University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Faculty Publications from the Harold W. Manter Laboratory of Parasitology Parasitology, Harold W. Manter Laboratory of

12-1994

Phylogenetic Systematic Analysis of the Trichostrongylidae (Nematoda), with an Initial Assessment of Coevolution and Biogeography

Eric P. Hoberg United States Department of Agriculture, Agricultural Research Service, [email protected]

J. Ralph Lichtenfels United States Department of Agriculture, Agricultural Research Service

Follow this and additional works at: https://digitalcommons.unl.edu/parasitologyfacpubs

Part of the Biodiversity Commons, Evolution Commons, and the Parasitology Commons

Hoberg, Eric P. and Lichtenfels, J. Ralph, "Phylogenetic Systematic Analysis of the Trichostrongylidae (Nematoda), with an Initial Assessment of Coevolution and Biogeography" (1994). Faculty Publications from the Harold W. Manter Laboratory of Parasitology. 723. https://digitalcommons.unl.edu/parasitologyfacpubs/723

This Article is brought to you for free and open access by the Parasitology, Harold W. Manter Laboratory of at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Faculty Publications from the Harold W. Manter Laboratory of Parasitology by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. J. Parasitol., 80(6), 1994, p. 976-996 ? AmericanSociety of Parasitologists1994

PHYLOGENETICSYSTEMATIC ANALYSIS OF THE TRICHOSTRONGYLIDAE(NEMATODA), WITH AN INITIALASSESSMENT OF COEVOLUTIONAND BIOGEOGRAPHY

E. P. Hoberg and J. R. Lichtenfels UnitedStates Departmentof Agriculture,Agricultural Research Service, Biosystematic Parasitology Laboratory, BARCEast, Building1180, 10300Baltimore Avenue, Beltsville, Maryland 20705-2350

ABSTRACT:Phylogenetic analysis of the subfamiliesof the Trichostrongylidaebased on 22 morphologicaltrans- formationseries produceda single cladogramwith a consistencyindex (CI) = 74.2%.Monophyly for the family was supportedby the structureof the female tail and copulatorybursa. Two major clades are recognized:the Cooperiinaeclade with the basal Cooperiinaeand Libyostrongylinae+ Trichostrongylinae,and the Graphidiinae clade with the basal Graphidiinaeand Ostertagiinae+ Haemonchinae.Dendrograms presented by Durette- Desset (1985) (CI = 56.1%)and Lichtenfels(1987), based on the key to the Trichostrongylidaeby Gibbons and Khalil (1982) (CI = 59.0%),were found to be relatively inefficientin describingcharacter evolution and in supportingputative relationshipsamong the subfamilies.Based on the currentanalysis, the intestine appearsto have constitutedthe ancestralhabitat for the trichostrongylidswith the stomach/abomasumhaving been in- dependentlycolonized in each clade. Assessment of host associationssuggests extensive colonization but also a high degreeof coevolution with Bovidae and Cervidaefor Ostertagiinae+ Haemonchinae.Biogeography for this assemblageis complex, but this analysisis compatiblewith a Palearcticor Eurasianorigin for Cooperiinae, Haemonchinae,and Ostertagiinae.

Trichostrongyle nematodes are significant par- Cram, 1927) and more recently by the configu- asites among sylvatic and domesticated rumi- ration of the synlophe and characteristic genital nants. An understanding of phylogenetic rela- structures (Durette-Desset and Chabaud, 1977, tionships of these nematodes has a bearing on 1981; Durette-Desset, 1983; Gibbons and Kha- elucidating aspects of host associations, parasite lil, 1982, 1983). behavior (including pathogenesis), epidemiolo- The family has received extensive attention gy, evolution, and biogeographic history. These since it was established, e.g., Yorke and Maple- nematodes may also prove to be exceptionally stone (1926), Baylis and Daubney (1926), and useful in defining the distributional history for has been reviewed in detail by Travassos (1937), ruminants in the Holarctic, e.g., Dr6zdz (1967) Skrjabin et al. (1952, 1954), Durette-Desset and and Hoberg et al. (1993a). Chabaud (1977, 1981), Gibbons and Khalil The superfamily Trichostrongyloidea Cram, (1982) and Durette-Desset (1983, 1985). The first 1927 constitutes one of the most diverse and synoptic treatment of the family was presented complex taxa within the Strongylida or bursate by Travassos (1937) whose classification of the nematodes (Durette-Desset, 1983, 1985). Al- Trichostrongylidae contained 13 subfamilies. though not adopted in the present study, recently This represented a rather broad concept for the it has been proposed that this taxon be elevated trichostrongylids and included many taxa later to subordinal rank as the Trichostrongylina (see referred to other families and subfamilies of the Durette-Desset and Chabaud, 1993). Among the Trichostrongyloidea (see Durette-Desset, 1983). trichostrongyloids, the family Trichostrongyli- More restrictive concepts for the trichostrongyl- dae Leiper, 1912 represents groups of consider- oids were developed by Skrjabin et al. (1952, able veterinary importance as parasites of ru- 1954) who recognized 15 subfamilies and 18 minants, other mammals, and avian hosts. These tribes within the Trichostrongylidae. The cur- nematodes are typically monoxenous and occur rently accepted taxonomy for the family followed in the intestine or stomach of their terrestrial from the first exhaustive studies of the genital vertebrate hosts. Morphologically these taxa were cone and copulatory bursa (Andreeva, 1958; originally defined by a reduced buccal capsule Chabaud, 1959; Chabaud et al., 1970; Gibbons and distinctive copulatory bursa (Leiper, 1912; and Khalil, 1983) and synlophe (Durette-Desset and Chabaud, 1977). Trichostrongylidae accord- ing to Durette-Desset and Chabaud (1977, 1981), Received 14 March 1994; revised 25 May 1994; ac- Gibbons and Khalil (1982), and Durette-Desset cepted 25 May 1994. (1983) is now considered to contain 6 subfami- 976 HOBERGAND LICHTENFELS-TRICHOSTRONGYLIDAEPHYLOGENY 977 lies including 4 economically significant groups: tenfels (1987) from the keys presented by Gib- Trichostrongylinae Leiper, 1908, Ostertagiinae bons and Khalil (1982) considered a broader ar- Lopez-Neyra, 1947, Haemonchinae Skrjabin and ray of characters and 2 rather than 3 major Shul'ts, 1952, Cooperiinae Skrjabin and Shik- lineages were recognized. An evaluation of these hobalova, 1952, and the relatively minor Libyo- hypotheses revealed concepts of relationships strongylinae Durette-Desset and Chabaud, 1977, among taxa that were developed from assess- and Graphidiinae Travassos, 1937. ments of single characters, overall morphological Although Gibbons and Khalil (1982) and Dur- similarity, or unique combinations of shared ette-Desset (1983) both recognized 6 subfami- primitive characters. Although each of the lies, there was discordance between these clas- subfamilies was recognized, the reliability of con- sifications with respect to the numbers of genera clusions about phylogeny of the trichostrongylids recognized, particularly within the Ostertagiinae and relationships among the subfamilies were (see Lichtenfels and Hoberg, 1993) and to their equivocal (Jansen, 1989; Hoberg and Lichten- placement at the level of subfamily. The appar- fels, 1992; Hoberg et al., 1993b). ent disparity resulted largely from conclusions In the present study, phylogenetic analyses derived from the evaluation of 2 different sets of (Hennig, 1966; Wiley, 1981) of the Trichostron- morphological characters that were applied to gylidae were initiated to: (1) test monophyly of the construction of taxonomic keys at the generic the family; (2) assess relationships among the level. Concepts of Durette-Desset (1983) were subfamilies; (3) define the relationship of the Os- based primarily on the structure of the copula- tertagiinae and Graphidiinae (see Hoberg et al., tory bursa, whereas those of Gibbons and Khalil 1993b); (4) consider the implications of subfam- (1982) were based on a broader array of struc- ily phylogeny for trends in character evolution; tural attributes (Lichtenfels, 1987). and (5) begin preliminary assessments of host, Although all previous treatments of the tricho- habitat, and biogeographic associations of par- strongylids implied degrees of morphological asites. Results of this study provide the first phy- similarity for inclusive groups of species, genera, logenetic systematic analysis of the Trichostron- tribes, and subfamilies, specific evolutionary hy- gylidae, as an extension of the seminal research potheses for the family were seldom considered. of Chabaud (1959), Durette-Desset and Chabaud The first explicit evolutionary hypothesis for the (1977, 1981), Gibbons and Khalil (1982), and 14 families of the Trichostrongyloidea, including Durette-Desset (1983, 1985, 1992). In as much the Trichostrongylidae, was presented as a den- as systematics forms the foundation for predic- drogram derived from the pioneering studies of tive hypotheses, assessment of the phylogenetic Chabaud (1959), Durette-Desset and Chabaud relationships of the trichostrongylids at the (1977, 1981), and Durette-Desset (1983, 1985). subfamilial level can form the infrastructure for Although Gibbons and Khalil (1982) had not a more refined understanding of nematode be- intended to present a classification, their con- havior, parasite-host coevolution, and biogeog- cepts were summarized by Lichtenfels (1987) in raphy. an attempt to define relationships among the subfamilies. MATERIALSAND METHODS The competing hypotheses outlined by Lich- Phylogeneticanalysis (Hennig, 1966; Wiley, 1981; tenfels (1987) represented 2 rather disparate views Wiley et al., 1991) of the subfamilies of the Tricho- of evolution and postulated relationships among strongylidaewas conductedbased on an evaluationof charactersof adult nematodes. the 6 subfamilies. Durette-Desset morphological Repre- (1981, 1983, sentativesof all currentlyrecognized subfamilies of the 1985) and Durette-Desset and Chabaud (1977, Trichostrongylidaewere examined. 1981) indicated that the Trichostrongylidae were Characterswere polarized by taxonomic outgroup divided into 3 lineages, depending on the mor- comparison (Lundberg,1972; Wiley, 1981; Wiley et with referenceto the phology of bursal rays 2 and 3. Each of these al., 1991), strongylatesuperfam- ilies Strongyloideaand Ancylostomatoideaand to oth- lineages was further broken into subfamilies, 1 er familiesof the Trichostrongyloidea.As the synlophe more primitive than the other. Other characters is an attributelimited to the Trichostrongyloidea,these were then forced to conform to relationships de- characterswere polarizedspecifically with referenceto fined by the structure of the bursa and thus ex- families Molineidae, Heligmosomidae,and Heligmo- nellidae.Selection of charactersand for tensive trends in evolution were argumentation parallel pre- characterstate polarity were found to be primarilycon- dicted by necessity (Durette-Desset, 1985, 1992). sistent with previousdecisions developed by Chabaud In contrast, the relationships derived by Lich- et al. (1970), Durette-Dessetand Chabaud(1977, 1981) 978 THEJOURNAL OF PARASITOLOGY, VOL. 80, NO.6, DECEMBER1994

TABLEI. Charactermatrix for the subfamiliesof the Trichostrongylidae.

Character 1 1 1 1 1 1 1 1 1 1 2 2 2 Subfamily 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2

Strongylida* 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Trichostrongylinae 2 0 1 1 0 0 1 1 0 1 0 1 0 0 0 0 0 0 0 1 0 0 Libyostrongylinae 1 0 1 0 1 0 0 1 0 1 0 1 0 0 0 1 0 0 1 0 0 0 Haemonchinae 0 1 0 0 1 1 0 0 1 0 0 0 0 0 0 0 1 1 0 1 1 1 Cooperiinae 1 0 0 1 0 0 0 0 0 0 1 1 0 0 0 1 0 0 1 0 0 1 Ostertagiinae 1 1 0 0 1 1 0 0 1 0 0 0 1 1 1 0 1 1 0 0 1 1 Graphidiinae 0 1 0 0 0 1 0 0 1 0 1 0 0 0 0 0 0 1 0 0 0 0

* Outgroups including the superfamilies Strongyloidea and Ancylostomatoidea and the families Molineidae, Heligmonellidae, and Heligmosomidae of the superfamily Trichostrongyloidea. and Durette-Desset (1983, 1985), particularly for the with multistatecharacters ordered. The consistencyin- synlophe and the bursa. Further refinement of defini- dex (CI) was calculatedfor individual charactersand tions of characters was based on the literature with for the overall analysis.Calculation of the homoplasy specific reference to Travassos (1937), Skrjabin et al. slope ratio (HSR) and CIndom and followed (1952, 1954), Sarwar (1956), Andreeva (1958), Lich- Meier et al. (1991) and Klassen et CIadjus•dal. (1991). Com- tenfels (1977), Durette-Desset (1982a) for Cooperiinae, parisons to previous hypotheses(specifically those of Durette-Desset (1982b) for Ostertagiinae, Gibbons and Durette-Dessetand Chabaud [1981], Durette-Desset Khalil (1982, 1983), Gibbons (1981), Hoberg and Lich- [1985],and Lichtenfels[1987] were based on the USER tenfels (1992), and Hoberg et al. (1993b, 1993c). DEFINEDTOPOLOGY function of PAUP. The den- Some characters were found to be variable within drograms for the Trichostrongylidaedeveloped by in-group taxa. In such cases where both the plesio- Durette-Dessetand Chabaud (1981) and Lichtenfels morphic and apomorphic condition were present, giv- (1987), the latterbased on Gibbons and Khalil (1982), en characters were coded as apomorphic for the were redrawn and employed as USER DEFINED subfamily to recognize the acquisition of the derived TREES to compare these alternative hypotheses di- state in at least some of the in-group genera. This de- rectly. Charactersused in the current analysis were cision influenced coding for 7 characters as follows mapped onto these trees and optimized by FARRIS (with derived state listed first):buccal tooth (character OPTIMIZATION.The overall CI and those for in- 5-Libyostrongylinae with presenceor absenceof this dividual characterswere calculated,and the HSR and attribute);form of the bursa (character12-Cooperi- CIadjus,,,tedwere determined as a basis for evaluating the inae with both 1-3-1 and 2-3; character13--Osterta- efficiencyand strengthof the competinghypotheses for giinae with both 2-2-1 and 2-1-2); dorsal lobe (char- trichostrongylidphylogeny. acter 17--Ostertagiinaewith reducedand long dorsal Host habitat (localizationin the host), host associ- lobe); spicule length (character18--Ostertagiinae and ations, and geographicdistributions ofsubfamilial taxa the Haemonchinaewith longand shortspicules); dorsal wereexamined by mappingthese dataonto the parasite ray symmetry(character 21--Ostertagiinae and Hae- phylogenyproduced in the currentstudy. monchinae with asymmetric and symmetric dorsal lobes); vulval flap (character22-Cooperiinae, Oster- RESULTS tagiinae, and Haemonchinae,presence or absence of flaps). Potential influence on the topology of clado- Characters gram(s)due to these decisions for coding of variable characterswas examined with an alternatematrix in 1) Body length of male. Three states: 0 = rel- which characterstates as presentedabove were coded atively long (consistently >10 mm); 1 = medium as plesiomorphic. in length (generally 6-10 mm); 2 = minuscule Five multistate characterswere split into indepen- (typically <5 mm). dent transformationseries and recodedas binarychar- 2, 3) Cephalic vesicle (presence and structure) acters(see Glen and Brooks, 1985) to accountfor der- ivation of some characterstates as follows: cephalic (Fig. 1). The structure of the cephalic vesicle is vesicle (characters2, 3), synlophe,number of ridges(9, variable within the Trichostrongylidae and was 10), bursa form (12, 13), "7" papillae (15, 16), and split into 2 independent transformation series bursalrays 2 and 3 (19, 20). (see Glen and Brooks, 1985). It is represented by A summaryof the 22 homologousseries, represent- a and cervical ex- ing 23 characterstates, is presented below and in a well-developed large cephalic numerical charactermatrix (Table I). Plesiomorphic pansion demarcated from the body in the Coo- statesare coded as 0, apomorphicas 1 or 2. Characters periinae (and most outgroups) (0, 0) (Fig. la). A have been illustratedto depict the plesiomorphicand reduced, poorly developed cephalic and cervical apomorphic for each attribute(Figs. state 1-14). expansion is characteristic of the Ostertagiinae, Phylogenetic analysis was conducted with PAUP version 2.4 (Swofford, 1985). The following options Graphidiinae, and Haemonchinae (1, 0) (Fig. ib), were used: ALLTREES,FARRIS OPTIMIZATION, whereas cephalic modification is absent in the HOBERGAND LICHTENFELS-TRICHOSTRONGYLIDAEPHYLOGENY 979

Trichostrongylinae and Libyostrongylinae (0, 1) b); 1 = paired papillae (Fig. 9c) (see Hoberg et (Fig. Ic). Character 2. Two states: 0 = well de- al. [1993b] regarding detailed argumentation for veloped; 1 = reduced. Character 3. Two states: this character). O = well developed; I = absent. 15, 16) Dorsal raylets or "7" papillae and the 4) Buccal cavity (Fig. 2). Two states: 0 = pres- accessory bursal membrane (Fig. 9). Split into ent; 1 = present, poorly developed. independent transformation series, the "7" pa- 5) Buccal tooth (Fig. 3). Two states: 0 = absent; pillae may be papilliform with a reduced mem- 1 = present (considered to include the "neodont" brane (0, 0) (Fig. 9a), elongate and well-devel- of Durette-Desset, 1985). oped supporting a prominent accessory bursal 6) Cervical papillae (form) (Fig. 4). Two states: membrane (1, 0) (Fig. 9c), or with a highly mod- 0 = sensilla-like, often situated in a pit; 1 = large, ified membrane (0, 1) (Fig. 9d). Character 15. thornlike, often triangular, projecting from body Two states: 0 = papilliform with a reduced or surface. modified membrane; 1 = elongate with a highly 7) Excretory pore (Fig. 5). Two states: 0 = developed membrane. Character 16. Two states: aperture not contained in a depression of the 0 = papilliform and reduced or elongate; 1 = body surface; 1 = aperture situated in permanent modified. v-shaped depression. 17) Dorsal lobe of bursa and dorsal ray (Fig. 8) Synlophe (presence and development) (Fig. 10). Two states: 0 = long dorsal lobe and ray; 1 6). Two states: 0 = present and extending into = reduced or short dorsal lobe and ray. posterior region of the body; 1 = reduced, limited 18) Spicules (relative length). Two states: 0 = to specific regions of the cuticle (usually near the short; 1 = long. vulva) or absent. 19, 20) Bursal rays 2 and 3 (relative length and 9, 10) Synlophe (number of ridges) (Fig. 7). position) (Fig. 12). Split into independent trans- Split into independent series, the synlophe may formation series, the tips of rays 2 and 3 may be be composed of relatively few ridges as in the equal in length and parallel (0, 0) (Fig. 12a), ray Cooperiinae (0, 0) (Fig. 7a), may be composed 2 may be less in length than ray 3 with the tips of a high number of ridges as in the Haemon- being convergent (1, 0) (Fig. 12c), or ray 2 may chinae, Ostertagiinae, and Graphidiinae (0, 1) be shorter than ray 3 with the tips being divergent (Fig. 7b), or may be reduced or absent as in the (0, 1) (Fig. 12b). Character 19. Two states: 0 = Libyostrongylinae and Trichostrongylinae (1, 0) equal and parallel; 1 = convergent. Character 20. (Fig. 7c). Character 9. Two states: 0 = low num- Two states: 0 = equal; 1 = divergent. ber; 1 = high number. Character 10. Two states: 21) Dorsal ray (symmetry) (Fig. 10). O = sym- 0 = low number; 1 = absent or reduced. metric dorsal ray; 1 = asymmetric dorsal ray. 11) Synlophe (ridge height). Two states: 0 = 22) Vulval flap (presence) (Fig. 11). O = absent; ridges of near equal height; 1 = lateral ridges or 1 = present. lateralmost ridge substantially smaller than those in the dorsal and ventral fields. Phylogenetic analysis 12, 13) Bursa (form and type) (Fig. 8). Among A single phylogenetic tree for the relationships outgroups, the bursa may be 2-1-2 or 2-3 (e.g., of the 6 subfamilies of the Trichostrongylidae Strongyloidea and Molineidae); derived states are resulted from the analysis of 22 morphological represented by 1-3-1 (Trichostrongylinae, Libyo- attributes (represented by 23 character states) (Fig. strongylinae, and Cooperiinae with a limited 15). The hypothesis was strongly supported with number of genera of the latter with 2-3) or 2-2-1 a CI = 74.2% (minimum length = 23 steps; pos- (Ostertagiinae and only among a limited number tulated steps = 31), HSR = 51.3%, and Cladjusted = ofheligmosomes). This multistate character was 19.2% (Clactua, - Clrandom = Cladjusted), indicat- split to recognize the potential of independent ing substantial phylogenetic information was origin of derived states for the form of the bursa: contained in the cladogram. Consistency indices the bursa may be 2-3 or 2-1-2 (0, 0) (Fig. 8a, b), for individual characters are present in Table II. 1-3-1 (1, 0) (Fig. 8c), or 2-2-1 (0, 1) (Fig. 8d). Homoplasy was postulated for parallel devel- Character 12. Two states: 0 = 2-3 or 2-1-2; 1 = opment in 5 characters (body length, buccal tooth, 1-3-1. Character 13. Two states: 0 = 2-3 or 2-1- synlophe height, bursal rays 2 and 3, and vulval 2; 1 = 2-2-1. flap) and reversal in 3 characters (buccal cavity, 14) Ventral raylet or "0" papilla (structure) "7" papillae and bursal rays 2 and 3). Postulated (Fig. 9). Two states: 0 = single papilla (Fig. 9a, instances of homoplasy were evenly distributed 980 THEJOURNAL OF PARASITOLOGY, VOL. 80, NO.6, DECEMBER1994

b 4 b a _ 2a 38

!A L i

4b 'b:flb b

tA "

a b \

FIGURES1-7. Charactersused in phylogeneticanalysis of the subfamiliesof the Trichostrongylidae(not to scale). In each figure,the plesiomorphiccondition is indicated by "a" and the apomorphiccondition by "b," "c," etc. 1. Cephalicvesicle, denoted by regionbetween pointers(characters 2/3): la. Largevesicle in Cooperia curticei(Giles, 1892) (from Gibbons and Khalil, 1982). lb. Reduced vesicle in Longistrongylusnamaquensis (Ortlepp, 1963) (from Gibbons, 1977). ic. Absence of vesicle in Trichostrongyluscolubriformis (Giles, 1892) (fromGibbons and Khalil, 1982). 2. Buccalcavity, indicatedby pointer(character 4): 2a. Presentand developed HOBERGAND LICHTENFELS-TRICHOSTRONGYLIDAEPHYLOGENY 981 through the tree and largely restricted to terminal within the clade but undergo postulated reversal taxa and branches. in the Libyostrongylinae (4) and the Trichostron- Monophyly for the Trichostrongylidae is sup- gylinae (16, 19). The sister-group association ported by 2 synapomorphies (defined by refer- within the Cooperiinae clade of the Libyostron- ence to taxonomic outgroups): the length of ray gylinae + Trichostrongylinae is supported by 3 4 of the copulatory bursa (Fig. 13) and the ab- synapomorphies: absence of a cephalic vesicle sence (putative secondary loss) of the tail spine (3) and reduction or absence of the synlophe (8, in the female (Fig. 14), both largely consistent 10). characters that were excluded from the analysis. In contrast, the Graphidiinae clade is defined With respect to a long ray 4, this attribute is by a reduced cephalic vesicle (character 2), thorn- present among all subfamilies, although in a lim- like cervical papillae (6), a high number of ridges ited number of the Ostertagiinae this ray is short comprising the synlophe (9), and relatively long (e.g., Spiculopteragia [Orloff, 1933], Mazama- spicules (18). The sister-group relationship with- strongylus Cameron, 1935, Rinadia Grigorian, in the Graphidiinae clade of the Ostertagiinae + 1951, and Apteragia Jansen, 1958). Thus, with Haemonchinae is supported by 2 synapomor- the presence of both apomorphic and plesio- phies: reduced dorsal lobe (17), asymmetric dor- morphic states for this character in the Oster- sal ray (21), and secondarily by the presence of tagiinae (and coding as apomorphic for the ge- a buccal tooth (5) and vulval flap (22). nus), the character may be interpreted as A minimal level of parallelism in 5 characters undergoing a reversal among a limited number is evident and distributed across both clades (Fig. of genera in this taxon. This is compatible with 15). The medium length of the body (1) is found a secondary reduction in the length of the 4th in the Cooperiinae and Libyostrongylinae of the ray, whereas all other genera of the Ostertagiinae Cooperiinae clade and the Ostertagiinae of the (see Gibbons and Khalil 1982; Durette-Desset Graphidiinae clade. A buccal tooth (5) is present 1983, 1989; Jansen, 1989) have the synapomor- in some Libyostrongylinae and the Haemon- phic condition that defines the family Tricho- chinae + Ostertagiinae. A reduction in the height strongylidae. of the lateralmost ridge (11) is evident in the Two major clades sharing a sister-group rela- Cooperiinae and the Graphidiinae. Bursal rays tionship were recognized within the Trichostron- 2 and 3 are divergent (20) in the Trichostrongyl- gylidae (Fig. 15). The Cooperiinae clade (the bas- inae and the Haemonchinae. Vulval flaps (22) al Cooperiinae and the Libyostrongylinae + are present in the Ostertagiinae + Haemonchin- Trichostrongylinae) and the Graphidiinae clade ae and the Cooperiinae. (the basal Graphidiinae and the Ostertagiinae + The influence of differences in coding for vari- Haemonchinae) were each defined by 4 synapo- able characters was examined. An alternate ma- morphies. Considering the Cooperiinae clade, a trix with these specific characters coded as ple- 1-3-1 bursa (character 12) unequivocally defines siomorphic (5 for Libyostrongylinae; 12, 22, for this inclusive group. Additionally, the absence Cooperiinae; 13, 17, for Ostertagiinae; and 18, ofa buccal cavity (4), a modified accessory bursal 21, 22, for Haemonchinae and the Ostertagiinae) membrane and "7" papillae (16), and conver- was analyzed. A single tree resulted with a CI = gence of ray 2 and 3 (19) are generally constant 76.9%; there were no changes in the overall to-

in Ostertagiaostertagi (Stiles, 1892) (from Andreeva, 1958). 2b. Poorly developed in Trichostrongylusretortae- formis (Zeder, 1800) (from Yorke and Maplestone, 1926). 3. Buccal tooth indicated by pointer (character5): 3a. Absence of tooth in Cooperiafuelleborni Hung, 1926 (from Gibbons, 1981). 3b. Presence of tooth in Haemonchuscontortus (Rudolphi, 1802) (from Gibbons and Khalil, 1982). 4. Cervicalpapillae in dorsoventral view, showing relative position in cervical region and structure(arrows) (character 6): 4a. Sensilla-likepapillae in Cooperiasp. (from Gibbons, 1981). 4b. Thornlikepapillae in Haemonchuscontortus (from Gibbons, 1979). 5. Excretorypore in lateral view (pointers)(character 7): 5a. Unmodified pore in Cooperiafuelleborni (from Gibbons, 1981). 5b. Pore in v-shaped notch in T. colubriformis(from Gibbons and Khalil, 1982). 6. Synlophe (character8): 6a. Cervicalregion, showing ventral view and ridges (pointers)in Cooperianeitzi Monnig, 1932 (from Hoberg et al., 1993c). 6b. Absence of synlophe in T. colubriformis(from Gibbons and Khalil, 1982). 7. Synlophe in transversesection near midbody (characters9/10): 7a. Low numbersof ridgesin C. curticei(from Gibbons, 1981). 7b. High number of ridges in O. ostertagi(from Gibbons and Khalil, 1982). 7c. Reduction of synlophe in Libyostrongylusdouglassi (Cobbold, 1882) (from Gibbons and Khalil, 1982). 982 THEJOURNAL OF PARASITOLOGY,VOL. 80, NO. 6, DECEMBER1994

8 c d a b .•

9 14 kk a Cob

41

b a a C

a aJ ,IK ~J1/ 74n

8I 12

14 13 13 "

FIGURES 8-14. Characters used in phylogenetic analysis of the subfamilies of the Trichostrongylidae (not to scale). In each figure, the plesiomorphic condition is indicated by "a," and the apomorphic condition by "b," "c," etc. 8. Bursa form with disposition of rays indicated by pointers (characters 12/13): 8a. A 2-3 bursa in Cooperioides kenyensis Daubney, 1933 (from Durette-Desset, 1983). 8b. A 2-1-2 bursa in Ostertagia spp. (from Andreeva, 1958). 8c. A 1-3-1 bursa in T. retortaeformis (from Skrjabin et al., 1954). 8d. A 2-2-1 bursa in Teladorsagia spp. (from Andreeva, 1958). 9. Structure of "O" (arrows) and "7" (pointers) papillae in ventral HOBERGAND LICHTENFELS-TRICHOSTRONGYLIDAEPHYLOGENY 983 pology, relationships, or support of nodes and TABLEII. Consistencyindices for individual charac- terminal branches of the tree. However, 3 char- ters used in analysis of the Trichostrongylidae. acters became constant the (13, 21, 22), reducing Char- number of phylogenetically informative char- acter num- acters in the analysis to 19. ber Character CI Homoplasy*

Alternate hypotheses for the Trichostrongylidae 1 Body length 0.667 P 2 Cephalic vesicle 1.0 - Explicit dendrograms for the relationships of 3 Cephalic vesicle 1.0 - 4 Buccal cavity 0.50 R the trichostrongylids, as presented in the litera- 5 Buccal tooth 0.50 P ture, were redrawn as cladograms to promote 6 Cervical papillae 1.0 - their evaluation with respect to the present phy- 7 Excretory pore 1.0 - 8 (presence) 1.0 logeny using the TOPOLOGY function of PAUP Synlophe - 9 Synlophe (no. of ridges) 1.0 - (Swofford, 1985). Characters from the current 10 Synlophe (no. of ridges) 1.0 - study were mapped onto these alternative trees 11 Synlophe (height) 0.50 P 12 Bursa (form) 1.0 - and optimized to allow a determination of the 13 Bursa (form) 1.0 - 14 efficiency of each of the competing hypotheses. "0" Papilla(e) 1.0 - 15 "7" Papillae 1.0 - The forced topological comparison (including the 16 "7" Papillae 0.50 R CI, HSR, and CIadjusted), based on a single char- 17 Dorsal ray (development) 1.0 - 18 acter database, emphasizes putative differences Spicules (length) 1.0 - 19 Bursa (rays 2, 3) 0.50 R in character evolution and relative support for 20 Bursa (rays 2, 3) 0.50 P relationships at the level of subfamilies within 21 Dorsal ray (symmetry) 1.0 - the Trichostrongylidae. 22 Vulval flap 0.50 P The hypothesis developed by Durette-Desset * Postulated homoplasy as indicated by: P = parallelism; R = reversal. and Chabaud (198 1) and Durette-Desset (1983, 1985) specifies sister-group relationships for 3 distinct lineages (Cooperiinae + Libyostrongyl- versal being required to describe relationships inae; Graphidiinae + Ostertagiinae; and Tricho- depicted in the tree (Table III). strongylinae + Haemonchinae) (Fig. 16). The Monophyly was assumed for the Trichostron- resulting cladogram (as a user-defined topology) gylidae, but support for sister-group relation- had a CI = 56.1% (minimum length = 23; steps ships was minimal. The Cooperiinae + Libyo- = required = 41); HSR 1.0 and CIadjusted= 1.1%, strongylinae are defined by 2 synapomorphies indicating minimal phylogenetic information was (character 16, dorsal raylets; 19, bursal rays 2 contained in the hypothesis. Homoplasy was and 3). The Graphidiinae + Ostertagiinae and postulated for 15 characters with parallel devel- Trichostrongylinae + Haemonchinae are de- opment in 10 attributes and 5 instances of re- fined by 4 characters (character 2, cephalic ves-

view (characters14, 15, 16): 9a. Single "0" papilla and reduced "7" and membranein T. colubriformis(from Gibbons and Khalil, 1982). 9b. Single "0" papilla and highly modified "7" and membranein Cooperiarotun- dispiculumGibbons and Khalil, 1980 (from Gibbons, 1981). 9c. Paired "0" papilla and elongate "7" papillae with accessorybursal membraneCamelostrongylus mentulatus (Railliet and Henry, 1909) (from Gibbons and Khalil, 1982). 10. Structureof dorsal lobe and ray (characters17 and 21): 10a. Long, symmetricdorsal ray and lobe in ventral view in Cooperiapectinata Ransom, 1907 (from Gibbons and Khalil, 1982). 10b. Short, asym- metric dorsal ray and lobe in dorsal view in H. contortus(from Gibbons and Khalil, 1982). 11. Vulval flap seen in left lateral view (character22): 1la. Absence of flap in Cooperiaconnochaetes Boomker, Horak and Alves, 1979 (fromGibbons, 1981). 1ib. Presenceof flapin Teladorsagiacircumcincta (Stadleman, 1894) (from Lancaster and Hong, 1990). 12. Bursal rays 2 and 3 showing length and disposition of tips as indicated by pointers (characters19/20): 12a. Rays equal and parallelin Ostertagiaspp. (from Andreeva, 1958). 12b. Ray 2 < 3 and divergent in Haemonchus longistipesRailliet and Henry, 1909 (from Gibbons, 1979). 12c. Ray 2 < 3 and convergentin C. kenyensis(from Durette-Desset, 1983). 13. An elongate bursal ray 4 (pointers)is a putative synapomorphyfor the Trichostrongylidae:13a. Short ray in Amphibiophiluschabaudi Skrjabin, 1916 (from Durette-Desset,1983). 13b. Long ray 4 in H. contortus(from Gibbons and Khalil, 1982). 14. A roundedtail in females (pointers)is a putative synapomorphyfor the Trichostrongylidae:14a. Tail with spine in Nematodirus archariSokolova, 1948 (from Rickardand Lichtenfels,1989). 14b. Roundedtail without spine in Parostertagia heterospiculumSchwartz and Alicata, 1933 (from Gibbons and Khalil, 1982). 984 THE JOURNALOF PARASITOLOGY,VOL. 80, NO. 6, DECEMBER1994

O Od 'a, 43 92O OL 2 ~' 11,22 081 52

17,21H

-4 ,12, 16, 19 2,6,9,1 8

CI: 74.2%

Homoplasy 4 A~

FIGURE 15. Cladogramderived from the analysis presentedin the currentstudy showing phylogenetichy- pothesis for the 6 subfamilies of the Trichostrongylidae. This hypothesis has a CI = 74.2%, CIadju, = 19.2% and HSR = 51.3%,indicating strong support for the hypothesis.Two majorclades are recognized: the Cooperiinae clade comprised of Cooperiinae,Libyostrongylinae, and Trichostrongylinaeand the Graphidiinaeclade with Graphidiinae,Ostertagiinae, and Haemonchinae.Characters have been mappedonto the tree, and apomorphic attributesare noted by arrowsdefining specific nodes and branches,whereas postulated homoplasy is indicated by asterisks(parallelism and convergence)and stars (reversals);numbering is consistent with the data matrix, and definitionsof charactersare presentedin the text.

icle; 6, cervical papillae; 9, synlophe; 18, spic- The hypothesis developed by Lichtenfels (1987) ules) that undergo reversal in the Trichostron- was based on morphological studies conducted gylinae. A synapomorphy supporting the Oster- by Gibbons and Khalil (1982), which resulted in tagiinae + Graphidiinae is lacking and the Hae- the development of a key for the family. This monchinae + Trichostrongylinae are defined by hypothesis specifies sister-group relationships for a single attribute (20, bursal rays 2 and 3). Ab- 2 distinct lineages (Trichostrongylinae + Libyo- sence of unequivocal support for the hypothesis strongylinae and Haemonchinae as the sister- is indicated by a high level of homoplasy in the group of the Graphidiinae + Ostertagiinae + terminal branches, a high HSR, and low CIadjusted Cooperiinae) (Fig. 17). The resulting cladogram (Fig. 16). (as a user-defined topology) had a CI = 59.0% Parallelism was not evenly distributed but was (minimum length = 23; steps required = 39); associated with terminal branches (taxa) in each HSR = 95.7% and Cladjusted= 4.0%, indicating of the 3 lineages. Specifically most instances of minimal phylogenetic information was con- parallel development occurred between the Os- tained in the hypothesis. Homoplasy was pos- tertagiinae and Haemonchinae (4 characters: tulated for 15 characters with parallel develop- character 5, buccal tooth; 17, dorsal lobe; 21, ment in 9 attributes and 6 instances of reversal dorsal ray; and 22, vulval flap) and the Tricho- being required to describe relationships depicted strongylinae and the Libyostrongylinae + Coo- in the tree (Table IV). periinae (5 characters: 3, cephalic vesicle; 4, buc- Monophyly was assumed for the Trichostron- cal cavity; 8, synlophe; and 10, synlophe). gylidae, but support for sister-group relation- HOBERGAND LICHTENFELS-TRICHOSTRONGYLIDAEPHYLOGENY 985

1 1 4,11,22 1,7 5, 17,2 1,2 2 1 1

$CGC1, 20

CI: 56.1% -61

Homoplasy **C

FIGURE16. Cladogramfor the Trichostrongylidaebased on the dendrogrampresented by Chabaud and Durette-Desset(1977) and Durette-Desset(1983, 1985, 1992). The tree representsa user-definedtopology and allows direct comparisonto the presentanalysis (Fig. 15). This hypothesishas a CI = 56.1%,CIaudu = 1.1%, and HSR = 1.0, indicating that minimal phylogeneticinformation is presented.The distributionof putative apomorphicattributes is indicated by arrows,and homoplasy is representedby asterisks(parallelism and con- vergence)or stars(reversal); numbering of charactersis consistentwith the data matrixand definitionspresented in the text. Terminal branches are as follows: COOP = Cooperiinae, LIBY = Libyostrongylinae,GRAP = Graphidiinae,OSTE = Ostertagiinae,HAEM = Haemonchinae,and TRIC = Trichostrongylinae.

ships was minimal. The Trichostrongylinae + TABLEIII. Consistency indices for individual char- Libyostrongylinae are defined by 3 synapomor- actersused in analysisof Trichostrongylidaewhen ap- plied to the hypothesis of Durette-Desset(1985). phies (character 3, cephalic vesicle; and 8 and

10, synlophe). The Haemonchinae clade is de- Character fined by 5 characters (character 2, cephalic ves- number CI Homoplasy* icle; 6, cervical papillae; 9, synlophe; 18, spicules; 1 0.50 R 22, vulval flap) that undergo reversal in the 2 0.50 R 3 0.50 P Cooperiinae or Graphidiinae. A synapomorphy 4 0.50 P supports the Graphidiinae + Ostertagiinae and 5 0.333 P Cooperiinae (11, synlophe). Absence of unequiv- 6 0.50 R 7 1.0 - ocal support for the hypothesis is indicated by a 8 0.50 P low Cladjusted, a high HSR, and a high level of 9 0.50 R in the terminal branches. 10 0.50 P homoplasy 11 0.50 P Homoplasy was not evenly distributed among 12 0.50 P taxa (Fig. 17). Specifically, most instances ofpar- 13 1.0 - 14 1.0 - allel development occurred between the Oster- 15 1.0 - tagiinae and Haemonchinae (3 characters: char- 16 1.0 - acter 5, buccal tooth; 17, dorsal lobe; 21, dorsal 17 0.50 P 18 0.50 R ray) and the Cooperiinae with the Trichostron- 19 1.0 - gylinae and the Libyostrongylinae (4 characters: 20 1.0 - buccal 16, "7" 19, 21 0.50 P 4, cavity; 12, bursa; papillae; 22 0.333 P bursa). Among characters with postulated rever- sal, 4 of 6 were associated with the Cooperiinae * Postulated homoplasy as indicated by: P = parallelism; R = reversal. 986 THE JOURNALOF PARASITOLOGY,VOL. 80, NO. 6, DECEMBER1994

~s, C) AQPo' 11 5,17,21 2,6,9,18 4,20 5,16,19 13,14,15 4,12, 1, 1 -< 1,7 5,17,20,21 22

-'11

1,12

43,8,10 <2,6,9,18,22

Cl: 59.0%

Homopiasy (Cej

FIGURE17. Cladogram for the Trichostrongylidae based on the dendrogram presented by Lichtenfels (1987) from studies by Gibbons and Khalil (1982). The tree represents a user-defined topology, allowing direct com- parison to competing hypotheses (Figs. 15, 16). The hypothesis has a CI = 59.0%, CIadjustd = 4.0%, and HSR = 95.7%, indicating minimal phylogenetic information is presented and that putative relationships are not well supported. The distribution of putative apomorphic characters is indicated by arrows, and homoplasy is rep- resented by asterisks (parallelism and convergence) or stars (reversal). Characters are consistent with those presented in the data matrix and are defined in the text. Terminal branches are as follows: TRIC = Trichostrongyl- inae, LIBY = Libyostrongylinae, HAEM = Haemonchinae, GRAP = Graphidiinae, OSTE = Ostertagiinae, and COOP = Cooperiinae.

TABLEIV. Consistency indices for individual char- (2, cephalic vesicle; 6, cervical papillae; 9, syn- acters used in analysis of Trichostrongylidae when ap- lophe; 18, spicules). plied to the dendrograms presented by Lichtenfels (1987) based on a key to genera of the Trichostron- gylidae by Gibbons and Khalil (1982). DISCUSSION of the 6 subfamilies of Character Phylogenetic analysis number CI Homoplasy* the Trichostrongylidae resulted in a single phy-

1 0.667 P logenetic tree, which allows recognition ofmono- 2 0.50 R phyly for the family (Fig. 15). Strong character 3 1.0 support was evident for each of the subfamilies, 4 0.50 P 5 0.333 P and 2 major clades were postulated. Minimal 6 0.50 R levels of convergence or parallelism (homoplasy) 7 1.0 were to describe the of the tree 8 1.0 - required topology 9 0.50 R (Fig. 15; Table II). 10 1.0 - Important diagnostic characters may be rec- 11 0.50 R 12 0.50 P ognized for the family and each putative clade. 13 1.0 - Synapomorphies for the family include the length 14 1.0 - of the fourth ray of the copulatory bursa and the 15 1.0 - 16 0.50 P absence of a tail spine in females (Figs. 13, 14). 17 0.50 P The Graphidiinae clade (Graphidiinae and Os- 18 0.50 R + is a 19 0.50 P tertagiinae Haemonchinae) diagnosed by 20 0.50 P high number of ridges comprising the synlophe 21 0.50 P and large protruding, thornlike cervical papillae. 22 0.50 P Within the Graphidiinae clade, the sister-group * Postulated homoplasy as indicated by: P = parallelism; R = reversal. relationship of the Haemonchinae and Osterta- HOBERGAND LICHTENFELS-TRICHOSTRONGYLIDAEPHYLOGENY 987 giinae is supported by the structure of the dorsal Trichostrongylidae developed by Durette-Desset ray and also by the presence of an esophageal and Chabaud (1977, 1981), Durette-Desset (1983, tooth and vulval flap. In contrast, the strongly 1985, 1992), and Lichtenfels (1987) (from Gib- reduced buccal cavity and bursal pattern diag- bons and Khalil, 1982) were found to have min- nose the Cooperiinae clade (Cooperiinae and imal explanatory power in elucidating the dis- Libyostrongylinae + Trichostrongylinae). Ad- tribution of character states observed among taxa ditionally, the extreme reduction or absence of (Figs. 15-17). Compared to the hypothesis pre- the synlophe further supports recognition of a sented herein, these earlier attempts at phylo- sister-group relationship for the Trichostrongyl- genetic reconstruction were less efficient in ex- inae and Libyostrongylinae. plaining character evolution (Tables II-IV) as Before comparing the competing hypotheses demonstrated by a lower overall Cladjusted and for the Trichostrongylidae, the methodological higher HSR (19.2% and 51.3%, respectively in problem of coding polymorphic characters (ple- Fig. 15; versus 1.1% and 1.0 in Fig. 16; and 4.0% siomorphic versus apomorphic when both states and 95.7% in Fig. 17). These statistics are con- of a character are present) within a supraspecific sistent with a high level of homoplasy in the taxon must be addressed. In such instances, de- hypotheses by Durette-Desset (1983, 1985) and cisions for coding could lead to misinterpretation Lichtenfels (1987), which involved parallelism of relationships. For instance, if a specific char- or reversal for 15 of 22 characters (Tables II-IV; acter is coded as apomorphic, then the possibility Figs. 15-17). Although each of the subfamilies of recognizing secondary reversal within a lim- was recognized in these earlier studies, the reli- ited number of taxa is obscured. Conversely, if ability of conclusions about phylogeny of the coded as plesiomorphic, then the application as trichostrongylids and relationships among the a synapomorphy at higher levels becomes am- subfamilies were equivocal (Hoberg and Lich- biguous. This situation occurred with respect to tenfels, 1992; Hoberg et al., 1993b). the length of the fourth ray of the copulatory Durette-Desset (1981, 1983, 1985) and Dur- bursa in the Ostertagiinae. In the first instance, ette-Desset and Chabaud (1977, 1981) proposed a long ray would constitute a synapomorphy for that the Trichostrongylidae were divided into 3 the Trichostrongylidae as presented in the cur- lineages, depending on the morphology of bursal rent study (with recognition of secondary rever- rays 2 and 3. Each of these lineages was further sal only in a limited number of genera in the broken into a "primitive" and "specialized" Ostertagiinae). In contrast, a short ray could con- subfamily based primarily on the structure of the stitute a plesiomorphy for the Ostertagiinae (in- synlophe. The hypothesis developed by Durette- terpreted as a secondary reversal for the entire Desset (1985) specified sister-group relationships subfamily), thus obscuring acquisition of the de- for 3 distinct lineages (Cooperiinae + Libyo- rived state in a number of genera. strongylinae; Graphidiinae + Ostertagiinae; An alternative interpretation for polymor- Trichostrongylinae + Haemonchinae). How- phism is that the inclusive taxon may not be ever, this topologydoes not appearto have strong monophyletic. Limited instances of polymor- support based on consideration of the morpho- phism may indicate that some currently recog- logical data base utilized in the present study. nized subfamilies contain genera that do not be- Support for the dendrogram of Durette-Desset long to the inclusive group. Thus, the necessity (1985) is initially determined and subsequently of a hierarchical approach where successively constrained by the relationship of a single char- lower taxonomic groups are examined is evident. acter (bursal rays 2 and 3). Consequently, other Consequently, if such characters are to be used characters require multiple trends in parallel and in phylogenetic reconstruction, complete and ac- progressive development (specifically characters curate definition is requisite. With respect to de- 5, 17, 21, and 22 between the Haemonchinae cisions for other variable characters in the cur- and the Ostertagiinae defined by character 20; rent analysis, coding of states as plesiomorphic and 3, 4, 8, and 10 between the Trichostrongyl- or apomorphic did not influence the topology, inae and the Libyostrongylinae + Cooperiinae recognition ofsister-group relationships, or length defined by character 19). Parallelism was not of the tree. However, it did reduce the number evenly distributed (Fig. 16), with most instances of phylogenetically informative characters from occurring between the Ostertagiinae and Hae- 22 to 19. monchinae and the Trichostrongylinae and the Alternative evolutionary hypotheses for the Libyostrongylinae + Cooperiinae. Additionally, 988 THEJOURNAL OF PARASITOLOGY,VOL. 80, NO. 6, DECEMBER1994 the branch with Ostertagiinae+ Graphidiinae trends in some characters(in this case, parallel lacked a definingsynapomorphy (Fig. 16). trendsin evolution of the synlophe and the cop- The putative relationshipof the Ostertagiinae ulatorybursa). and the Graphidiinaehas received considerable Consequently, a difference in the proposed attention (Durette-Desset, 1981, 1982b, 1985; mode of evolution may constitute the basis for Jansen, 1989; Hoberg et al., 1993b). Durette- disparities in these phylogenetic relationships Desset and Chabaud(1981) and Durette-Desset postulatedfor the Trichostrongylidae.In the hy- (1985) had suggestedderivation of the Osterta- potheses by Durette-Dessetand Chabaud(1977) giinae via multiple origins from 2 genera of the and Durette-Desset(1982b, 1985), a strongpro- Graphidiinae (see summary in Hoberg et al. gressivecomponent is evident within each of the [1993b]) (Fig. 18A). If this view is correct, by 3 lineages (Libyostrongylinae-Cooperiinae; definition the Ostertagiinaewould be polyphy- Trichostrongylinae-Haemonchinae;Graphidi- letic (derived independently from 1 or more inae-Ostertagiinae), which undergo parallel ancestorsreferred to anothertaxon), whereas the trends in the evolution of specific charactersof Graphidiinaewould become paraphyletic(a tax- the synlophe and bursa (Figs. 15, 16; Tables II, on with a common ancestor,but with exclusion III).Orthogenesis or directedevolution, with pro- of 1 or more descendants)(Hennig, 1966;Wiley, gressivetrends in characterdevelopment and par- 1981; Wiley et al., 1991). The resultingclassifi- allel trends involving a series of primitive and cation (cladistic or otherwise)would thus have advancedtaxa, has been discredited,and hypoth- been inconsistent with the phylogenetichistory eses based on such assumptionsare often both of these trichostrongylidsubfamilies (see opin- teleologicaland tautologicalin nature(see Mayr, ions on classification in Khalil and Gibbons 1982; Klassen, 1992; Brooks and McLennan, [1981] and Jansen [1989]). 1993).Although the majorityof characterpolarity Alternative hypotheses for these subfamilies decisions in the current analysis are consistent suggestedthat the Ostertagiinaeand Graphidi- with those developedby Durette-Dessetand Cha- inae could be sister-groups(sharing a common baud (1977, 1981) and Durette-Desset (1983, ancestor;also not supporteddue to the absence 1985)(via a form ofoutgroupcomparison), it was of a synapomorphyfor both subfamilies) (Fig. found to be unnecessaryto invoke extensivepar- 18B),or that the Ostertagiinaeand Graphidiinae allelism (for terminaltaxa) to explain the evolu- were more closely relatedto other subfamiliesof tion of these nematodes.The data matrixapplied Trichostrongylidae(Hoberg et al., 1993b) (Fig. to the 2 topologiesindicates that the currenthy- 18C). Resolution of this situation follows from pothesis is more internallyconsistent and pro- the recognition of putative synapomorphiesfor vides a more efficient explanationfor character the Haemonchinae + Ostertagiinae and un- evolution (Figs. 15, 16; Tables II, III). Thus, the equivocal synapomorphiesdiagnosing the Os- differencein concepts developed for the phylog- tertagiinae(to the exclusion of the Graphidiinae) eny of the subfamiliesof the Trichostrongylidae presented in the current study (Fig. 15). Thus, appearsto reside not in cladistic methodology, the hypothesis for multiple origins of the Oster- characterselection, or character-statepolariza- tagiinae is refuted, and the Graphidiinaeas the tion, but in the likely mechanismsof evolution. basal taxon of the Graphidiinae-cladeis postu- The relationships presented by Lichtenfels lated as the sister-groupfor the Haemonchinae (1987) for 2 major lineages within the Tricho- + Ostertagiinae(Fig. 15). strongylidaealso appear to lack strong support Aside from the putative phylogeny for the (Figs. 15, 17;Tables II, IV). This hypothesis,first Graphidiinaeand Ostertagiinae,the currenthy- outlined by Lichtenfels(1987) only for compar- pothesis for overall relationshipsof the subfam- ative purposes with the classificationproposed ilies differsconsiderably from that proposed by by Durette-Desset(1983, 1985), was based on a Durette-Desset and Chabaud (1981) and Dur- key to genera of the Trichostrongylidaedevel- ette-Desset (1985, 1992). The latter hypothesis oped by Gibbons and Khalil (1982). The system depicts an implicitly orthogeneticview of evo- of Gibbonsand Khalilincluded a varietyof char- lution (see Brooks and McLennan [1993] for a acters not evaluated by Durette-Desset (1983): discussion of the influence of orthogenesis on structureof the synlophe, bursa, genital cone, concepts of parasiteevolution). As such, it spec- cervical papillae, cephalic vesicle, and buccal ifies a primitiveand advancedcomponent in each tooth. Althoughthe resultingdendrograms differ lineage linked by progressive modification or substantially(Figs. 16, 17), the system of Gib- HOBERGAND LICHTENFELS-TRICHOSTRONGYLIDAE PHYLOGENY 989

Ostertagia Teladorsagia Marshallagia Spiculopteragia Longistrongylus Mazamastrongylus OSTERTAGIINAE 1 A 'i /

Graphidium Hyostrongylus

GRAPHIDIINAE

OR

Ostertagiinae Graphidiinae Ostertagiinae Graphidiinae

B/ BC

? FIGURE 18. Relationships postulated for the Graphidiinaeand the Ostertagiinae.18A. Multiple origins hypothesisfor the Ostertagiinaefrom the Graphidiinaepresented by Durette-Dessetand Chabaud(1977, 1981), and Durette-Desset (1981, 1985). Two lineages are recognizedwithin the Ostertagiinae,being derived from either Graphidiumor Hyostrongylus;under this scenarioOstertagiinae would be polyphyleticand Graphidiinae paraphyletic.18B. Alternatively,Graphidiinae could be the sister-groupof the Ostertagiinae.18C. Or, these subfamilies could be more closely related to other subfamilies of the Trichostrongylidae.The hypothesis for independentorigins of 2 lineageswithin the Ostertagiinaewas refutedby the currentanalysis, which recognizes Graphidiinaeas the sister-groupfor the Ostertagiinae+ Haemonchinae. bons and Khalil as interpreted by Lichtenfels The topology of the current hypothesis and (1987) was only marginally more efficient than that outlined by Lichtenfels (1987) also differed that of Durette-Desset (1983, 1985) (Tables III, substantially. In the latter, the basal dichotomy IV). was defined by the presence or absence of the 990 THEJOURNAL OF PARASITOLOGY,VOL. 80, NO. 6, DECEMBER1994 synlophe (Libyostrongylinae+ Trichostrongyl- utilize the intestine (Cooperiinae),or are found inae and Haemonchinae + Graphidiinae + in the stomachand intestine(Libyostrongylinae), Cooperiinae and Ostertagiinae).Although the or the intestine, stomach, or abomasum of the Haemonchinae,Ostertagiinae, and Graphidiinae definitive host (Trichostrongylinae).Thus, if the were grouped together, this clade also included intestine is the putative ancestral habitat, the the Cooperiinae in contrast to the present hy- stomach/abomasumwould have been colonized pothesis (Figs. 15, 17). With respect to the data independentlyon a minimum of 2 or 3 occasions base for the currentstudy, extensive parallelism during the radiation of the family (Fig. 19). A and reversalis requiredto describethe topology habitat shift is postulatedfor the ancestorof the of the dendrogram presented by Lichtenfels Graphidiinaeclade to account for the limited (1987) (Table IV; Fig. 17), and thus the hypoth- distributionof the Graphidiinae,Ostertagiinae, esis is not consideredto be a reliableestimate of and Haemonchinaein the stomach and aboma- phylogenyfor the Trichostrongylidae. sum. Withinthe Cooperiinaeclade, the common ancestor of the Libyostrongylinae + Tricho- Habitat,host, and biogeographicdistributions strongylinaemay have become associated with Previous attempts at understandingthe habi- the stomach/abomasumwhile maintaining the tat associations(localization within the host), de- ancestralhabitat. Alternatively,colonization of gree of parasite-hostcoevolution, and historical a new habitatcould have occurredindependently biogeographyfor the trichostrongylidshave been within the Trichostrongylinaeand the Libyo- limited. Skrjabinet al. (1954) reviewed general strongylinae.Refinement of this hypothesis and aspects of habitat utilization and host distribu- resolutionof the relationshipsof the Libyostron- tion (specificity).Durette-Desset and Chabaud gylinae and Trichostrongylinaewill be depen- (1977, 1981) and Durette-Desset(1982a, 1982b, dent upon generic- and species-level analyses 1985, 1992) provided a synopticexamination of within these subfamilies. the relationshipsof the Trichostrongyloideaand Hostassociations the Trichostrongylidaewith respectto theirhosts and geographicdistributions as an extension of The host range of the Trichostrongylidaeis hypotheses for parasite evolution. More inclu- ratherbroad, including a number of avian and sively, within the TrichostrongylidaeDr6zdz mammalianorders, but most generaand species (1967) and Jansen(1989) discussedthe evolution are the predominantnematodes parasitizingru- of the Ostertagiinaeand their cervid and bovid minants and artiodactylsthroughout the world hosts. Otherconsiderations have been limited to (Durette-Dessetand Chabaud,1977, 1981; Dur- species-level associations within particular ette-Desset, 1983, 1985). Based on the phylo- subfamilies, e.g., Hoberg et al. (1993a). Below genetic conclusions summarized by Durette- we provide a preliminaryevaluation of the cur- Desset (1985, 1992), the most highly evolved rent hypothesisfor the trichostrongylidswith re- subfamilies, Haemonchinae, Cooperiinae, and spect to habitat, hosts, and biogeography. Ostertagiinae,are primarilyparasites of the Bov- The evolution of site selection by trichostron- idae and Cervidae. However, these subfamilies gylids may be evaluated with respectto the par- were not consideredto be closely related (sum- asite phylogeny (Fig. 19). Mappingand optimi- marized in Fig. 16), each having been derived zation of data for habitat utilization onto the independentlyfrom a more primitive subfamily. presentphylogeny for the subfamiliesallows rec- Thus, parasitespresent in Bovidae and Cervidae ognition of the intestine as the ancestralhabitat were consideredmore closely relatedto those in for the family. This conclusion is in agreement ratites (Libyostrongylinae + Cooperiinae) or with Skrjabinet al. (1954) and corroboratedwith groupsof ancientmammals (Graphidiinae + Os- referenceto the intestinaldistribution of species tertagiinaeand Trichostrongylinae+ Haemon- of Molineidae (within the Trichostrongyloidea), chinae), and their currentdistributions could be Strongyloidea, and Ancylostomatoidea (out- explained by colonization rather than coevolu- groups in the currentanalysis). tion (Durette-Desset,1985, 1992). An extension Withinthe Trichostrongylidae,all membersof of this hypothesis suggestedthat diversification the Graphidiinae clade occur as adults in the of the Trichostrongylidaewas apparentlyiniti- stomachor abomasumof the definitivehost. The ated in the Eocene, with those taxa that occur in Cooperiinae clade has species that exclusively modern ruminantshaving been derived via in- HOBERGAND LICHTENFELS-TRICHOSTRONGYLIDAE PHYLOGENY 991

(INT) (INT/S/ABO) (INT/S) (INT) (S)

(S /AB O)

+ (8)

(INT): intestine (S): stomach (ABO): abomasum

FIGURE19. Cladogramfor the Trichostrongylidaeshowing the hypothesis presentedin the currentstudy with habitatlocalization in the gastrointestinalsystem mappedonto the parasitetree. Labelsare as follows:INT = intestine, S = stomach (in nonruminants),and ABO = abomasum.Relationships depicted are consistentwith the intestineas the ancestralhabitat; the stomach/abomasumwas colonizedonce in the ancestorof the Graphidi- inae clade and once or twice in the Cooperiinae clade to account for the distribution of some species of Libyostrongylinaeand Trichostrongylinae. dependent colonization in the Miocene, coincid- Although this level of resolution is not possible ing with the origins and radiation of the Bovidae with the current parasite phylogeny, certain gen- and Cervidae. Additionally, the predominant in- eralities about host associations among rumi- fluence on the distribution, evolution, and ra- nants are warranted. diation of the trichostrongylid fauna was consid- The overall distribution of ruminant hosts for ered to be ecological rather than phylogenetic. the trichostrongylids is compatible with a min- Thus, similarity in host diet, specifically herbiv- imum of 2 events of independent colonization ory, was recognized as the putative determinant in the Cooperiinae and Graphidiinae clades (Fig. of diversification (Durette-Desset, 1985, 1992). 20). Within the former clade, no clear relation- Jansen (1989) indicated that this scenario for ships among hosts and subfamilies are apparent, the radiation of the Trichostrongyloidea and although the Cooperiinae and the Trichostron- Trichostrongylidae did not appear well support- gylinae occur in ruminants. However, consid- ed. The phylogenetic analysis presented in the ering the Graphidiinae clade, coevolution of the current study provides a mechanism to assess Haemonchinae and the Ostertagiinae with bovid the hypotheses for diversification presented by and cervid hosts appears to have been extensive. Durette-Desset (1985, 1992). It is clear that host The dominance of these subfamilies among the associations are exceedingly complicated, as in- Bovidae, Camelidae, Cervidae, and Antiloca- dicated by the distribution of host groups mapped pridae indicates that the common ancestor of the onto the parasite phylogeny (Fig. 20). This may Haemonchinae + Ostertagiinae was already a be indicative that colonization at some level has parasite of the pecoran artiodactyls (Fig. 20). Co- strongly influenced diversification of this group evolution can secondarily be inferred by refu- of nematodes in agreement with concepts (but tation of the multiple origins hypothesis of the not specific hypotheses) outlined above. How- Ostertagiinae from the Graphidiinae (ofDurette- ever, in this case the relative importance of host Desset and Chabaud, 1977, 1981) in the present switching versus coevolutionary processes can study. With monophyly and a sister-group re- only be determined in the context ofgeneric-level lationship recognized for the Haemonchinae and and species-level analyses of parasites and hosts. Ostertagiinae, it is postulated that radiation of 992 THEJOURNAL OF PARASITOLOGY, VOL. 80, NO.6, DECEMBER1994

0

eiv3 0 ?o ?o 0•0 C) o ,o eS 22"i,,,," ooo 8, oo'\ An= Antilocapridae Av= Avian hosts Ba= Bathyergidae Bo= Bovidae Ca= Camelidae Ce= Cervidae H= Hyracoids La= Lagomorphs Le= Lemuridae P= Primates Ra= Ratites Ro= Rodents So= Soricidae Su= Suidae Ta= Tayassuidae Tr= Tragulidae FIGURE 20. Cladogramfor the Trichostrongylidaeshowing the hypothesispresented in the currentstudy and associationswith host groups mappedonto the parasitetree. these taxa coincided with the apparently rapid these geographic ranges have been influenced di- diversification of basal modern ruminants in the rectly by host biogeography and secondarily by Late Oligocene to Early Miocene about 23-28 climate, the latter probably as a determinant of million yr ago (see Vrba, 1985; Kraus and Mi- the suitability of habitat for development of free- yamoto, 1991; Allard et al., 1992). With respect living larval stages (Jansen, 1989; Suarez and to the Ostertagiinae, an hypothesis for extensive Cabaret, 1991). In the Holarctic, an emergent coevolution with bovids and cervids had pre- Beringia connecting Alaska and Chukhotka dur- viously been outlined by Dr6zdz (1965, 1967). ing much of the Tertiary and intermittently since Based on the present analysis, the hosts for the the late Pliocene (Matthews, 1981; Herman and basal trichostrongylids as the common ancestor Hopkins, 1980) was a determinant of distribu- of the Cooperiinae + Graphidiinae clades re- tion for hosts and parasites. Notably, dispersal mains to be determined. of Holarctic faunas across Beringia during the Tertiary and their cyclic vicariance and isolation Biogeography during the Late Pliocene and Quaternary would The trichostrongylids are cosmopolitan in dis- have directly influenced the diversification of tribution, with contemporary ranges occurring these parasite host assemblages. With respect to across all major biogeographic regions except Palearctic origins for most ruminant hosts of Antarctica (Fig. 21). Overall these are complex Cooperiinae, Ostertagiinae, and Haemonchinae, biogeographic relationships that must eventually initial diversification of these subfamilies thus be considered within the context of paleogeog- appears to have occurred in the Old World (Dur- raphy and paleocontinental reconstruction. De- ette-Desset, 1985, 1992) (Fig. 21). However, in tailed studies within each subfamily are required the late Tertiary and Pleistocene the predomi- to elucidate ancestral areas and address the rel- nant movement of Cervidae and Bovidae was ative importance of dispersal versus vicariance from the Palearctic to the Nearctic (Rausch, 1963; in the development of the fauna. Historically, Repenning, 1967; Hoffmann, 1976; Kurt6n and HOBERGAND LICHTENFELS-TRICHOSTRONGYLIDAE PHYLOGENY 993

ei ??" 0 ?o"9 ~?"' B No, Pa

Af= Africa As= Asia (& Australia) Ma= Madagascar Ne= Nearctic No= Neotropical Pa= Palearctic

FIGURE 21. Cladogramfor the Trichostrongylidae showing the hypothesis presented in thecurrent study and geographicrelationships for taxamapped onto the parasitetree.

Anderson, 1980; Geist, 1985), and large seg- However,late Pleistoceneextinctions of largeru- mentsof these faunasmay have experiencedrange minantscould also have had a substantialimpact expansion into North America during that in- on the fauna. Extirpation and local extinction terval. Thus, for ruminant parasites range ex- among many ruminantssuch as the saiga Saiga pansion would have largely been from west to tatarica(Linnaeus), Nearctic camelids, and other east across Beringiaduring the Quaternary(see Pleistocene and Recent mammals, e.g., Guthrie Hobergand Rickard, 1988; Hoberget al., 1989). (1968), Kahlke (1991), and Kurtrn and Ander- Likewise, extensive faunal interchangefollowed son (1980), is likely to have influencedthe dis- the emergenceof the PanamanianIsthmus link- tributionsof some contemporarytrichostrongy- ing North and South America in the Pliocene lids. Later domestication of bovids in Eurasia (Marshallet al., 1982; Vermeij, 1991). With re- (Epsteinand Mason, 1984; Ryder, 1984) and the spect to the latter,the extremeisolation of South subsequent transport of infected hosts beyond America through the Tertiary and the recent their typical ranges during contemporarytimes emergence of a permanent land connection be- has secondarilyresulted in a broad distribution tweenthe Nearcticand Neotropicalregions would for many taxa throughout the world, e.g., An- furtherrefute the putative phylogeneticand bio- drews (1973) and Rickard et al. (1993). These geographiclinkage of the Trichostrongylinaeand anthropogenicallydriven events, and a low level Haemonchinaeproposed by Durette-Dessetand of host specificity for many species of tricho- Chabaud(1977). strongyles(Suarez and Cabaret, 1991) will com- A synoptic elucidation of the historical bio- plicate our ability to recover the historical de- geography of hosts and parasites may be dra- velopment of host and geographicrelationships matically confounded by the degree of climato- for the Trichostrongylidae.In this respect, a logical perturbationand habitat fragmentation baseline for assessing the role of contemporary during the late Pliocene and Quaternary.Isola- introductionsof parasitesmay be derived from tion in refugialzones duringglacial maxima and examination of the faunas currentlypresent in poststadialvicariance of populationsacross Ber- exotic ruminantsof Australiaand New Zealand. ingia appearsto have promoted speciationwith- Resolution of the biogeographichistory for this in some assemblages,e.g., Hoberget al. (1993a). parasite-host assemblagewill follow from anal- 994 THE JOURNALOF PARASITOLOGY,VOL. 80, NO. 6, DECEMBER1994

yses at the level of genera and species of tricho- Rickard-Ballweber reviewed and commented on strongyles within the context ofpaleocontinental earlier versions of the manuscript. reconstruction. LITERATURECITED Summaryand conclusions ALLARD, M. W., M. M. MIYAMOTO, L. JARECKI, F. A preliminary phylogenetic analysis has been KRAUS,AND M. R. TENNArNT.1992. DNA sys- developed for the subfamilies of the Trichostron- tematics and evolution of the artiodactylfamily Bovidae. Proceedingsof the National Academyof gylidae. The relationships recognized in the cur- SciencesUSA 89: 3972-3976. rent analysis differed markedly from those that ANDREEVA,N. K. 1958. Atlas of helminths (Stron- have been proposed for this group at the subfam- gylata) of domestic and wild ruminants of Ka- ily level. Competing hypotheses suggested quite zakhstan.Veterinary Institute of the KazakhSec- tion ofVASKhNIL,Tashkent. [English Translation disparate relationships for the 6 subfamilies (Figs. (1978), Agricultural Research Service, United 15-17; Tables II-IV). The discordance in topol- States Departmentof Agriculture;Amerind Pub- ogy and tree length between the current hypoth- lishing Company,New Delhi, 206 p.] esis and those presented by Durette-Desset (1985) ANDREWS,J. R. H. 1973. A host-parasitechecklist of helminths of wild ruminantsin New Zealand. and Lichtenfels (1987) resulted largely from con- New ZealandVeterinary Journal 21: 43-47. clusions about relationships in the latter studies BAYLIS, H. A., AND R. DAUBNEY. 1926. Synopsis of that were based on assessments of single char- the families and generaof Nematoda. BritishMu- acters, overall morphological similarity, or unique seum of Natural History, London, United King- combinations of shared primitive characters. The dom, 277 p. D. D. Par- current is re- BROOKS, R., AND A. McLENNAN. 1993. hypothesis strongly supported, ascript, parasitesand the languageof evolution. quires a lower degree of parallel and convergent SmithsonianInstitution Press, Washington, D.C., evolution, has diagnosable synapomorphies for 429 p. all branches, and provides a more consistent ex- CHABAUD, A. G. 1959. Remarques sur la systema- des Bulle- planation of character evolution and phylogeny tique Nrmatodes Trichostrongyloidea. tin de la Soci&t6Zoologique de France 84: 473- at the subfamilial level. Previous hypotheses for 483. the diversification of the Trichostrongylidae were --, F. PUYLAERT, O. BAIN, A. J. PETTER,AND M. found to be untenable based on phylogenetic and C. DURETTE-DESSET. 1970. Remarques sur biogeographic criteria presented in the current l'homologieentre les papillescloacales des Rhab- dites et les c6tes dorsalesdes Strongylida.Comptes study. Rendus Hebdomadairesdes Seances de l'Acade- Work in progress will attempt to recognize ad- mie des Sciences (Paris)271: 1771-1774. ditional morphological characters that define the CRAM,E. 1927. Bird parasitesof the nematode sub- trichostrongylids and those useful in studying the orders Strongylata, Ascaridata, and Spirurata. United States National Museum Bulletin 140: 1- relationships within the family. The current anal- 465. form the foundation for eventual yses generic- DRrZDZ,J. 1965. Studieson helminthsand helmin- and species-level evaluations of the Ostertagi- thiases in Cervidae.I. Revision of the subfamily inae and other subfamilies. In doing so, we hope OstertagiinaeSarwar, 1956 and an attemptto ex- to elucidate basic concepts in the evolution of plain phylogenesisof its representatives.Acta Par- asitologicaPolonica 13: 445-481. the trichostrongyles and other parasitic nema- 1967. Studieson helminthsand helminthia- todes and provide a predictive data base with ses in Cervidae III. Historical formation of hel- which to evaluate historical biogeography and minthofaunain Cervidae.Acta ParasitologicaPo- the interrelationships of coevolution and host lonica 14: 287-300. M. C. 1981. A on the switching in the origins and diversification of this DURETTE-DESSET, hypothesis systematicposition of the Ostertagiinaewithin the economically significant fauna. Trichostrongylidea.In Workshopno. 14, system- atics and biology of Ostertagiasens. lat. (Nema- ACKNOWLEDGMENTS toda: Trichostrongylidae),J. Jansen and L. M. Gibbons (eds.). Parasitology82: 175-177. in We thank Arthur Abrams for assistance 1982a. Sur les divisions g6niriques des preparation of the figures. Patricia Pilitt assisted N6matodesCooperiinae. Annales de Parasitologie in describing some morphological characters and Humaine et Compar6e57: 383-387. S1982b. Sur les divisions des reviewed an early version of the manuscript. gin6riques Ostertagiinae.Annales de Parasito- Daniel Brooks and shared N~matodes Greg Klassen kindly logie Humaine et Compar6e57: 375-381. ideas concerning phylogenetic analysis and or- S1983. Keys to the generaof the superfamily thogenesis in evolution. Scott Gardner and Lora Trichostrongyloidea.No. 10. In CIH keys to the HOBERGAND LICHTENFELS-TRICHOSTRONGYLIDAE PHYLOGENY 995

nematodeparasites of vertebrates,R. C. Anderson HERMAN,Y., ANDD. M. HOPIUNS.1980. Arctic Ocean and A. G. Chabaud(eds.). CommonwealthAgri- climate in late Cenozoic time. Science 209: 557- culturalBureaux, Farnham Royal, England,p. 1- 562. 86. HOBERG,E. P., ANDJ. R. LICHTENFELS.1992. Mor- . 1985. Trichostrongyloidnematodes and their phology of the synlophe and genital cone of Pa- vertebratehosts: Reconstructionof the phylogeny rostertagia heterospiculum(Trichostrongylidae) of a parasiticgroup. Advances in Parasitology24: with comments on the subfamilial placement of 239-306. the genus. SystematicParasitology 22: 1-16. M. FRUETEL,AND L. G. 1989. Syn- S1989. Nomenclaturepropos6e pour les es- , RICKARD. p~ces d6critesdans la sous-familledes Ostertagi- lophe of Nematodirusodocoilei (Trichostrongyl- inae Lopez-Neyra,1947. Annalesde Parasitologie oidea) from deer and caribou in North America Humaine et Comparke64: 356-373. with comments on the evolution of Nematodirus 1992. Phylogenie des nematodes Tricho- spp. amongthe Cervidae(Artiodactyla). Canadian strongyloideavue a travers celle de leurs hotes Journalof Zoology 67: 1489-1494. vertebres.Parassitologia 34: 1-16. --, J. R. LICHTENFELS,AND P. A. PILITT.1993a. , ANDA. G. CHABAUD.1977. Essai de classi- Comparativemorphology of Ostertagiamossi and ficationdes N6matodesTrichostrongyloidea. An- Ostertagia dikmansi (Trichostrongylidae)from nales de ParasitologieHumaine et Comparke52: Odocoileus virginianusand comments on other 539-558. Ostertagia spp. from the Cervidae. Systematic 1981. Nouvel essai de classifi- Parasitology24: 111-127. -, AND . cation des N6matodes Trichostrongyloidea.An- - , - , AND . 1993b. Synlophe of nales de ParasitologieHumaine et Comparbe56: Hyostrongylusrubidus (Trichostrongylidae), with 297-312. evaluation of structuralcharacters supporting af- - , AND -. 1993. Nomenclature des Stron- filiationwith the Ostertagiinae.Journal of the Hel- gylidae au-dessus du groupe-famille.Annales de minthologicalSociety of Washington60:219-233. ParasitologieHumaine et Compar&e68: 111-112. - , - , AND . 1993c. Synlophe of EPSTEIN,H., ANDI. L. MASON. 1984. Cattle. In Evo- Cooperianeitzi (Trichostrongylidae:Cooperiinae) lution of domesticatedanimals, I. L. Mason (Ed.). with comments on vulval inflationsand hypertro- LongmanInc., New York, New York, p. 6-27. phy of cuticularridges among the trichostrongy- GEIST,V. 1985. On Pleistocene bighorn sheep: Some lids. Journal of the HelminthologicalSociety of problemsof adaptation,and relevanceto today's Washington60: 153-161. Americanmegafauna. Wildlife Society Bulletin 13: - , AND L. G. RICKARD. 1988. Morphology of 351-359. the synlophe of Nematodirusmaculosus (Tricho- GIBBONS,L. M. 1977. Revision of the genera Lon- strongyloidea)with comments on the evolution of gistrongylusLeRoux 1931, KobusinemaOrtlepp, Nematodirusspp. among the Caprinae(Artiodac- 1963 and BigalkenemaOrtlepp, 1963 (Nematoda: tyla). Proceedingsof the HelminthologicalSociety Trichostrongylidae).Journal ofHelminthology 51: of Washington55: 160-164. 41-62. HOFFMANN,R. S. 1976. An ecological and zoogeo- 1979. Revision of the genus Haemonchus graphicalanalysis of animal migrationacross the Cobb, 1898 (Nematoda:Trichostrongylidae). Sys- Beringland bridgeduring the QuaternaryPeriod. tematic Parasitology1: 3-24. In Beringiain the Cenozoic,V. L. Kontrimavichus 1981. Revision of the Africanspecies of the (ed.). Akademia Nauk, SSSR, Vladivostok, Rus- genus CooperiaRansom, 1907 (Nematoda,Trich- sia, p. 354-367. ostrongylidae).Systematic Parasitology 2:219-252. JANSEN, J. 1989. A concise history of the Ostertagi- , AND L. F. KHALIL. 1982. A key for the iden- inae Lopez-Neyra,1947 (Nematoda:Trichostron- tificationof generaof the nematodefamily Tricho- gyloidea)and a discussionon its composition.Acta strongylidae Leiper, 1912. Journal of Helmin- Leidensia58: 151-159. thology 56: 185-233. KAHLKE, R.-D. 1991. Pleistocene distributional and ,AND . 1983. Morphology of the genital evolutionaryhistory of the genusSaiga Gray, 1843 cone in the nematode family Trichostrongylidae (Mammalia, Artiodactyla,Bovidae) in the Pale- and its value as a taxonomic character.In Con- arctic. VertebrataPalasiatica 29: 314-322. cepts of nematodesystematics, A. R. Stone, H. M. KHALIL, L. F., AND L. M. GIBBONS. 1981. The Platt, and L. F. Khalil (eds.). SystematicsAsso- subfamily OstertagiinaeSarwar, 1956. Parasitol- ciation SpecialVolume 22, Academic Press,Lon- ogy 82: 177-181. don, United Kingdom, p. 261-271. KLASSEN,G. J. 1992. Coevolution: A history of the GLEN,D. R., ANDD. R. BROOKS. 1985. Phylogenetic macroevolutionaryapproach to studying host- relationships of some strongylatenematodes of parasiteassociations. Journalof Parasitology78: primates.Proceedings of the HelminthologicalSo- 573-587. ciety of Washington52: 227-236. - , R. D. MooI, ANDA. LOCKE.1991. Consis- GUTHRIE,R. D. 1968. Paleoecology of the large- tency indices and random data. Systematic Zo- mammalcommunity in interiorAlaska during the ology 40: 446-457. LatePleistocene. American Midland Naturalist 79: KRAUS,F., ANDM. M. MIYAMOTO.1991. Rapid clad- 346-363. ogenesis among the pecoranruminants: Evidence HENNIG, W. 1966. Phylogenetic systematics. Uni- from mitochondrialDNA sequences. Systematic versity of Illinois Press, Urbana, 263 p. Zoology 40: 117-130. 996 THEJOURNAL OF PARASITOLOGY, VOL. 80, NO.6, DECEMBER1994

KURTtN,B., ANDE. ANDERSON.1980. Pleistocene from ruminantsin North Americawith a descrip- mammals of North America. Columbia Univer- tion of the synlophe and the female. Canadian sity Press, New York, New York, 442 p. Journalof Zoology 67: 1708-1714. LANCASTER,M. B., AND C. HONG. 1990. The iden- - , E. P. HOBERG, N. M. ALLEN, G. L. ZIMMERMAN, tification of females within the subfamily Oster- AND T. M. CRAIG. 1993. Spiculopteragia spicu- tagiinae Lopez-Neyra, 1947. Veterinary Parasi- lopteraand S. asymmetrica(Nematoda: Tricho- tology 35: 21-27. strongyloidea)from red deer (Cervuselaphus) in LEIPER,R. T. 1912. Check-listofhelminthes parasitic Texas. Journalof Wildlife Diseases 29: 512-515. in cattle.Journal of the LondonSchool of Hygiene RYDER,M. L. 1984. Sheep. In Evolution of domes- and TropicalMedicine 1:115-123. ticated animals, I. L. Mason (ed.). LongmanInc., LICHTENFELS,J.R. 1977. Differencesin cuticularridges New York, New York, p. 63-85. among Cooperiaspp. of North American rumi- SARwAR,M.M. 1956. Studieson some trichostrongy- nants with an illustratedkey to species. Proceed- lids of ruminantsfrom the Indo-Pakistansub-con- ings of the HelminthologicalSociety of Washing- tinent. Biologia 2:145-215. ton 44:111-119. SKRIUABIN,K. I., N. P. SHIKHOBALOVA,AND R. S. 1987. Phylogenetic inference from adult SHUL'TS.1954. Essentialsof nematodologyIII. morphologyin the Nematoda;with emphasis on Trichostrongyloidsof animalsand man. Moscow, the bursatenematodes, the Strongylida;advance- Academyof SciencesUSSR. [EnglishTranslation ments (1982-1985) and recommendationsfor fur- by Israel programfor ScientificTranslations, Je- ther work. InternationalJournal for Parasitology rusalem, 1960, 704 p.] 17: 269-279. - , T. I. POPOVA,S. N. BOEV,AND , ANDE. P. HOBERG. 1993. The systematics of S. L. DELIAMURE.1952. Keys to the parasitic nematodesthat cause ostertagiasis in domesticand nematodesDE',III. Strongylata.Moscow, Academyof wild ruminantsin North America:An updateand Sciences USSR. [English Translation by Israel a key to species. VeterinaryParasitology 46: 33- Program for Scientific Translations, Jerusalem, 53. 1961, 890 p.] LUNDBERG,J. G. 1972. Wagner networks and ances- SUAREZ,V. H., AND J. CABARET. 1991. Similarities tors. SystematicZoology 21: 398-413. between species of the Ostertagiinae(Nematoda: MARSHALL,L. G., S. D. WEBB,J. J. SEPKOSKI,JR., AND Trichostrongyloidea)in relationto host-specificity D. M. RAUP. 1982. Mammalianevolution and and climatic environment. Systematic Parasitol- the greatAmerican interchange. Science 215:1351- ogy 20: 179-185. 1357. SWOFFORD,D. L. 1985. Phylogeneticanalysis using MATTHEWS,J. V., JR. 1981. Tertiary and Quaternary parsimony,version 2.4. Illinois Natural History environments:Historical background for an anal- Survey. ysis of the Canadianinsect fauna.Memories of the TRAVASSOs,L. 1937. Revisao da familiaTrichostron- EntomologicalSociety of Canada108: 31-86. gylidaeLeiper, 1912. MonographiasInstituto Os- MAYR,E. 1982. Processesof speciation. In Mecha- waldo Cruz 1:1-512. nisms of speciation,C. Barigozzi(ed.). Liss, New VERMEu, G. J. 1991. Whenbiotas meet:Understand- York, New York, p. 1-19. ing biotic interchange.Science 253: 1099-1104. MEIER,R., P. KORES,AND S. DARWIN. 1991. Ho- VRBA,E. S. 1985. African Bovidae: Evolutionary moplasy slope ratio:A bettermeasurement of ob- events since the Miocene. South AfricanJournal servedhomoplasy in cladisticanalyses. Systematic of Science 81: 263-266. Zoology 40: 74-88. WILEY,E. O. 1981. Phylogenetics,the theory and RAUSCH,R. L. 1963. A review of the distributionof practiceof phylogeneticsystematics. John Wiley HolarcticRecent mammals. In Symposiumon Pa- and Sons, Inc., New York, New York, 439 p. cific Basinbiogeography, J. L. Gressit(ed.). Tenth - , D. SIEGEL-CAUSEY,D. R. BROOKS,AND V. A. Pacific Science Congress,Bishop Museum Press, FUNK. 1991. The compleat cladist. A primerof Honolulu, Hawaii, p. 29-43. phylogenetic procedures.University of Kansas, REPENNING,C. A. 1967. Palearctic-Nearctic mam- Museum of Natural History, Special Publication maliandispersal in the LateCenozoic. In The Ber- 19, 158 p. inglandbridge, D. M. Hopkins(ed.). Stanford Uni- YORKE,W., ANDP. A. MAPLESTONE.1926. The nema- versity Press, Stanford,California, p. 288-311. tode parasitesof vertebrates.J. and A. Churchill, RICKARD,L. G., ANDJ. R. LICHTENFELS.1989. Nema- London, U.K., 536 p. todirusarchari (Nematoda: Trichostrongyloidea)