absence of eggs were determined for 660 of thecollected ovisac-bearing individuals and 92 males. Finally,eggs were counted and measured, and egg and clutchvolumes (# eggs * mean egg volume) were calculated for129 ovisac-bearing individuals. Following Sassaman et al.(1997), New Mexican Triops spp. were divided into threegroups (see Table 3 for a summary and distribution of threeforms): T. newberryi (Fig. 2A), T. longicaudatus ‘‘long’’(Fig. 2B), and T. longicaudatus ‘‘short’’ (Fig. 2C) based onthe mean number of legless rings (Fig. 3, see Fig. 4 forphotos of legless rings) and sexual composition of thepopulation.

Triops newberryi was the only notostracan present in 8of 14 sample sites, including all four flood retention ponds,T. longicaudatus ‘‘long’’ was the only notostracan at onesite, and T. longicaudatus ‘‘short’’ was the sole notostracanin one site. Four playa lakes contained mixed populationsof T. newberryi and T. longicaudatus ‘‘short’’ (Table 3).

Morphological Measurements

Female T. longicaudatus ‘‘long’’ had the greatest number oflegless rings, with an average of 9.5 (range 5 9-10; Fig. 3).

Triops newberryi ovisac-bearing individuals had a mean of8.1 (7-10) legless rings, and T. longicaudatus ‘‘short’’ovisac-bearing individuals had a mean of 6.0 (5-7) leglessrings. Mean number of legless rings for males were 11.4(11-12.5) for T. longicaudatus ‘‘long’’ and 9.6 (8-11) for T.newberryi, both significantly greater (P , 0.0001) than thefemale of the same morphological form. Ovisac-bearingindividuals of the three forms differed significantly in thenumber of teeth on the posterior margin (P , 0.0001;Fig. 3): T. longicaudatus ‘‘short’’ (mean # teeth 5 42.2;range 5 32-53) had the most, followed by T. newberryi(36.8; 29-48), and T. longicaudatus ‘‘long’’ (30.4; 25-38).Male T. longicaudatus ‘‘long’’ had a mean of 29.4 posteriormargin teeth (range 5 25-33) and T. newberryi had a meanof 37.8 (27-46), neither of which were significantlydifferent than their respective ovisac-bearing individuals(P 5 0.219 and P 5 0.111, respectively). Triops newberryiindividuals had the longest maximum carapace length(16.6 mm), followed by T. longicaudatus ‘‘long’’(15.9 mm), and T. longicaudatus ‘‘short’’ (14.8 mm).Individual T. newberryi were significantly longer (P ,0.0001) than coexisting T. longicaudatus ‘‘short’’ individ-uals, and displayed fewer posterior margin carapace teeth(P , 0.0001). Carapace lengths of male T. longicaudatus‘‘long’’ were not significantly longer than female carapacelengths (P 5 0.273), but T. newberryi males weresignificantly longer (P , 0.001) than ovisac-bearingindividuals. The mystax was present in all examined males,and absent in all examined individuals bearing ovisacs.

Carapace length was significant as a covariate of clutchsize. Corrected for differences in carapace length, clutchsize differed significantly (P , 0.0001) among the threeforms, while individual egg volume differed significantly(P , 0.0001; Fig. 5) between T. longicaudatus ‘‘short’’ andT. newberryi and between the former and T. longicaudatus‘‘long.’’ The eggs of the T. longicaudatus ‘‘short’’ weresmallest, and at a comparable body size, T. longicaudatus‘‘long’’ carried the most eggs, followed by T. longicaudatus‘‘short’’ and then T. newberryi. The smallest individualfound carrying eggs was 6.7 mm (carapace length) in T.longicaudatus ‘‘short,’’ 8.7 mm in T. newberryi, and,11.6 mm in T. longicaudatus ‘‘long.’’ Egg color also varied

Morphological form Pond Sample name

Accession #

COI ND1

T. newberryi PL 12-1 20III09_11 HQ908498 –T. newberryi PL 12-2 9IV09_14 HQ908499 HQ908570T. newberryi PL 17-1 9IV09_16 HQ908496 HQ908568T. newberryi PL 17-2 9IV09_15 HQ908497 HQ908569Triops sp. T. long. NC006079 NC006079 NC006079T. cancriformis T. cancriformis NC004465 NC004465 NC004465T. cancriformis T. cancriformis EF675903 EF675903 –T. cancriformis T. cancriformis DQ369317 DQ369317 –T. australiensis T. australiensis EF189677 EF189677 –T. australiensis T. australiensis DQ889135 DQ889135 –T. australiensis T. australiensis DQ310624 DQ310624 –Triops sp. Triops n. sp. DQ310625 DQ310625 –Triops sp. Triops sp. DQ310623 DQ310623 –Lepidurus couesii Lepidurus couessi DQ310622 DQ310622 –Lepidurus apus Lepidurus apus EF189669 EF189669 –Lepidurus sp. Lepidurus sp. AF209067 AF209067 –

Table 2. Continued.

Pond


among live individuals of the three forms, and this colordifference remained unchanged after more than two yearsof preservation in 95% EtOH (Fig. 4). Both T. long-icaudatus forms carried dark red eggs, while the eggs of T.newberryi varied from light yellow, to orange, to pinkish-orange, but were not as dark as the eggs of either T.longicaudatus form. Coexisting populations of T. long-icaudatus ‘‘short’’ and T. newberryi differed in clutch size(P , 0.001) and egg volume (P , 0.0001).

Qualitatively, the carapace surface differed between T.longicaudatus ‘‘long,’’ and the other two forms, with theformer having a smooth carapace, with few to no teethother than those on the posterior margin, while thecarapaces of T. longicaudatus ‘‘short’’ and T. newberryiwere covered in miniscule teeth (Fig. 2). These small teethwere also often present among the larger posterior marginteeth in T. newberryi and T. longicaudatus ‘‘short’’ (andwere counted if emanating directly from the margin), butwere rarely present in T. longicaudatus ‘‘long.’’ Addition-ally, the carapace of T. longicaudatus ‘‘long’’ was pale,with little to no mottling, while both other forms haddarker, sometimes mottled, carapaces (Fig. 2).

Apparent sex ratios differed among the three forms; T.newberryi populations were predominantly ovisac-bearingindividuals, with mean per pond proportion of males thatranged from 0.0 to 0.26 (Table 3), T. longicaudatus ‘‘short’’populations were entirely comprised of ovisac-bearingindividuals, and the single collection of T. longicaudatus‘‘long’’ had a sex ratio close to 1:1 (proportion of males 50.45, see Table 3).

Molecular Analysis

Our molecular analysis used a 710 bp COI mitochondrialDNA sequence (GenBank accession HQ908496-HQ908-567), and a 517 bp ND1 mitochondrial DNA sequence(GenBank accession HQ908568-HQ908635), from 84individuals (see Table 2). The inference of insertion/deletion events was not necessary to align either geneportion, i.e., there were no gaps in the finished alignment.The 27 individuals identified as T. longicaudatus ‘‘short’’shared a single haplotype for both COI and ND1. Thesingle individual from Lordsburg Playa exhibited the samehaplotype. The 24 individual T. longicaudatus ‘‘long’’ alsoshared a single haplotype for both COI and ND1 but bothwere different from the haplotypes observed for T.longicaudatus ‘‘short.’’ The 21 individual T. newberryifrom 12 ponds consisted of seven COI haplotypes and sixND1 haplotypes. Some individuals shared gene-specifichaplotypes, so with the two gene sequences combined,seven overall haplotypes were present. Five pondscontained individual T. newberryi that shared haplotypes;two of those same ponds also contained individuals whosesequences differed by as much as 1.6%. Uncorrectedpairwise genetic distances (p) for the combined data rangedfrom 0.0% to 1.6% among T. newberryi individuals (0.0%-1.6% for COI, 0.0%-1.7% for ND1). Uncorrected pairwisedistances between T. longicaudatus ‘‘short’’ and T. long-icaudatus ‘‘long’’ were 5.6% (5.3% for COI, 5.7% forND1), and ranged from 2.2% to 2.7% (2.3%-2.8% for COI,3.3%-4.5% for ND1)between T. longicaudatus ‘‘short’’ andT. newberryi, and from 4.4% to 5.0% (4.1%-5.1% for COI,4.1%-5.7% for ND1)between T. longicaudatus ‘‘long’’ andT. newberryi. There was only one amino acid change in theCOI sequence, in a single individual of T. newberryi, butseven amino acid substitutions were inferred in the ND1sequence: three in T. longicaudatus ‘‘long’’ and two in eachof T. newberryi, and T. longicaudatus ‘‘short.’’

The Bayesian analysis conducted using MrBayes (Ron-quist and Huelsenbeck, 2003) resulted in 19,000 sampledtrees after removing 1000 trees to account for burn-in (theruns converged to a stationary distribution after approxi-mately 500 sampled trees). The consensus reconstruction ofthese 19,000 trees, including posterior probabilities, isshown in Figure 6. The analysis was consistent withmonophyly of T. newberryi and T. longicaudatus ‘‘long,’’with additional support for a clade comprising T. newberryiand T. longicaudatus ‘‘long.’’ Triops longicaudatus ‘‘short’’forms an unresolved group sister to the later clade. Themost likely (Ln 5 24623.73726) tree resulting from theanalysis using Garli (Zwickl, 2006; tree is not shown,bootstrap values shown in Fig. 6) supports the monophylyof T. longicaudatus ‘‘long’’ and a clade comprising T.

Fig. 3. Distribution of number of legless abdominal rings and number ofcarapace posterior margin teeth for ovisac-bearing individuals among thethree morphs of the genus Triops collected in New Mexico in summer2008. Black bars represent T. newberryi, grey bars represent T.longicaudatus ‘‘short,’’ and white bars represent T. longicaudatus ‘‘long.’’


newberryi and T. longicaudatus ‘‘long.’’ Reducing thenumber of identical sequences had no effect on eithertopology or branch support in the Garli analyses. Theparsimony search using TNT (Goloboff et al., 2008)resulted in 1810 trees with a length of 417. In contrast toboth likelihood-based analyses, the parsimony analysisstrongly supported the monophyly of all three species

(Fig. 7), as well as supporting a clade comprising T.newberryi and T. longicaudatus ‘‘short.’’

DISCUSSION

The notostracan fauna of the northern Chihuahuan Desertconsists of three genetically and morphologically distinct

Fig. 4. Comparison of morphological features of the three forms of Triops collected in New Mexico in summer 2008. Ventral portion of abdomen of (A)T. newberryi, (B) T. longicaudatus ‘‘long,’’ and (C) T longicaudatus ‘‘short,’’ showing difference in number of legless rings, with rings numbered ascounted by authors. Eggs of (D) T. newberryi, (E) T. longicaudatus ‘‘long’’ and (F) T. longicaudatus ‘‘short,’’ showing differences in egg color amongspecies, and egg color variation within T. newberryi.


Tab

le3

.S

um

mar

yo

fv

aria

tion

inm

orp

ho

log

yan

dfe

cun

dit

yin

ov

isac

-bea

rin

gin

div

idual

so

fth

eth

ree

mo

rph

olo

gic

alfo

rms

of

Tri

op

sco

llec

ted

inep

hem

eral

wat

ers

of

the

No

rther

nC

hih

uah

uan

des

ert.

#M

easu

red

#le

gle

ssri

ngs

Car

apac

ele

ngth

(mm

)#

post

.m

argin

cara

pac

ete

eth

Egg

vol

(mm

3)

Clu

tch

size

(#eg

gs)

Pro

port

ion

fem

ale

Mea

n(R

ange)

Mea

n(R

ange)

Mea

n(R

ange)

Mea

n(R

ange)

Mea

n(R

ange)

Tri

op

sn

ewber

ryi

FP

02

22

8.1

1(7

-9)

13

.43

(11

.3-1

5.1

)3

6.0

5(3

0-4

3)

0.0

31

(0.0

27-0

.03

5)

94

.60

(78

-11

7)

0.7

3F

P0

36

58

.10

(7-9

)1

1.7

9(1

0.7

-14

.1)

37

.86

(31

-47

)0

.026

(0.0

23-0

.03

)6

5.0

0(5

2-7

6)

0.7

7F

P0

41

98

.13

(7-9

)6

.94

(5.9

-7.9

)3

5.2

1(2

9-4

0)

––

0.7

3F

P0

67

7.7

1(7

-8)

9.3

9(7

.9-1

1.2

)3

3.1

4(3

2-3

5)

––

0.7

8P

L0

53

88

.13

(7.5

-9.5

)5

.63

(4.6

-6.7

)3

7.6

6(3

3-4

4)

––

0.7

9P

L0

65

28

.53

(7.5

-10)

10

.26

(7.5

-12

.5)

34

.94

(30

-43

)0

.028

(0.0

23-0

.03

2)

46

.09

(34

-60

)0

.96

PL

07

78

8.0

7(7

.5-9

)1

0.0

7(7

.3-1

4.4

)3

9.0

4(3

0-4

8)

0.0

31

(0.0

26-0

.03

8)

43

.69

(21

-71

)0

.89

PL

08

19

.00

12

.29

36

.00

––

1.0

0P

L1

04

47

.98

(7.5

-10)

13

.50

(10

.7-1

6.6

)3

6.7

3(3

1-4

2)

0.0

35

(0.0

3-0

.03

9)

84

.30

(46

-14

0)

0.7

4P

L1

14

67

.76

(7-8

.5)

13

.88

(10

.4-1

5.6

)4

1.1

1(3

4-4

8)

0.0

35

(0.0

29-0

.04

1)

88

.88

(30

-13

5)

0.8

5P

L1

23

88

.46

(8-9

)1

3.4

9(1

1.3

-15

.4)

34

.83

(30

-41

)0

.031

(0.0

29-0

.03

4)

59

.57

(37

-87

)1

.00

PL

17

17

8.2

1(7

.5-9

)1

7.1

0(1

2.5

-20

.2)

37

.38

(32

-42

)0

.032

(0.0

28-0

.03

5)

21

2.0

0(1

33

-23

3)

0.9

4A

ll4

27

8.1

3(7

-10

)1

1.1

6(4

.6-2

0.2

)3

6.7

6(2

9-4

8)

0.0

31

(0.0

23-0

.04

1)

86

.77

(21

-23

3)

0.8

6

Tri

op

slo

ng

icau

da

tus‘‘

sho

rt’’

PL

05

45

5.9

8(5

.5-6

.5)

5.1

4(3

.8-6

.6)

45

.76

(38

-53

)–

–1

.00

PL

06

14

6.3

9(6

-7)

8.9

8(7

.7-1

0.2

)4

2.7

9(3

9-4

7)

0.0

21

(0.0

18-0

.02

7)

46

.80

(43

-64

)1

.00

PL

07

48

5.9

8(5

.5-6

.5)

8.4

1(6

.9-1

0.3

)4

1.6

5(3

5-4

9)

0.0

24

(0.0

18-0

.02

8)

52

.85

(29

-81

)1

.00

PL

08

75

5.7

4(5

-6.5

)1

2.0

6(8

.0-1

4.7

5)

38

.83

(32

-47

)0

.028

(0.0

19-0

.03

4)

92

.44

(47

-14

5)

1.0

0P

L0

93

06

.12

(5.5

-7)

7.6

7(6

.7-9

.5)

42

.03

(35

-46

)0

.019

(0.0

15-0

.02

1)

27

.00

(18

-43

)1

.00

All

21

26

.04

(5-7

)8

.45

(3.8

-14

.75

)4

2.2

1(3

2-5

3)

0.0

23

(0.0

15-0

.03

4)

54

.77

(18

-14

5)

1.0

0

Tri

op

slo

ng

icau

da

tus‘‘

long’’

PL

23

21

9.5

0(9

-10

)1

2.9

5(1

1.0

-15

.9)

30

.35

(25

-38

)0

.030

(0.0

27-0

.03

4)

23

2.5

0(1

76

-26

1)

0.5

4


forms, all potentially valid species. These three forms hadbeen previously delineated by Sassaman (1991) andSassaman et al. (1997) and recognized in other recentstudies (Maeda-Martınez et al., 2000a; Obregon-Barboza etal., 2001; Murugan et al., 2002). Our study presentsadditional evidence suggesting the distinctiveness of allthree taxa.

Sassaman et al. (1997), using allozymes and meristicdata, divided Triops collected from sites throughout thesouthwestern United States into two species, with onespecies further separated into two forms based onmorphology (long-abdomen and short-abdomen) and sexratio (gonochoric vs. female only). They identified thislatter species as T. longicaudatus based on similarities inmorphology between their long-abdomened, gonochoricmorph and LeConte’s (1846) description and Linder’s(1952) re-description of the species. However, althoughthey could differentiate the two morphs genetically Sassa-man et al. (1997) did not deem this sufficient to distinguishthem as separate species without additional genetic supportusing a larger sampling of reproductive and morphologicaltypes. Sassaman et al. (1997) identified the second speciesas T. newberryi, choosing this affinity over the morpho-logically similar T. aequalis because of Packard’s (1871,1883) account of the distribution of Triops spp. (westernUS for T. newberryi; Kansas, eastern Mexico for T.aequalis) and page precedence (Packard, 1871), despitethe fact that Packard’s description notes that T. aequalisovisac-bearing individuals have 8 legless rings (as doSassaman’s specimens) while T. newberryi ovisac-bearing

individuals have 10 legless rings. Our T. newberryiaveraged eight legless rings, and we found only a singleovisac-bearing individual (out of 668 examined) in T.newberryi-occupied ponds with 10 legless rings. Wetherefore question the reliance of geographic range overmorphological similarity for species identity by Sassamanet al. (1997). However, to minimize confusion, we willcontinue to refer to this form as T. newberryi for the presentpaper, but note that a reexamination of Packard’s typespecimens is needed.

More recently, Maeda-Martinez et al. (2000b) examineddigestive enzyme differences between the two morphs of T.longicaudatus, and found differences in overall proteolyticfunction, in the number of molecular forms of the digestiveenzyme trypsin, and in the inhibitability of these forms byphenylmethanesulfonyl fluoride (PMSF) and 10 mM tosyl-Lys chloromethyl ketone (TLCK). Murugan et al. (2002),using morphology and portions of two mitochondrial rRNAgenes from seven populations, i.e., ponds, of T. long-icaudatus ‘‘long’’ and ‘‘short’’ and 2-7 individuals perpopulation, recognized six of their seven sampled popula-tions as separate phylogenetic species because all geneticvariation was between populations; there was no variationamong individuals from the same pond. Their study weaklysupported a sister relationship between a gonochoric T.longicaudatus ‘‘long’’ population, and a female-only T.longicaudatus ‘‘short’’ population, to the exclusion of theremainder of the sampled populations, which consisted ofboth male-less and androdioecious T. longicaudatus ‘‘short.’’However, this interpretation is uncertain because the authorswere not always clear in identifying the morphological formof all their populations and their sampled gene segmentsshowed very little variation. The only clade in their resultanttrees with high (. 90%) bootstrap support is one composedof all the North American populations.

Genetic Differences

Our results (Figs. 6, 7) support the monophyly of one, two,or all three forms depending upon the analysis type, whichare additionally incongruent in their support of relation-ships among the three groups. The distinctiveness of thethree forms (and the unresolved relationships among theforms) is supported by inferred amino acid differences inthe ND1 gene, with T. newberryi and T. longicaudatus‘‘short’’ each having two unique amino acid changes, and T.longicaudatus ‘‘long’’ having three unique changes. Triopslongicaudatus ‘‘short’’ and T. longicaudatus ‘‘long’’ bothexhibited no within-form genetic variation. While the lackof variation for T. longicaudatus ‘‘long’’ may be due in partto all sequenced individuals being collected from the samesource pond, T. longicaudatus ‘‘short’’ was collected fromsix ponds. Triops newberryi, in contrast, had a greater, butstill relatively low (0-1.6%) amount of genetic variation,both within and among ponds. Variation of up to 1.6%existed within ponds, while several closely relatedindividuals inhabited different ponds (Figs. 6, 7). Theoccurrence of multiple haplotypes within a pond suggeststhat multiple colonization events may have occurred atthose locations. Pairwise distances among forms wassubstantially higher than within-form distances, with T.

Fig. 5. Clutch number and mean egg volume as a function of carapacelength (CL) for individual ovisac-bearing individuals of the three morphsof the genus Triops collected in New Mexico in summer 2008. Ten eggsfrom each female were used to calculate mean egg volume. Black-filledcircles represent T. newberryi, grey-filled circles represent T. longi-caudatus ‘‘short,’’ and white-filled circles represent T. longicaudatus‘‘long.’’


newberryi differing from T. longicaudatus ‘‘short’’ and T.longicaudatus ‘‘long’’ by at least 2.2% (2.3% for COI) and4.4% (4.1% for COI) respectively, and the latter two formsdiffering by 5.6% (5.3% for COI). These within-formdistances are comparable to distances found using COI intwo other species-complexes of Triops, but among-formdistances are substantially lower in our analyses. Zierold etal. (2007) compared COI sequences among 74 individualsof T. cancriformis, finding Kimura 2-parameter (k2p)distances ranging from 0.0-1.6% (corresponding to 0.0-1.9% uncorrected p-distances). They found k2p distancesgreater than 10% between T. cancriformis and T.mauritanicus [although T. mauritanicus had officially beenelevated from T. cancriformis mauritanicus by Korn et al.(2006), it was still regarded a subspecies of T. cancriformisby Zieorold et al. (2007)]. Similarly, Murugan et al. (2009)found COI distances among three putative Australianspecies of Triops were greater than 10%, but found novariation within the two species with multiple sampledsequences. Additionally, a review of 852 crustacean COIsequences by Lefebure et al. (2006) found within-speciesdistances ranging from 0.0-7.9% and among species(within genera) distances ranging from 15.4-33.3%.Compared with these studies, our genetic distances do notstrongly support the separation of these forms into species.

There is a striking discordinance between the likelihood-based results and the parsimony-based results, with little tono support for the monophyly of either T. longicaudatus‘‘short’’ or T. newberryi in the former analyses but strongsupport in the latter analysis. Definitive resolution of thephylogenetic relationships among the forms will requireadditional genetic analyses with a greater range of variation.Work is presently in progress on microsatellites that willfacilitate resolving the ambiguity of the present results.

Differences in Morphology

Although we followed Sassaman et al. (1997) in assigningindividuals to one of three forms of Triops spp. based onthe number of legless rings and sexual composition of thepopulation, these three lineages can be distinguished byadditional morphological characters. Despite both Linder(1952) and Longhurst (1955c) considering it to be a‘‘useless’’ systematic character due to high levels ofvariation within species, the number of legless abdominalrings was remarkably consistent within forms (validated byother forms of evidence, see below), and we agree witholder (Packard, 1871, 1883; Rosenberg, 1847) and newer(Sassaman et al., 1997; Maeda-Martınez et al., 2000a)researchers who considered it very useful.

While there was overlap, there was a significantdifference among the three forms in the mean number ofteeth on the posterior margin of the carapace, with T.longicaudatus ‘‘short’’ having the most, followed by T.newberryi, and T. longicaudatus ‘‘long’’ having the least.This character was used by Murugan et al. (2009) indistinguishing two putative Australian species of Triops,although, similar to COI distances, differences were largerthan among our forms, with no overlap between species.This character has not been considered useful in earlierexaminations of North American Triops spp., and by itself

cannot be used to identify individual specimens, buttogether with other characters it appears to be useful inseparating the three forms.

Under magnification there was also a noticeable differ-ence in the dorsal surface of the carapace among the threeforms (Fig. 2). Both T. longicaudatus ‘‘short’’ and T.newberryi exhibited numerous miniscule teeth emanatingfrom the dorsal carapace, while the surface of T. long-icaudatus ‘‘long’’ was smooth, with no dorsal teeth presenton the carapace. This character has not been mentionedpreviously in North American Triops spp. and was examinedby Murugan et al. (2009) but was not useful in distinguishingspecies. In general, the gross morphology of T. long-icaudatus ‘‘short’’ and T. newberryi are much more similarto each other than either is to T. longicaudatus ‘‘long.’’

The most obvious character observed when we firstexamined collected individuals was the consistent differ-ence in egg color between T. newberryi and T. long-icaudatus ‘‘short.’’ Egg color was strongly correlated tolegless ring counts: coexisting individuals separated intoforms solely by number of legless rings invariably also haddifferent colored eggs. We have found no mention of eggcolor in any previous discussion of North American Triopsspp., which is puzzling for such an obvious character fordistinguishing different morphs, especially in co-existingpopulations. We have found only a single reference to eggcolor in T. longicaudatus at all, in a redescription of Koreanspecimens (Yoon et al., 1992), in which egg color wasdescribed as brown. We found egg color to be as reliable amethod of differentiation as counting rings in mixedpopulations, and much faster to diagnose, albeit limitedto egg-carrying individuals.

Males of T. longicaudatus ‘‘long’’ and T. newberryi hadsignificantly more legless rings than their ovisac-bearingcounterparts, suggesting an additional character for sexualdifferentiation. Sexual dimorphism in the number of leglessrings has been noted by other researchers (Packard, 1883;Linder,1952; Longhurst, 1955c; Sassaman et al., 1997).

Differences in Growth and Fecundity

Comparing size among the three forms, the largestindividual found was in T. newberryi, followed by T.longicaudatus ‘‘long,’’ and then T. longicaudatus ‘‘short.’’However, when comparing overall mean carapace lengthsamong the three forms, T. longicaudatus ‘‘long’’ werelargest, following by T. newberryi and T. longicaudatus‘‘short.’’ Comparing growth (and reproduction) among thepopulations is complicated by the length of time theephemeral wetland has been inundated; the longer theinundation, the larger the size of inhabiting individuals, andnot all ponds fill with water at the same time. Triopslongicaudatus ‘‘long’’ individuals were from a single pondand had little variation in carapace lengths. Both T.newberryi and T. longicaudatus ‘‘short’’ were found inmultiple ponds with presumably varying times sinceinundation, and these populations exhibited large amountsof carapace-length variation among ponds. Different agesacross ponds would be expected to yield different sizedspecimens in different ponds and it would affect thecalculation of mean carapace length for the form across


ponds. Desiccated eggs of Triops spp. (encysted embryos)hatch within days of inundation with water (Takahashi,1977; Scott and Grigarick, 1979; Su and Mulla, 2002), andindividuals grow quickly; knowing the amount of time aplaya has contained water is crucial for comparing growthand fecundity among different species.

Unfortunately, pond fill dates were not known to theappropriate scale (within several days) for our samples. Weresolved this issue by examining specimens from pondswith coexisting morphological forms. Scott and Grigarick(1979) showed that cysts of all Triops spp. generally hatchwithin 1-3 days of immersion in water, with no differencesin time to hatch among forms, so coexisting forms would

presumably hatch contemporaneously, allowing suchpopulations to be directly compared.

We were thus able to compare the T. newberryi and T.longicaudatus ‘‘short’’ populations that coexisted in threeplayas. The fourth playa inhabited by both species onlycontained a single individual in the sample that wasprovisionally identified as T. newberryi, and was thereforenot included in the comparison. In all cases where theforms co-inhabited the same pond (Table 3) T. newberryihad longer carapace lengths, larger eggs, and smaller clutchsizes (number of eggs). Additionally, the smallest individ-uals observed carrying eggs in T. longicaudatus ‘‘short’’were smaller than the smallest egg-carrying T. newberryi.

Fig. 6. Majority rules consensus phylogram of 19,000 trees resulting from a Bayesian analysis of 1,217 bp and 84 specimens of Notostraca. Numbers ontree show branch support: the left number is Bayesian posterior probability obtained from the program MrBayes, the right number is likelihood bootstrapvalue, obtained from the program Garli. Posterior probabilities are shown if greater than 0.75, likelihood bootstrap values are shown if greater than 50%.Collapsed branches have posterior probabilities less than 0.5. Terminal taxa sampled by authors are labeled by pond in which they were found (FP 5 floodretention pond; PL 5 playa); numbers after pond label are for identification of individual sample. Small parallel dashes on branches among the outgroupsrepresent removal of 20% of the branch length to facilitate viewing of the ingroup. Branch lengths are proportional to the expected number of substitutionsper site that has occurred along that branch. Bars uniting the three morphs are at right.


What cannot be inferred from our limited data is how longeggs are carried in the egg case, and how quicklyindividuals can produce another brood after releasing itseggs. However, there have been several studies on growthand reproduction of North American Triops spp. examiningthis issue. Scott and Grigarick (1978) examined fecundityof what they called T. longicaudatus (either T. newberryi orT. longicaudatus ‘‘short’’) in the laboratory and found thateggs first appeared in sacs 13 days post-hatch (9-12 days inoutdoor ponds), and were carried for 19 hours to severaldays, with deposition occurring between molts. Weeks(1990) and Weeks and Sassaman (1990) found that T.longicaudatus ‘‘short’’ produced more eggs over its lifetimethan T. newberryi (both in pure and mixed lab populations)

but grew slower and to a smaller maximum size. Obregon-Barboza et al. (2001) similarly tested growth andreproductive rates in Baja, Mexico populations of whatthey referred to as the ‘‘short’’ and ‘‘long’’ form of Triopssp. From their descriptions of the two forms it is possibletheir ‘‘short’’ form is equivalent to our T. longicaudatus‘‘short,’’ and their ‘‘long’’ form may be equivalent to our T.newberryi. Obregon-Barboza et al. (2001) showed that theirlong form grew faster than the short form but the latterdeposited many more eggs in its lifetime, despite havingsmaller clutch sizes and smaller maximum body size. Theyfurther showed that their short form (T. longicaudatus‘‘short’’?) accomplished this by depositing multiple broodsdaily, and could deposit over 1200 eggs in a 24 h period

Fig. 7. Strict consensus cladogram of the 1,810 shortest trees resulting from a parsimony analysis using TNT. Numbers on tree show bootstrap branchsupport. Values are shown if values are greater than 50%. Labeling as in Figure 4.


(mean 5 306 eggs per 24 h), while their long form (T.newberryi?) deposited at most a single brood daily (mean5 7.9 eggs per 24 h). Our results and previous researchsuggests that although T. newberryi appears to grow fasterand/or to a larger maximum size, T. longicaudatus ‘‘short’’appears to mature at a smaller size and lays relativelygreater numbers of smaller eggs.

The morphological characters distinguishing T. new-berryi from T. longicaudatus ‘‘short’’ were consistent insamples from multiple ponds, including ponds in which theforms coexisted, supporting the separate identity of theseforms. However, we could not reach the same conclusionfor T. longicaudatus ‘‘long,’’ which has only been collectedfrom a single playa lake in our sampling. This site (PL23)was more alkaline and saline by an order of magnitude(Table 1) than any other sampled site, and the waterappearance was a pale milky color. It is unknown ifdifferences in overall appearance (lack of carapace color,smoothness of carapace) and specific morphologicalcharacters are due to the different water conditions ofPL23, or if they are lineage-based. Attempts to find thismorph in additional ponds are currently underway.

Differences in Reproductive System

Sassaman et al. (1997) differentiated the three forms ofNorth American Triops initially and primarily by theirreproductive system. Populations of T. longicaudatus‘‘long’’ were found with generally equal numbers of malesand ovisac-bearing individuals, and breeding studiesshowed that ovisac-bearing individuals did not produceviable eggs unless in the presence of males. This form wasreferred to as ‘‘gonochoric.’’ The authors assumed thatfertilization by males had occurred when viable eggs wereproduced, although actual mating was never observed, andthere was no control treatment (co-occurring ovisac-bearing individuals), and no histology was performed toverify the absence of testicular tissue in ovisac-bearingindividuals. Populations of T. longicaudatus ‘‘short’’ wereinvariably female-only (Sassaman et al., 1997). Theseovisac-bearing individuals were believed to be functionallyhermaphroditic, and could produce viable eggs in theabsence of other individuals. While some researchers havelabeled some male-less reproducing Triops individualsparthenogenetic (Zaffagnini and Trentini, 1980), allozymestudies in at least some species have shown thatheterozygotes produce both homozygotic and heterozygoticoffspring, suggesting that self-fertilization occurs althoughthis can also occur with several modes of automicticparthenogenesis (Pearcy et al., 2006); this was found forboth T. longicaudatus ‘‘short’’ and T. newberryi (Sassamanet al., 1997). Populations of T. newberryi were biasedtoward ovisac-bearing individuals, but frequently containedmales. As in T. longicaudatus ‘‘short’’ these ovisac-bearingindividuals could produce viable eggs in the absence ofmales, and were considered hermaphroditic.

Hermaphroditism and Androdioecy

Until recently, histological examination of sexual identityin Triops spp. has been primarily limited to the Old World

species T. cancriformis Bosch, 1881. These studies(Bernard, 1892, 1895; Longhurst, 1955b; Zaffagnini andTrentini, 1980) have demonstrated that adult individualslacking egg sacs also lack ovarian tissue, and are likely tobe males. While individuals with ovisacs invariably haveovarian tissue, some also have testicular tissue in the formof sometimes-large testes lobes. These individuals cansuccessfully reproduce in the absence of males and arethought to be self-fertilizing hermaphrodites, althoughhermaphroditism can only be distinguished from partheno-genesis with thorough genetic study (Pearcy et al., 2006). Ina study of putative T. longicaudatus from Japan, Akita(1966, 1971) found that populations either only containedovisac-bearing individuals -or were mixed sexed (and oftenmale-dominated). Individuals in the single-sex populationshad ovarian tissue and either large, well-developed testeslobes, or small, rudimentary lobes. Ovisac-bearing individ-uals in the mixed sex populations did not have testes lobes,and could not produce viable eggs in the absence of males.

Longhurst (1954, 1955b) examined several species ofTriops, including putative T. longicaudatus from Califor-nia, and found females, males, and hermaphrodites, butgave no information regarding the morphology of theexamined individuals, or their population source. A recentcytological study (Garcıa-Velazco et al., 2009) of MexicanTriops spp. examined long (seven or more legless rings)and short (less than seven legless rings) populations. Theshort-abdomened populations, which appear to be consis-tent with our T. longicaudatus ‘‘short,’’ consisted of ovisac-bearing individuals with both ovaries and large testes lobes.However, in contrast to the results of both Sassaman et al.(1997) and the current study, Garcıa-Velazco et al. (2009)found putative males (individuals with neither ovisacs norovarian tissue) in 6 of their 17 sampled ‘‘short’’populations. While it is possible that our male-less T.longicaudatus ‘‘short’’ populations are the same species,clarification requires genetic study. The populations oflong-tailed Triops spp. examined by Garcıa-Velazco et al.(2009) could be either T. longicaudatus ‘‘long’’ or T.newberryi, but the authors did not mention whether thepopulations differed in the number of legless rings.However, Garcıa-Velazco et al. (2009) did determine thatthree of the long-tailed populations were gonochoric, withgenerally equal sex ratios and ovisac-bearing individualslacking testicular lobes. The remaining populations werethought to be hermaphrodite-only or hermaphrodite-dom-inated, containing individuals with ovarian tissue andtesticular lobes. Based solely on reproductive system, thepopulations with sex ratios approaching one may bepresumed to be our T longicaudatus ‘‘long,’’ and theputative hermaphroditic long-tailed form our T. newberryi,but without further population-specific morphologicalinformation, this is uncertain.

The co-existence of self-fertilizing hermaphrodites andmales is called androdioecy, and is extremely rare inanimals; simultaneous hermaphroditism is thought totypically lead to a male-less population, because theproduction of males would have a high and presumablyunnecessary energetic cost (Pannell, 2002, 2008; Weeks etal., 2006). While androdioecy has been widely studied in


the branchiopod clam shrimp Eulimnadia texana Packard,1871 (Sassaman, 1989; Sassaman and Weeks, 1993; Weekset al., 2008), studies on this reproductive system in Triopsspp. are still in the early stages (Sassaman, 1991; Sassamanet al., 1997; Garcıa-Velazco et al., 2009; Murugan et al.,2009), but has been hypothesized to exist in at least fourspecies: T. cancriformis (Zaffagnini and Trentini, 1980;Engelmann et al., 1997; Zierold et al., 2007; Zierold et al.,2009), T. granarius Lucas 1864 (Akita, 1966), T. newberryi(Sassaman, 1991; Sassaman et al., 1997; Garcıa-Velazco etal., 2009), and T. longicaudatus ‘‘short’’ (Garcıa-Velazco etal., 2009).

Our results agree with those of Sassaman et al. (1997),with a gonochoric T. longicaudatus ‘‘long,’’ an allhermaphroditic or parthenogenetic T. longicaudatus‘‘short,’’ and an androdioecious T. newberryi. UnlikeGarcıa-Velazco et al. (2009), we have not yet found anyandrodioecious populations identified as T. longicaudatus‘‘short.’’

Conclusions and Future Studies

It is our opinion that the three forms detailed in this studyare probably sufficiently distinct in morphology to warrantelevation to species level. While among-form geneticdistances are not necessarily as high as those found amongspecies in other studies, the well-supported monophyly ofat least two of the forms and the consistent morphologicaldifferences, even in co-existing populations, argues thatthese forms are on independent evolutionary pathways.However, this study examined only a small portion of therange of North American Triops, and formal speciesdesignation of the New Mexican forms should await amore extensive and time-consuming systematic revision,including re-examination of previously examined speci-mens, including the type specimens. Much work needs tobe completed for this to occur; the earlier speciesdescriptions were often vague and used inconsistent orill-defined terminology, the types are housed in widespreadlocations (USNM, Washington DC; British Natural HistoryMuseum, London; Museum Historie Naturalle, Paris;Peabody Museum, New Haven, Conn), and the samplingfor at least one form (T. longicaudatus ‘‘long’’) is presentlyinsufficient for separating playa-specific differences fromlineage-specific differences. We have concerns aboutSassaman et al.’s (1997) preliminary assignation of theeight legless ring form to T. newberryi, and believe T.aequalis may be a more appropriate designation. We haveless reservation concerning Sassaman et al.’s (1997)assignation of the long-abdomened (10 legless rings),gonochoric morph to T. longicaudatus, and their suggestionthat the short-abdomene (six legless rings), unisexualmorph, if elevated to species-level, may be called T.oryzaphagus. As mentioned by Sassaman et al. (1997),Rosenberg’s (1947) descriptions of T. oryzaphagus and T.biggsi are brief and the only explicit difference was thenumber of legless rings, five and six, respectively. Bothspecies had exclusively female populations, and Sassamanet al. (1997) considered it likely that they were the samespecies, in which case T. oryzaphagus has page precedence.However, it is unclear if Sassaman et al. (1997) examined

either type specimen, so whether these are actuallydifferent species, and whether either is synonymous withour T. longicaudatus ‘‘short’’ must await an examination ofthe type specimens.

ACKNOWLEDGEMENTS

The authors thank J. Alleman, N. Harings, M. Schiavon, M. Serena and B.Leinauer for help with field work. This research was supported by the NewMexico Agricultural Experiment Station, and support to D. Cowley and R.Sallenave from the Cooperative State Research, Education and ExtensionService, US Department of Agriculture under Agreement numbers 2008-34461-19061 and 2008-40549-04328 in cooperation with New MexicoState University and the Texas A & M University System.

REFERENCES

Akita, M. 1966. Life history of Triops longicaudatus (LeConte) in Japan I.Sex ratio and mode of reproduction. Zoological Magazine 75: 178-182.

———. 1971. On the reproduction of Triops longicaudatus (LeConte).Zoological Magazine 80: 242-250.

Bernard, H. M. 1892. The Apodidæ. A Morphological Study. MacMillanand Co., New York.

———. 1895. Hermaphroditism among the Apodidae. The Annals andMagazine of Natural History 17: 296-309.

Bilton, D. T., J. R. Freeland, and B. Okamura. 2001. Dispersal infreshwater invertebrates. Annual Review of Ecology and Systematics32: 159-181.

Brendonck, L. 1996. Diapause, quiescence, hatching requirements: whatwe can learn from large freshwater branchiopods (Crustacea: Bran-chiopoda: Anostraca, Notostraca, Conchostraca). Hydrobiologia 320:85-97.

———, and L. de Meester. 2003. Egg banks in freshwater zooplankton:evolutionary and ecological archives in the sediment. Hydrobiologia491: 65-84.

———, and B. J. Riddoch. 1999. Wind-borne short-range egg dispersal inanostracans (Crustacea: Branchiopoda). Biological Journal of theLinnean Society 67: 87-95.

———, D. C. Rogers, J. Olesen, S. C. Weeks, and W. R. Hoeh. 2008.Global diversity of large branchiopods (Crustacea: Branchiopoda) infreshwater. Hydrobiologia 595: 167-176.

Brochet, A. L., M. Gauthier-Clerc, M. Guillemain, H. Fritz, A. Waterkeyn,A. Baltanas, and A. J. Green. 2010. Field evidence of dispersal ofbranchiopods, ostracods and bryozoans by teal (Anas crecca) in theCamargue (southern France). Hydrobiologia 637: 255-261.

Cook, C. E., Q. Yue, and M. Akam. 2005. Mitochondrial genomes suggestthat hexapods and crustaceans are mutually paraphyletic. Proceedings ofthe Royal Society of London, B 272: 1295-1304.

de Waard, J. R., V. Sacherova, M. E. A. Cristescu, E. A. Remigio, T. J.Crease, and P. D. N. Hebert. 2006. Probing the relationships ofbranchiopod crustaceans. Molecular Phylogenetics and Evolution 39:491-502.

Engelmann, M., T. Hahn, and G. Hoheisel. 1997. Utrastructural character-ization of the gonads of Triops cancriformis (Crustacea, Notostraca) frompopulations containing both females and males: no evidence forhermaphroditic reproduction. Zoomorphology 117: 175-180.

Figuerola, J., A. J. Green, and T. C. Michot. 2005. Invertebrate eggs canfly: evidence of waterfowl-mediated gene flow in aquatic invertebrates.American Naturalist 165: 274-280.

———, ———, and L. Santamarıa. 2003. Passive internal transport ofaquatic organisms by waterfowl in Donana, south-west Spain. GlobalEcology and Biogeography 12: 427-436.

Folmer, O., M. Black, W. Hoeh, R. Lutz, and R. Vrijenhoek. 1994. DNAprimers for amplification of mitochondrial cytochrome c oxidasesubunit I from diverse metazoan invertebrates. Molecular MarineBiology and Biotechnology 3: 394-299.

Gand, G., J. Garric, J. Schneider, H. Walter, J. Lapeyrie, C. Martin, and A.Thiery. 2008. Notostraca trackways in Permian playa environments ofthe Lodeve basin (France). Journal of Iberian Geology 34: 73-108.


Garcıa-Velazco, H., H. Obregon-Barboza, C. Rodrıguez-Jaramillo, and A.M. Maeda-Martınez. 2009. Reproduction of the tadpole shrimp Triops(Notostraca) in Mexican waters. Current Science 96: 91-97.

Goloboff, P., J. Farris and K. Nixon. 2008. TNT, a free program forphylogenetic analysis. Cladistics 24: 774-786.

Green, A. J., K. M. Jenkins, D. Bell, P. J. Morris, and R. T. Kingsford.2008. The potential of waterbirds in dispersing invertebrates and plantsin arid Australia. Freshwater Biology 53: 380-392.

Haukos, D. A., and L. M. Smith. 1994. The importance of playa wetlandsto biodiversity of the Southern High Plains. Landscape and UrbanPlanning 28: 83-98.

Korn, M., F. Marrone, J. L. Perez-Bote, M. Machado, M. Cristo, L. C. D.Fonseca, and A. K. Hundsdoerfer. 2006. Sister species within the Triopscancriformis lineage (Crustacea, Notostraca). Zoologica Scripta 35:301-322.

LeConte, J. 1846. A new species of Apus. American Journal of Science andthe Arts 2: 274-275.

Lefebure, T., C. J. Douady, M. Gouy, and J. Gibert. 2006. Relationshipbetween morphological taxonomy and molecular divergence withinCrustacea: proposal of a molecular threshold to help species delimina-tion. Molecular Phylogenetics and Evolution 40: 435-447.

Linder, F. 1952. Contributions to the morphology and taxonomy of theBranchiopoda Notostraca, with special reference to the North Americanspecies. Proceedings of the United States National Museum 103: 1-69.

Longhurst, A. R. 1954. Reproduction in Notostraca (Crustacea). Nature173: 781-783.

———. 1955a. Evolution in the Notostraca. Evolution 9: 84-86.

———. 1955b. The reproduction and cytology of the Notostraca(Crustacea, Phyllopoda). Proceedings of the Zoological Society ofLondon 125: 671-680.

———. 1955c. A review of the Notostraca. Bulletin of the BritishMuseum (Natural History). Zoology 3: 3-57.

Maddison, W. P., and D. R. Maddison. 2009. Mesquite: a modular systemfor evolutionary analysis. Version 2.73 http://mesquiteproject.org.

Maeda-Martınez, A. M., H. Obregon-Barboza, H. Garcıa-Velazco, and G.Murugan. 2000a. A proposal on the phylogeny and the historicalbiogeography of the tadpole shrimp Triops. Anostracan News 8: 1-4.

———, V. Obregon-Barboza, M. G. Navarrete-Del Toro, H. Obregon-Barboza, and F. L. Garcıa-Carreno. 2000b. Trypsin-like enzymes fromtwo morphotypes of the ‘living fossil’ Triops (Crustacea: Branchiopoda:Notostraca). Comparative Biochemistry and Physiology B 126: 317-232.

Mantovani, B., M. Cesari, A. Luchetti, and F. Scanabissi. 2008.Mitochondrial and nuclear DNA variability in the living fossil Triopscancriformis (Bosc, 1801) (Crustacea, Branchiopoda, Notostraca).Heredity 100: 496-505.

Murugan, G., A. M. Maeda-Martınez, H. Obregon-Barboza, and N. Y.Hernandez-Saavedra. 2002. Molecular characterization of the tadpoleshrimp Triops (Branchiopoda: Notostraca) from the Baja CaliforniaPeninsula, Mexico: New insights on species diversity and phylogeny ofthe genus. Hydrobiologia 486: 101-113.

———, H. Obregon-Barboza, A. M. Maeda-Martınez, and B. V. Timms.2009. Co-occurrence of two tadpole shrimp, Triops cf. australiensis(Branchiopoda: Notostraca), lineages in middle Paroo, north-westernNew South Wales, with the first record of Triops hermaphrodites for theAustralian continent. Australian Journal of Zoology 57: 77-84.

Obregon-Barboza, H., A. M. Maeda-Martınez, and G. Murugan. 2001.Reproduction, molting, and growth of two Mexican uniparental forms ofthe tadpole shrimp Triops (Branchiopoda: Notostraca) under arecirculating culture system. Hydrobiologia 462: 173-184.

Packard, A. S. 1871. Preliminary notice of new North AmericanPhyllopoda. American Journal of Science and the Arts 2: 108-113.

———. 1883. A monograph of the phyllopod Crustacea of North America.12th Annual Report of the U.S. Geological and Geographical Survey ofthe Territories 2: 295-516.

Pannell, J. R. 2002. The evolution and maintenance of androdioecy.Annual Review of Ecology and Systematics 33: 397-425.

———. 2008. Consequences of inbreeding depression due to sex-linkedloci for the maintenance of males and outcrossing in branchiopodcrustaceans. Genetics Research, Cambridge 90: 73-84.

Pearcy, M, O. Hardy, and S. Aron. 2006. Thelytokous parthenogenesis andits consequences on inbreeding in an ant. Heredity 96: 377-382.

Posada, D., and K. A. Crandall. 1998. Modeltest: testing the model ofDNA substitution. Bioinformatics 14: 817-818.

Proctor, V. W. 1964. Viability of crustacean eggs recovered from ducks.Ecology 45: 656-658.

Remigio, E. A., and P. D. N. Hebert. 2000. Affinities among anostracan(Crustacea: Branchiopoda) families inferred from phylogenetic analysesof multiple gene sequences. Molecular Phylogenetics and Evolution 17:117-128.

Richter, S., J. Oleson, and W. C. Wheeler. 2007. Phylogeny ofBranchiopoda (Crustacea) based on a combined analysis of morpho-logical data and six molecular loci. Cladistics 23: 301-336.

Ronquist, F., and J. P. Huelsenbeck. 2003. MRBAYES 3: Bayesianphylogenetic inference under mixed models. Bioinformatics 19: 1572-1574.

Rosenberg, L. E. 1947. New species of Apus from California. Transactionsof the American Microscopal Society 66: 70-73.

Ruedemann, R. 1922. On the occurrence of an Apus in the Permian ofOklahoma. Journal of Geology 30: 311-318.

Sassaman, C. 1989. Inbreeding and sex ratio variation in female-biasedpopulations of a clam shrimp, Eulimnadia texana. Bulletin of MarineScience 45: 425-432.

———. 1991. Sex ratio variation in female-biased populations ofNotostracans. Hydrobiologia 212: 169-179.

———, and S. C. Weeks. 1993. The genetic mechanism of sexdetermination in the conchostracan shrimp Eulimnadia texana. TheAmerican Naturalist 141: 314-328.

———, M. A. Simovich, and M. Fugate. 1997. Reproductive isolation andgenetic differentiation in North American species of Triops (Crustacea:Branchiopoda: Notostraca). Hydrobiologia 359: 125-147.

Scott, S. R., and A. A. Grigarick. 1978. Observations on the biology andrearing of the tadpole shrimp Triops longicaudatus (Leconte) (Notos-traca: Triopsidae). The Wasmann Journal of Biology 36: 116-126.

———, and ———. 1979. Laboratory studies of factors affecting egghatch of Triops longicaudatus (Leconte) (Notostraca: Triopsidae).Hydrobiologia 63: 145-152.

Su, T., and M. S. Mulla. 2002. Factors affecting egg hatch of the tadpoleshrimp Triops newberryi, a potential biological control agent ofimmature mosquitoes. Biological Control 23: 18-26.

Takahashi, F. 1977. Pioneer life of the tadpole shrimps, Triops spp.(Notostraca: Triopsidae). Applied Entomological Zoology 12: 104-117.

Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: Molecularevolutionary genetics analysis (MEGA) software version 4.0. MolecularBiology and Evolution 24: 1596-1599.

Umetsu, K., N. Iwabuchi, I. Yuasa, N. Saitou, P. F. Clark, G. Boxshall, M.Osawa, and K. Igarashi. 2002. Complete mitochondrial DNA sequenceof a tadpole shrimp (Triops cancriformis) and analysis of museumsamples. Electrophoresis 23: 4080-4084.

Vanschoenwinkel, B., S. Gielen, M. Seaman, and L. Brendonck. 2008.Any way the wind blows - frequent wind dispersal drives species sortingin ephemeral aquatic communities. Oikos 117: 125-134.

Waterkeyn, A., O. Pineau, P. Grillas, and L. Brendonck. 2010a.Invertebrate dispersal by aquatic mammals: a case study with nutriaMyocastor coypus (Rodentia, Mammalia) in Southern France. Hydro-biologia 654: 267-271.

———, B. Vanschoenwinkel, S. Elsen, M. Anton-Pardo, P. Grillas and L.Brendonck. 2010b. Unintentional dispersal of aquatic invertebrates viafootwear and motor vehicles in a Mediterranean wetland area. AquaticConservation: Marine and Freshwater Ecosystems 20: 580-587.

Weeks, S. C. 1990. Life-history variation under varying degrees ofintraspecific competition in the tadpole shrimp Triops longicaudatus(Leconte). Journal of Crustacean Biology 10: 498-503.

———, and C. Sassaman. 1990. Competition in phenotypically variableand uniform populations of the tadpole shrimp Triops longicaudatus(Notostraca: Triopsidae). Oecologia 82: 552-559.

———, C. Benvenuto, and S. K. Reed. 2006. When males andhermaphrodites coexist: a review of androdioecy in animals. Integrativeand Comparative Biology 46: 449-464.

———, T. F. Sanderson, M. Zofkova, and B. Knott. 2008. Breedingsystems in the clam shrimp family Limnadiidae (Branchiopoda,Spinicaudata). Invertebrate Biology 127: 336-349.

Yee, S. H., M. R. Willig, and D. L. Moorhead. 2005. Tadpole shrimpstructure macroinvertebrate communities in playa lake microcosms.Hydrobiologia 541: 139-148.

Yoon, S. M., W. Kim, and H. S. Kim. 1992. Redescription of Triopslongicaudatus (LeConte, 1846)(Notostraca, Triopidae) from Korea. TheKorean Journal of Systematic Zoology Special Issue No. 3: 59-66.


Zaffagnini, F., and M. Trentini. 1980. The distribution and reproduction ofTriops cancriformis (Bosc) in Europe (Crustacea Notostraca). MonitoreZoologico Italiano 14: 1-8.

Zierold, T., B. Hanfling, and A. Gomez. 2007. Recent evolution ofalternative reproductive modes in the ‘living fossil’ Triops cancriformis.BMC Evolutionary Biology 7: 161.

———, J. Montero-Pau, B. Hanfling, and A. Gomez. 2009. Sex ratio,reproductive mode and genetic diversity in Triops cancriformis.Freshwater Biology 54: 1392-1405.

Zwickl, D. J., 2006. Genetic algorithm approaches for the phylogeneticanalysis of large biological sequence datasets under the maximumlikelihood criterion. Ph.D. dissertation, The University of Texas at Austin.

RECEIVED: 27 September 2010.ACCEPTED: 17 January 2011.


Morphologic and Genetic Variation in Triops (Branchiopoda: Notostraca) from Ephemeral Waters of the...

Documents

Transcript of Morphologic and Genetic Variation in Triops (Branchiopoda: Notostraca) from Ephemeral Waters of the...