Phylogenetic Relationships of Some Strongylate Nematodes of Primates

Proc. Helminthol. Soc. Wash. 52(2), 1985, pp. 227-236

DAVID R. GLEN' AND DANIEL R. BROOKS Department of Zoology, University of British Columbia, 2075 Wesbrook Mall, Vancouver, B.C. V6T 2A9, Canada

ABSTRACT: Numerical phylogenetic analysis of 12 species of nematodes in the genus Oesophagostomum in- habiting primarily primates is performed. Based on 22 homologous series, treated as 25 numerical characters, the subgenera Ihlea, Lerouxiella, and Conoweberia are considered a single monophyletic group, the subgenus Conoweberia. Host relationships support an interpretation predominantly of coevolution, with little host switching. Biogeographic relationships implied by the phylogenetic tree support an interpretation of African origin, with subsequent dispersal to Southeast Asia, followed by secondary re-invasion of Africa by these helminths and their hosts.

As part of a study on the parasitological evi- buccal capsule to indicate subfamilies. For ex- dence pertaining to the phylogeny of the Great ample, Oesophagostomum belongs in the Oe- Apes, we have analyzed the phylogenetic rela- sophagostominae, a subfamily characterized by tionships of the members of the nematode genus small cylindrical buccal capsules (except those in Oesophagostomum Molin, 1861, which inhabit Australian marsupials), and it is likely that at primates. Oesophagostomum is a member of the this level a number of groups are either para- or superfamily Strongyloidea. This superfamily is polyphyletic. The Oesophagostominae contains almost exclusively parasitic in mammals, and is three tribes, two of which, Bourgelatioidinea and usually characterized by having a high degree of Oesophagostominea, possess a well-developed host specificity. Although a number of classifi- cervical groove. In the Bourgelatioidinea the cations of the Strongyloidea exist (Popova, 1955; groove encompasses the entire body and in the Yamaguti, 1961; Chabaud, 1965), we have relied Oesophagostominea it is restricted to the ventral heavily on the most recent attempt (Lichtenfels, surface. Two genera are placed in the Oesopha- 1980). One of the major problems with earlier gostominea, Daubneyia (Le Roux, 1940), found classifications is that too much emphasis was in African suids, and Oesophagostomum, re- placed on too few characters. ported from a number of mammals including primates. Authors of earlier classifications of the Stron- Within Oesophagostomum a number of sub- gyloidea relied heavily on the morphology of genera have been proposed (Table 2), and this buccal capsules. Separation of families and splitting has been widely criticized. Thornton's subfamilies was difficult and arbitrary and there (1924) criticisms stemmed from the fact that the was little indication of the relation between splitting up of the genus Oesophagostomum was parasite and host evolution below the super- based primarily on host records, and not on char- family level. (Lichtenfels, 1980:1) acteristics of the parasites themselves. Yamaguti Lichtenfels proposed two additional characters (1961) supported these criticisms as follows: in an attempt to make the classification more I agree with Goodey (1924), Thornton (1924), reflective of the evolutionary process; (1) the and Baylis and Daubney (1926) in that the shape of the female ovejector and (2) the ar- division of the genus into four subgenera (Oe- rangement of rays associated with the male cop- sophagostomum, Proteracrum and Hystera- ulatory bursa. Lichtenfels used these two char- crum of Railliet and Henry, 1913; and Cono- acters to separate two of the four families of weberia of Ihle, 1922) is unnecessary, and so Strongyloidea. Following this scheme, Oesopha- are the additional subgenera Bosicola (San- gostomum belongs to the family Chabertiidae ground, 1929), Ihleia [sic] of Travassos and (Table 1). Lichtenfels relied on the shape of the Vogelsang, 1932, Hydsonia, Le Roux, 1940 and Pukuia Le Roux, 1940. (p. 394)

1 Present address: Faculty of Medicine, University We examined in detail three of these subgenera, of Calgary, Calgary, Alberta, Canada. Lerouxiella, Ihlea, and Conoweberia. All species

Table 1. Partial classification of the Strongyloidea according to Lichtenfels (1980).

Superfamily Family Subfamily Tribe

Strongylidae Deletrocephalidae Syngamidae Strongyloidea Chabertiinae Chabertiidae Cloacininae Bourgelatioidinea Oesophagostominae Oesophagostominea Bourgelatiinea

in these subgenera have well-developed esoph- oped by Dr. J. S. Farris (State University of New York, ageal funnels, consisting of three sclerotized Stony Brook). plates, each with a denticle that may project into Character analysis the posterior portion of the buccal capsule. The A total of twenty-two characters was used in this subgenus Ihlea is characterized by the presence analysis. For a summary of oesophagostomine mor- of six denticles. The degree of development of phology, see Chabaud and Durette-Desset (1973). Some this character is unique to these three subgenera, characters, such as the shape of the external elements, form fairly simple binary characters, whereas others, and it separates them from other members of the such as the number of denticles, form more complex genus Oesophagostomum. multi-state characters. All twenty-two characters and their character states are summarized in Table 3. Some Materials and Methods characters, for example, the number of elements in the Specimens belonging to the subgenera Lerouxiella, external corona, are not sufficiently explained in the Conoweberia, and Ihlea were borrowed from various table, and the following sub-sections provide further parasite collections. Table 2 lists the species examined, explanation. and the museums from which specimens were ob- 1) NUMBER OF ELEMENTS IN THE EXTERNAL CROWN: tained. The specimens were stored in a mixture of 70% The only way to clearly establish the number of exter- ethanol and 5% glycerine, and cleared in lactophenol nal elements is to do en face examination of the anterior before examination. A drawing tube was used to pre- pare Figures 1-4. Specimens of O. susannae, O. raillieti, O. zukow- Table 2. Oesophagostomum species and specimen lo- skyi, and O. ventri were not available. For these species cation. original descriptions and diagrams were used. Papers in which additional characters and measurements were Museums proposed were also used (for example, Travassos and Vogelsang, 1932). Although more than 20 nominal Subgenus Species ABC species have been assigned to Ihlea, Conoweberia, and Lerouxiella, only 12 species are recognized as valid in Conoweberia blanchardi Travassos and X X this analysis. This agrees with the most recent key for Vogelsang, 1932 these subgenera, in which 11 species were recognized aculeatum Linstow, 1879 X (Chabaud and Durette-Desset, 1973). The only species bifurcum Creplin, 1849 XXX added is O. brumpti. On the basis of its overall length, ovatum Linstow, 1906 X shape of the buccal capsule, and size of its spicules it pachycephalum Molin, 1861 X X X is distinct from other members of the study group. raillieti Travassos and However, only two specimens were available, and ide- Vogelsang, 1932 ally more specimens are required to more accurately zukowskyi Travassos and establish the ranges of some characters. Vogelsang, 1932 The systematic technique used in this analysis was brumpti Raillieti and Henry, X cladistics or phylogenetic systematics (Hennig, 1966; 1905 Wiley, 1981). A number of precis of this methodology Ihlea stephanostomum Stossich, XXX have been published (Brooks et al., 1984, 1985), and 1904 we will add only statements that refer specifically to ventri Thornton, 1924 the present study. Polarization of character transfor- Lerouxiella xeri Ortlepp, 1922 X mations was based on the outgroup criterion. These suzannae LeRoux, 1940 character transformations were summarized in a nu- merically coded data matrix (Table 4). Analysis of the Museums: A—British Museum of Natural History (London); polarized character transformations was aided by use B—United States National Museum (Beltsville); C-Museum of the PHYSYS computer systematics system, devel- National D'Histoire Naturelle (Paris).

S.M.P

.. E.L.C.

-L.R

M.C.

.05mm Figure 1. Oesophagostomum bifurcum en face view. Figure 2. Oesophagostomum stephanostomum en Abbreviations: S.M.P., sub-medial papillae; O.A., oral face view. Abbreviations: S.M.P., sub-medial papillae; aperture; L.P., lateral papillae; E.L.C., external leaf E.L.C., external leaf crown; L.P., lateral papillae; M.C., crown; M.C., mouth collar. mouth collar. ends of worms. Many descriptions however, have been based on whole mounts, and the number of elements of the funnel, and in Ihlea there are six denticles pres- have usually been placed at ten. In a few studies en ent. As with the number of external elements, it is very face views have been examined, and in some of these difficult to count the number of denticles when looking the number of elements have been reported at 12-16, at a lateral section of a whole mount. A serial section and in three species the number reported was numer- through the funnel is necessary to accurately establish ous (30-40 elements). Preparation of en face views of the number of denticles. To the best of our knowledge all the species would have been useful in resolving this serial sections of O. pachycephalum have not been dispute. Unfortunately, however, not enough material made, but unfortunately not enough material was was available to permit sectioning. Regardless of available to permit sectioning. These would be partic- whether 10-16 elements are present, there are at least ularly interesting in light of its large number of external two clearly defined character states in the study group, elements and the shape (V) of the esophageal funnel. those species with few elements (10-16), and those with Only shared derived or synapomorphic traits are ca- many elements (30-40). pable of indicating natural groups. Because of this, de- 2) PATTERN OF ELEMENTS OF THE EXTERNAL CORONA— termining which characters are derived (apomorphic) EN FACE VIEW: There are two states evident for this and which are primitive (plesiomorphic) is very im- character. For most members of the subgenus Conowe- portant. The most general method for determining beria the external elements converge towards the center character polarity is the outgroup method (Lundberg, of the oral aperture (Fig. 1). In contrast, those in O. 1972; Wiley, 1981). This method is based on the fol- stephanostomum and O. ventri do not converge (Fig. lowing concept: 2). Given two characters that are homologues and found 3) PATTERN OF ESOPHAGEAL SCLEROTIZATION: All the within a single monophyletic group, the character species in the study group have a well-developed that is also found in the sister group is the plesio- esophageal funnel. In most species the funnel is de- morphic character whereas the character found within scribed as cup- or goblet-shaped, the sides being curved. the monophyletic group is the apomorphic charac- In members of the subgenus Ihlea and in O. pachy- ter. (Wiley, 1981, p. 139) cephalum the funnel is V-shaped, the sides being straighter. That is, the character state present in at least one mem- 4) SHAPE OF THE ELEMENTS OF THE EXTERNAL CORONA: ber of the study group (or ingroup), and in some species Two states are evident for this character. For most outside the ingroup (the outgroup), will be primitive. members of the subgenus Conoweberia the elements Usually the most suitable outgroup is the sister group, are pointed and triangular in shape. In those species or a group that is closely related to the ingroup. In with more than 30 external elements, the elements are order to determine which members of the genus Oe- more rounded and cylindrical in shape. sophagostomum are most suitable as outgroups, the 5) NUMBER OF DENTICLES IN THE ESOPHAGEAL FUNNEL: phylogenetic relationships of the various subgenera have The pattern and the degree of development of the been partially reconstructed (Fig. 3). Information for esophageal denticles is unique to the study group. In this cladogram was obtained from keys published by members of the subgenera Conoweberia and Leroux- Chabaud and Durette-Desset (1973) and Lichtenfels iella there are three denticles projecting into the lumen (1980). The cladogram suggests that the subgenera O.

suggests that primitively the papillae are close to the ventral groove. Character states designated "0" are the plesiomorphic states, and positive whole numbers indicate the derived states. Reversing this convention (for binary characters) would not affect the analysis so long as the same convention was followed for an entire data set. The coded values and their corresponding character states are summarized in Table 3. The character states found in each species of the ingroup are summarized in Table 4. A value of "9" was entered for missing data; there are two reasons for such entries. For some species, for example, O. raillieti, specimens were not available, and the original descriptions were not sufficiently detailed to establish the nature of all characters Figure 3. Partial phylogenetic reconstruction of the and character states used in this analysis. Also, sec- genus Oesophagostomum. Terminal taxa are subgen- tioning of specimens was not possible and consequent- era. ly some character states in a few species could not be determined. One of the statistics used to measure the fit of particular characters to a hypothesized phylogeny is the Bosicola and O. Oesophagostomum are the most ap- consistency (c) index (Farris, 1970). This value is cal- propriate outgroups. culated by dividing the minimum possible number of For two types of characters, however, the above out- character transformations by the actual number of groups were not sufficient for completely determining transformations hypothesized from the phylogeny. For the polarity of transformation series. In these cases the example, a c-index of 1.0 (100%) for a particular char- functional outgroup method outlined by Watrous and acter suggests that there is no postulated homoplasy Wheeler (1981) was used. This method is based on the for that character. Rather than list the c-index for each idea that, in those instances where the outgroup is un- character, an overall c-index is usually listed. This val- able to determine the polarity of a transformation, it ue represents an average fit of all the characters used is best to use the most plesiomorphic members of the to the hypothesized tree topology. ingroup as a functional outgroup. The "outgroup com- Three characters in this analysis were split into two parison need not be constrained by nomenclatural rank transformation series each. The reason for this was that or Linnaean hierarchical structures" (Watrous and leaving them as a single transformation series resulted Wheeler, 1981). Thus, determining the most plesio- in artificially low c-indices. For example, there are four morphic members of the ingroup is based on characters character state ranges for male body length: 6-13 mm, polarized using the initial outgroup (the subgenera Bo- 13-18 mm, 18-25 mm, and less than 6 mm. The range sicola and Oesophagostomum). An example of this in the outgroup is 6-13 mm and this state was coded method involves the esophageal funnel. All members as 0. This suggests two trends, one of increasing length of the study group have well-developed sclerotized fun- to 13-18 mm and 18-25 mm, and one of decreasing nels with projecting denticles. This character is absent length to less than 6 mm. If the four character states in the outgroup subgenera, and consequently these out- are coded as: 6-13 mm = 0, 13-18 mm = 1, 18-25 groups are unable to determine which character state, mm = 2, and <6 mm = 3, a c-index of 0.60 is calcu- cup-shaped or V-shaped, is primitive. However, on the lated. This reflects the fact that PHYSYS reads the basis of other characters, O. zukowskyi, O. bifurcum, transformation series as O. brumpti, and O. aculeatum, are the most plesiomorphic members of the ingroup, and because all these 0-1-2-3 species have a cup-shaped funnel, this character state which suggests that there is an increase in length fol- is hypothesized to be primitive. The same reasoning lowed by a secondary decrease. However, if the trans- was used to polarize the transformation series for the formation is coded as character involving the length of the esophageal funnel. The other type of character that required a functional 0-1-2 outgroup was that in which there was considerable variation in the outgroup. For example, in the subgenus O. Bosicola the cervical papillae are situated close to a c-index of 1.00 is calculated. In order to code for this the ventral groove, and in the subgenus Oesophagos- second transformation series the character male body tomum they are more posterior, being closer to the length was split into two transformation series. The esophageal swelling. Within the ingroup the papillae other two characters treated in this manner were female occur between the above two extremes. Without fully body length and male body width. resolving the relationships between the subgenera Bosi- cola and Oesophagostomum it is impossible to tell Results whether the trend in the ingroup is towards the papillae moving anteriorly (from that state inthe subgenus Oe- Twenty-two characters were studied and used sophagostomum), or posteriorly (from that state in the to reconstruct the phylogeny of the 12 species in subgenus Bosicola). Appealing to a functional outgroup the subgenera Conoweberia, Ihlea, and Leroux-

Table 3. Summary of characters and coded character Table 3. Continued. states. Character Character Character states Coding Character states Coding _ Transverse processes Absent 0 Male body length: Present 1 Increasing transformation 6-13 mm 0 Distance of vulva from 0.50-0.75 mm 0 13-1 8 mm 1 posterior end 0.35-0.50 mm 1 18-25 mm 2 Maximum female body width 0.30-0.50 mm 0 Decreasing transformation 6-13 mm 0 0.50-0.65 mm 1 <6 mm 1 0.65-1. 10 mm 2 Female body length: Maximum male body width: Increasing transformation 7-13 mm 0 Increasing transformation 0.30-0.60 mm 0 13-20 mm 1 0.60-1.20 mm 1 20-30 mm 2 Decreasing transformation 0.30-0.60 mm 0 Decreasing transformation 7-13 mm 0 <0.30mm 1 ^ / mm Number of elements in the 8-14 0 external corona 30-40 1 Pattern of external elements— Figure 1 0 en face section Figure 2 1 iella. Figure 4 represents the most parsimonious Pattern of esophageal Absent 0 phylogeny for these taxa and characters. The sclerotization Cup-shaped 1 character transformations used in this analysis V-shaped 2 have been mapped on the cladogram. Those Shape of external elements Triangular 0 characters postulated to exhibit some homoplasy Rounded 1 are marked with a star. The overall consistency Number of esophageal teeth Absent 0 index for this cladogram is 80%. This figure rep- 3 1 6 2 resents an average measure of fit for all characters Distance of ventral groove 0.16-0. 23 mm 0 used in the hypothesized tree topology. The in- from anterior end 0.23-0.30 mm 1 dividual c-indices for all the characters are listed 0.30-0.45 mm 2 in Table 5. A c-index of 80% is a high consis- Width of cephalic distension 0.1 2-0.25 mm 0 tency, representing a good fit of the data to the 0.25-0.32 mm 1 tree topology. Distance of cephalic papillae 0.30-0.40 mm 0 A number of characters in this analysis are from anterior end 0.40-0.50 mm 1 0.50-0.60 mm 2 used as absolute lengths. That is, they were not Length of esophagus 0.30-0.40 mm 0 converted to ratios. These include such charac- 0.70-1. Omm 1 ters as esophageal length, female tail length, and 1.0-1. 4 mm 2 length and width of the cephalic distension. In a Width of esophagus 0.09-0. 16 mm 0 few descriptions and diagnoses these characters 0.1 6-0. 20 mm 1 have been presented as ratios. During the present 0.20-0.30 mm 2 study these characters were coded as ratios (pro- Length of external elements 0.025-0.040 mm 0 portional to overall body length and width), and 0.012-0.025 mm 1 it was found that doing this did not affect the Ratio of buccal capsule length 1/1.3-1/1.5 0 to width 1/2.5-1/3.2 1 topology of the tree but did decrease the c-index. 1/3.6-1/4.1 2 For this reason these characters were left as ab- Width of buccal capsule 0.020-0.050 mm 0 solute lengths. Furthermore, removing these 0.050-0.070 mm 1 characters had little effect on the cladogram. Of 0.080-0. 140 mm 2 the 1 1 nodes, two collapsed resulting in one po- Length of esophageal funnel 0.024-0.040 mm 0 lytomy. The rest of the tree remained identical 0.040-0. 10 mm 1 with the fully resolved tree shown in Figure 4. Length of spicule 0.70-1.25 mm 0 0.50-0.70 mm 1 The monophyletic nature of the 1 2 species in 1.25-2.00 mm 2 this study is supported by three characters. Two Length of female tail 0.20-0.40 mm 0 of these involve the esophageal funnel. All mem- 0.1 3-0. 20 mm 1 bers in the study group have well-developed sclerotized funnels that are either cup- or

Table 4. Data for Oesophagostomum analysis.

Characters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Outgroup 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 blanchardi 1 0 1 0 0 0 1 0 1 1 0 2 1 1 1 1 1 1 2 0 0 0 1 0 0 aculeatum 0 0 1 0 0 0 1 0 1 1 0 0 0 0 1 1 0 1 0 1 1 0 0 0 0 bifurcum 0 0 0 0 0 0 1 0 1 0 0 0 0 0 1 1 0 0 0 0 9 0 0 0 0 brumpti 0 0 0 0 0 9 1 0 1 0 0 0 0 0 1 1 0 0 0 1 1 0 0 0 0 ovatum 1 0 1 0 0 9 1 0 1 2 1 2 9 1 1 2 2 1 2 9 9 0 9 1 0 pachycephalum 2 0 2 0 1 9 2 1 1 2 1 2 2 2 1 2 2 1 2 0 0 0 2 1 0 raillieti 1 0 1 0 0 9 1 0 1 2 9 9 2 9 1 1 2 1 2 0 0 0 2 1 0 susannae 0 0 1 0 0 9 1 0 1 1 0 1 0 0 1 1 0 1 2 1 1 1 1 0 0 ventri 2 0 2 0 1 1 2 1 2 2 1 2 2 2 1 2 2 1 2 0 0 0 2 1 0 xeri 0 0 1 0 0 0 1 0 1 1 0 1 0 0 1 1 0 1 2 1 1 1 1 0 0 zukowskyi 0 1 0 1 0 0 1 0 1 0 0 1 0 0 1 0 1 0 1 0 0 0 0 0 1 stephanostomum 2 0 2 0 1 1 2 1 2 2 1 2 2 2 1 2 2 1 2 0 0 0 2 1 0

V-shaped. Also, projecting into the lumen of the tained, but that they be considered junior syn- esophageal funnel are three (or six) denticles. The onyms of the subgenus Conoweberia. The reason third character unique to the study group in- for this is that if Ihlea and Lerouxiella are rec- volves the size of the elements of the external ognized, there is no way to group the remaining corona. In all species they are in the range of eight species in a manner consistent with their 0.012-0.025 mm. genealogy without assigning each species to its The parasite phylogeny presented in this study suggests that two of the subgenera studied (Ihlea and Lerouxielld) are monophyletic, and one Table 5. Consistency indices of characters used in the (Conoweberid) is paraphyletic. A paraphyletic analysis of Oesophagostomum. group is a group that includes a common ancestor and some but not all its descendants (Farris, Consis- 1974). The monophyly of Lerouxiella is sup- tency ported by the presence of transverse processes in No. Character index the buccal capsule, and the monophyly of Ihlea 1 Increasing male body length 100.00 by the presence of six esophageal denticles, and 2 Decreasing male body length 100.00 the en face pattern of the elements of the external 3 Increasing female body length 100.00 corona. Traditionally two species (O. stephanos- 4 Decreasing female body length 100.00 5 Number of external elements 100.00 tomum and O. ventri) have been placed in the 6 Pattern of external elements 100.00 subgenus Ihlea, and one of the major diagnostic en face view features of this subgenus has been the presence 7 Pattern of funnel sclerotization 100.00 of 30-40 external elements. The O. pachyceph- 8 Shape of external elements 100.00 9 Number of esophageal teeth 100.00 alum specimens examined in this study also had 10 Length of cephalic distension 100.00 30-40 elements. This suggests that this species 11 Width of cephalic distension 100.00 belongs in the subgenus Ihlea and not Conow- 12 Position of cephaiic papillae 66.67 eberia. In order to clarify this point the type spec- 13 Length of esophagus 100.00 imens should be examined. Also, serial sections 14 Width of esophagus J 00.00 15 Length of external elements 100.00 and an en face view are necessary to establish 16 Buccal capsule ratio 100.00 the number of esophageal denticles and the pat- 17 Buccal capsule width 66.67 tern of the external elements. Because the num- 18 Length of esophageal funnel 100.00 ber of O. pachycephalum specimens available 19 Length of spicules 100.00 20 Length of tail 50.00 was limited it was not possible to get permission 21 Position of vulva 50.00 to section this species. 22 Presence of transverse processes 100.00 Although this analysis recognizes that Ihlea 23 Female body width 100.00 and Lerouxiella are monophyletic, we do not 24 Increasing male body width 100.00 recommend that these two subgenera be main- 25 Decreasing male body width 100.00

Figure 4. Phylogeny of Oesophagostomum (Conowebena). Each slash mark designates synapomorphic occurrence of trait listed beside it. Stars indicate homoplasious characters. own subgenus. Ten subgenera would be required from more recent host transfers, phylogenies of for the 12 species in the study group. This would the parasites and the hosts are required. considerably complicate the existing classifica- The phylogeny of the Oesophagostomum tion. By assigning all 12 species to the subgenus species studied in this analysis suggest a pattern Conowebena, the existing classification is dis- that can be explained predominantly in terms of rupted least. Thus, it is proposed that all those coevolution. Table 6 presents a list of the 12 Oesophagostomum species with a well-devel- parasite species analyzed and the hosts in which oped esophageal funnel with denticles (three or they are found. Of these 12, all but three are six), and small elements of the external corona present in primates. be assigned to the subgenus Conowebena. The primate relationships that require postulating the fewest host transfers for the Oesopha- Discussion gostomum species in the study group is shown The existence of a particular parasite in a par- in Figure 5. According to this host phylogeny ticular host may be explained in terms of either seven host transfers are necessary to explain the a historical coevolutionary association, or a more host records. These transfers are mapped onto recent host transfer. In the case of coevolution the parasite phylogeny (Fig. 6). The first of these the parasite or its close relatives are postulated host transfers involved the infection of primates to have evolved with the host lineage. In con- by the common ancestor of all Conowebena trast, hypotheses of host transfers are required species. Postulating this as a transfer is supported when a parasite is reported from a host and there by the occurrence of the outgruop subgenera Bo- appears no congruence between the parasite phy- sicola and Oesophagostomum in ruminants and logeny and the host phylogeny. In order to dis- suids respectively, rather than New World mon- tinguish between those associations that have re- keys, lemurs, or insectivores. The four most ple- sulted from host/parasite coevolution, and those siomorphic parasite species, O. zukowskyi, O.

Table 6. Host records and distribution of species in the study group.

Parasite species Host genera Distribution

O. zukowskyi Papio Africa O. bifurcum Cercopithecus, Pa- Africa, pio, Macacus, S.E. Asia Cercocebus, Cy- nomolgus Homo, Pan O. brumpti Macacus, Homo, Africa Pongo O. aculeatum Macacus, Cebus, S.E. Asia Cynomolgus O. susannae Pedetes S. Africa O. xeri Xerus S. Africa O. blanchardi Hylobates, Pongo S.E. Asia O. raillieti Hytobates S.E. Asia Figure 5. Primate phylogeny that best fits Oesoph- O. ovatum Hylobates S.E. Asia agostomum (Conoweberid) phylogeny. This phylogeny O. pachycephalum Cercopithecus Africa is consistent with the one presented by Schwartz et al. O. stephanostomum Homo, Pan, Gorilla Africa, (1978). S. America O. ventri "Cat" S. America cat (the scientific name of the wild cat was not given in the original description) in Brazil (Thornton, 1924). If this species is valid its re- bifurcum, O. brumpti, and O. aculeatum, are all lationship with cats is clearly the result of a host reported from Old World monkeys (family Cer- transfer. Of the 12 parasite species studied in this copithecidae), and these associations are consis- analysis, eight are present in a host for which a tent with a coevolutionary hypothesis. Two of coevolutionary hypothesis is sufficient. The only these four parasite species have secondarily parasites for which hypotheses of host transfers transferred to additional primate hosts. O. bi- are necessary are O. xeri and O. susannae from furcum has infected humans and chimpanzees, rodents, O. pachycephalum from monkeys, and and O. brumpti has infected humans. Two par- O. ventri from cats. Of those species in primates asite species in the study group, O. xeri and O. only O. pachycephalum does not fit into a Co- susannae, are reported from South African ro- evolutionary framework. Furthermore, of the dents (ground-squirrels and springhares, respec- eight species that appear to have coevolved with tively); and their presence in these hosts is clearly primates, three have secondarily transferred to the result of a host transfer. A second transfer is additional primate hosts. These include the in- likely to have occurred from ground-squirrels to fection of humans and chimpanzees by O. bi- springhares, or vice-versa, but because it has no furcum, humans by O. brumpti, and orangutans bearing on the relationships with primates it has by O. blanchardi. not been included in Figure 6. According to Le The geographic distributions of the parasite Roux (1940) these two hosts live in close prox- species studied in this analysis are listed in Table imity, thus facilitating this host transfer. The three 6. Most of these distributions are listed as general Oesophagostomum species from gibbons are also distribution, for example, Africa and Southeast consistent with a coevolutionary hypothesis. One Asia. One of the reasons for this is that for some parasite species from gibbons {O. blanchardi) has primates, the parasites are removed in zoos and also been reported from orangutans. Because O. research settings that may be far from the original blanchardi is the most plesiomorphic of the host habitat. Consequently, the exact locations species in gibbons, its occurrence in orangutans of primates are not always specified. Also, the is postulated to be the result of a host transfer. collections from primates are uneven and incom- The presence of O. pachycephalum in monkeys plete, and accurate geographic distributions can is also postulated to be the result of a host trans- only be speculative. fer. Oesophagostomum stephanostomum has been One of the most important steps in discussing reported from humans, chimpanzees, and goril- the biogeography of a group of organisms is the las. Finally, O. ventri was described from a wild evolutionary history or phylogeny of that group

SOUTH EAST ASIA

Old World onkeys

Figure 6. Phylogeny of Oesophagostomum (Conowebend) with host and geographic changes.

of organisms. The biogeographic scenario de- jor geographic shifts affecting the Oesophagos- veloped in this thesis is dependent on the parasite tomum species in primates, a dispersal from phylogeny proposed herein. In Figure 6 the min- Africa to Southeast Asia followed by a return imum number of major geographic changes are from Southeast Asia to Africa. Finally, there has mapped onto the parasite cladogram. The out- been a dispersal event to South America (Brazil) groups used in this analysis were not useful in by the ancestor of O. stephanostomum and O. establishing the plesiomorphic geographic range ventri. Because O. stephanostomum is present in of the Oesophagostomum species in primates, humans, chimpanzees, and gorillas (and the lat- because both the subgenera Bosicola and ter two hosts do not appear to disperse very much) Oesophagostomum are ubiquitous. However, the it is very likely that humans have transmitted occurrence of O. zukowskyi, O. bifurcum, and O. this parasite to South America. The indigenous brumpti in Africa suggest that Africa is the ple- South Americans are hypothesized to have come siomorphic range for Oesophagostomum species from Asia, so it is possible that these species were in primates. This suggests that O. bifurcum has transmitted with the African slave trade. secondarily dispersed to Southeast Asia. The existence of O. aculeatum, O. blanchardi, O. rail- Summary lieti, and O. ovatum in Southeast Asian primates A well-corroborated phylogenetic tree of and O. xeri and O. susannae in South African members of the genus Oesophagostomum inhab- rodents suggest either of the following two sce- iting primates supports the hypothesis that the narios, each of which require hypothesizing two parasite species form a monophyletic group, the geographic changes. In the first scenario, the subgenus Conoweberia. The parasite species ex- common ancestor of O. aculeatum dispersed to hibit host relationships that are predominantly Southeast Asia, and those species in rodents have coevolutionary, with relatively few cases of host dispersed back to Africa. In the second scenario switching. The geographic history implied by the there have been two dispersal events to Southeast phylogenetic tree suggests an African origin, with Asia, the first involving O. aculeatum, and the dispersal to Southeast Asia, followed by re-in- second the common ancestor of O. blanchardi vasion of Africa. and the rest of the subgenus. In terms of the number of geographic changes both these scenar- Literature Cited ios are equally parsimonious. The occurrence of Brooks, D. R., J. N. Caira, T. R. Platt, and M. H. O. pachycephalum and O. stephanostomum in Pritchard. 1984. Principles and Methods of Phy- logenetic Systematics. Special Publ. No. 12, Mus. Africa suggest that there has been a dispersal Nat. Hist. Univ. Kansas. 92 pp. event returning these parasites to the plesio- , R. T. O'Grady, and D. R. Glen. 1985. Phy- morphic range. Thus, there appear to be two ma- logeny of the Cercomeria Brooks, 1982 (Platy-

helminthes). Proc. Helminthol. Soc. Wash. 52: Lundberg, J. G. 1972. Wagner networks and ances- 1-20. tors. Syst. Zool. 21:398-413. Chaubaud, A. 1965. Ordre des Strongylida. Pages Popova, T. I. 1955. Strongyloids of animals and man. 869-933 in P. P. Grasse, ed. Traite de Zoologie. Strongylidae. In Essentials of Nematodology. Vol. Vol. 4. Masson et Cie, Paris. 5. Akad. Nauk SSSR. Moscow. [Translated En- , and M. C. Durette-Desset. 1973. Description glish version.] 414 pp. d'un nouveau nematode oesophagostome, para- Schwartz, J. H., I. Tattersall, and N. Eldredge. 1978. site d'Hyemoschis au Gabon, et remarques sur le Phylogeny and classification of primates revisited. genre Oesophagostomum. Bull. Mus. Natn. Hist. Yearbook Phys. Anthro. 21:95-133. Nat., Paris, 3e ser. No. 184, Zoologie 123:1415- Thornton, H. 1924. A review of the oesophagostomes 1424. in the collections of the Liverpool School of Trop- Farris, J. S. 1970. Methods for computing Wagner ical Medicine. Ann. Trop. Med. Parasit. 18:393- trees. Syst. Zool. 19:83-92. 408. . 1974. Formal definitions of paraphyly and Travassos, L., and E. Vogelsang. 1932. Pesquizas polyphyly. Syst. Zool. 23:548-554. helminthologicas realisadas em Hamburgo. Con- Hennig, W. 1966. Phylogenetic Systematics. Univ. tribuicoa ao conhecimento das especies de Oe- Illinois Press, Urbana. 263 pp. sophagostomum des primatas. Mem. Inst. Oswal- Le Roux, P. L. 1940. On the division of the genus doCruzT. 26:251-327. Oesophagostomum Molin, 1861, into subgenera Watrous, L. E., and Q. D. Wheeler. 1981. The out- and the creation of a new genus for the oesopha- group comparison method of character analysis. gostomes of the wart-hog. J. Helminthol. 17:1-20. Syst. Zool. 30:1-11. Lichtenfels, J. R. 1980. Keys to genera of the Su- Wiley, E. O. 1981. Phylogenetics: The Theory and perfamily Strongyloidea. In R.C. Anderson, A. G. Practice of Phylogenetic Systematics. John Wiley Chabaud, and S. Willmott, eds. CIH Keys to the and Sons, New York. 439 pp. Nematode Parasites of Vertebrates. No. 7. Farn- Yamaguti, S. 1961. Systema Helminthum. The ham Royal, Bucks, England. Commonwealth Ag- Nematode Parasites of Vertebrates. Vol. 3. Inter- ricultural Bureaux. 41 pp. science Publishers, Inc., New York. 1,261 pp.

REQUEST FOR OUT-OF-PRINT BACK ISSUES

Your Society is very short on the following issues of the Proceedings: 27(1); 28(2); 30(1); 37(1); and the 1975 special issue on Parasites of Equines. If you have these issues and would like to donate them, please contact the Corresponding Secretary/Treasurer (M. D. Ruff, API, BARC-East, Belts- ville, MD 20705) or send the back issues to Helminthological Society of Washington, % Allen Press, Inc., 1041 New Hampshire St., Lawrence, KS 66044. We will be glad to reimburse you for the shipping. You can claim a tax deduction for their value and at the same time help your Society meet its orders for complete back issues.