Introduction
The common wallaroo or euro, Osphranter robustus (Gould, 1840) (formerly known as Macropus robustus Gould, 1840) is one of the most widely distributed large kangaroos in Australia and is common in all states and the Northern Territory apart from Victoria and Tasmania (Clancy & Croft, Reference Clancy, Croft, van Dyck and Strachan2008). In Victoria, there appears to be a single extant population in east Gippsland close to the border with New South Wales (Menkhorst, Reference Menkhorst and Menkorst1996), while it is completely absent from Tasmania. In spite of its abundance and distribution, only a single survey of its helminth parasites has been published, based on an examination of 30 wallaroos restricted to central and northern Queensland (Beveridge et al., Reference Beveridge, Chilton, Johnson, Smales, Speare and Spratt1998). Other parasite records from this kangaroo species are based on incidental collection, but O. robustus appears to harbour more species of helminth parasites than any other of the large kangaroos belonging to the genera Macropus Shaw, 1790 and Osphranter Gould, 1842 (Spratt & Beveridge, Reference Spratt and Beveridge2016).
The data presented here, based on the examination of a total of 170 wallaroos (or euros) from across the continent, are intended to contribute to this apparent void in knowledge of the helminth parasites of one of the commonest and most abundant kangaroo species in Australia. An attempt has also been made to determine the extent to which helminth species may be restricted to any of the three major mainland subspecies of O. robustus currently recognized, namely the eastern wallaroo, O. r. robustus (Gould, 1840) occurring along the Great Dividing Range of eastern Australia, the northern wallaroo, O. r. woodwardi (Thomas, 1901) restricted to the Kimberley region of Western Australia and the Top End of the Northern Territory, and the euro, O. r. erubescens (Sclater, 1870) occurring across much of the inland areas of Australia (Clancy & Croft, Reference Clancy, Croft, van Dyck and Strachan2008) (fig. 1). In addition, the extent to which wallaroos and euros share their helminth parasites with other sympatric macropodids was investigated.
Fig. 1. Sites at which specimens of Osphranter robustus were collected for examination of their parasites, together with the presumed distribution of the subspecies of O. robustus based on van Dyck & Strahan (Reference Dyck van and Strahan2008).
Materials and methods
Nematodes were collected opportunistically between 1975 and 2018 from fresh road-killed animals, from wallaroos killed by commercial shooters or, in rare cases, wallaroos killed by aboriginal hunters (table 1). A small number of samples were also obtained from animals collected during other unrelated studies (Banks et al., Reference Banks, Copeman and Skerratt2006). The oesophagus, stomach, bile ducts and small and large intestines were examined for helminths. Examination varied depending upon facilities available at collection sites and not all organs were examined in all kangaroos collected either due to trauma in the case of road-killed specimens or inadequate examination of the pylorus and small intestine due to the lack of facilities available at the time of collection.
Table 1. Localities and numbers of Osphranter robustus examined for gastrointestinal parasites (arranged by state and increasing latitude).
Various methods of collection were involved, including the preservation of entire gastrointestinal tracts in formalin or the collection of parasites from fresh individual organs and their preservation in either formalin, 70% ethanol or Berland's fluid (Gibson, Reference Gibson1979) for gastric nematodes. Cestodes were relaxed in water before being preserved in 10% formalin or 70% ethanol. As very large numbers of nematodes were invariably present in the stomach, a series of 50–200 nematodes selected at random from the stomach contents were examined. The number of gastric nematodes that needed to be examined in order to reveal all species was investigated by Vendl & Beveridge (Reference Vendl and Beveridge2014) for other macropodid host species (50–200) and has been extrapolated to O. robustus without explicit validation. Only presence–absence data were recorded. None of the methods used were suitable for the identification of infections with Strongyloides sp. Nematodes were cleared in lactophenol for morphological identification, while cestodes were stained in Celestine blue, dehydrated in ethanol, cleared in methyl salicylate and mounted in Canada balsam.
Representative specimens of each species have been deposited in the South Australian Museum, Adelaide (table 2). In the case of parasite species complexes involving cryptic species (the nematodes Hypodontus macropi (see Chilton et al., Reference Chilton, Jabbar, Huby-Chilton, Jex, Gasser and Beveridge2012) and Macroponema comani (see Tan et al., Reference Tan, Chilton, Huby-Chilton, Jex, Gasser and Beveridge2012)) and the cestode Progamotenia festiva (see Beveridge & Shamsi, Reference Beveridge and Shamsi2009), reference is made only to the genotype(s) present in O. robustus. The cestodes Progamotaenia macropodis and Triplotaenia undosa also constitute complexes with apparently host-specific genetic forms (Hu et al., Reference Hu, Gasser, Chilton and Beveridge2005; Hardman et al., Reference Hardman, Haukisalmi and Beveridge2012) but have not been investigated morphologically. In the case of Macropostrongyloides baylisi, allozyme data suggest that the parasite present in O. robustus is genetically distinctive (Beveridge et al., Reference Beveridge, Chilton and Andrews1993).
Table 2. Gastrointestinal helminth parasites of Osphranter robustus.
St, stomach; Si, small intestine; Li, large intestine; Bd, bile ducts; Pa, pyloric antrum; SP, host-specific; SH, shared; U, widely distributed; N, northern; Orw, in O. r. woodwardi; Ore, in O. r. erubescens; Orr, in O. r. robustus; E, eastern; SAM, South Australian Museum.
Host abbreviations: Lc, Lagorchestes conspicillatus Gould; Mf, Macropus fuliginosus (Desmarest); Mg, M. giganteus Shaw; Na, Notamacropus agilis (Gould); Np, N. parryi (Bennett); Nr, N. rufogriseus (Desmarest); Onu, Onychogalea unguifera (Gould); Oa, Osphranter antilopinus (Gould); Ob, O. Bernardus (Rothschild); Or, O. rufus (Desmarest); P, Petrogale spp.; W, Wallabia bicolor (Desmarest).
a Prevalences in O. r. robustus, O. r. woodwardi, O. r. erubescens.
b Indicates genotype specific to O. robustus.
c Cited in Beveridge et al. (Reference Beveridge, Spratt, Close, Barker and Sharman1989); no voucher located.
d Indicates new host record.
Prevalence data are presented using the definition of Bush et al. (Reference Bush, Lafferty, Lotz and Shostak1997) as the number of infected hosts divided by the number of hosts examined. Diversity was assessed using the reciprocal of Simpson's Index for the species as a whole and for each host subspecies of O. robustus separately (Magurran, Reference Magurran1988). The similarity between helminth communities in each of the host subspecies was assessed using Sorenson's Index (Magurran, Reference Magurran1988). The bootstrap method of Poulin (Reference Poulin1998) estimates the extent to which all parasite species of a given community had been recovered based on prevalence and sample size.
The prevalence data for each organ examined (stomach, small intestine, large intestine, etc.) are based on the number of organs examined rather than the total number of wallaroos collected. In the case of the strongylid genera Filarinema Mönnig, 1929 and Alocostoma Mawson, 1979 specific identification of females was not possible based on morphological features. For these genera, the occurrence of the species identifiable based on males only has been included. The data presented here include those of Beveridge et al. (Reference Beveridge, Chilton, Johnson, Smales, Speare and Spratt1998) from northern and central Queensland.
Data on the host specificity of the parasites encountered were derived primarily from Spratt & Beveridge (Reference Spratt and Beveridge2016), with additional information in Shuttleworth et al. (Reference Shuttleworth, Jabbar and Beveridge2016a, Reference Shuttleworth, Beveridge, Koehler, Gasser and Jabbarb) and Beveridge et al. (Reference Beveridge, Jex, Tan and Jabbar2018). Each parasite species was categorized as host-specific (‘SP’ in table 2) if it or a specific genotype of it (or two genotypes in the case of H. macropi) was known exclusively from O. robustus, or as a shared parasite (‘SH’ in table 2) if it occurred in additional hosts. In the latter instance, it was often possible to identify a primary host from published prevalence data (Beveridge et al., Reference Beveridge, Spratt, Close, Barker and Sharman1989, Reference Beveridge, Speare, Johnson and Spratt1992, Reference Beveridge, Chilton, Johnson, Smales, Speare and Spratt1998; Beveridge, Reference Beveridge1998, Reference Beveridge2002, Reference Beveridge2016; Aussavy et al., Reference Aussavy, Bernardin, Corrigan, Hufschmid and Beveridge2011; Spratt et al., Reference Spratt, Walter and Haycock2017), with the parasite occurring at a much higher prevalence in the shared host than in O. robustus. Where identifiable, the principal hosts have been indicated in bold in table 2. In some instances, parasites appeared to be generalists from their host ranges and prevalences, and in these instances no principal host is indicated. The rock wallaby genus Petrogale Gray, 1837 contains 17 species and they have been treated only at the generic level. Host genera or species which are not sympatric with O. robustus (e.g. Dendrolagus Müller, 1840, Thylogale (Gray, 1837), Notamacropus eugenii (Desmarest, 1817)) as well as records from captive animals only (e.g. Lagorchestes hirsutus Gould, 1844) were excluded.
The geographical distribution of the species occurring at a prevalence higher than 20% and those occurring primarily in O. robustus was plotted on a map. In instances where several wallaroos were examined at localities around a principal town (e.g. Charters Towers, Queensland; Karratha, Western Australia), the data have been combined. Parasite species were categorized as demonstrating either a relatively uniform distribution across the continent (‘U’ in table 2), those restricted to O. r. woodwardi (northern part of the Northern Territory and the Kimberley region of Western Australia; ‘Orw’ in table 2), those restricted to the northern regions of the continent (northern Western Australia, the Northern Territory, northern Queensland; ‘N’ in table 2), those restricted to the eastern margin of the country (‘E’ in table 2) and species with a highly localized distribution. Distribution maps have already been published for Cloacina atthis, C. clymene, C. crassicaudata, H. macropi, Macroponema beveridgei, Mac. cf. comani and P. festiva (Beveridge & Shamsi, Reference Beveridge and Shamsi2009; Chilton et al., Reference Chilton, Jabbar, Huby-Chilton, Jex, Gasser and Beveridge2012; Beveridge et al., Reference Beveridge, Ngyuen, Nyein, Cheng, Koehler, Shuttleworth, Gasser and Jabbar2014, Reference Beveridge, Jex, Tan and Jabbar2018).
Host nomenclature follows Jackson & Groves (Reference Jackson and Groves2015). Authorities for parasite names are provided in table 2 and are not repeated in the text.
Results
The current study examined the gastrointestinal parasites of 170 O. robustus from all mainland states in which the species occurs (apart from Victoria) as well as from the Northern Territory (fig. 1 and table 1) consisting of 87 O. r. robustus, 18 O. r. woodwardi and 65 O. r. erubescens. The sample included 85 males and 30 females; the sex of the remaining animals (usually collected by commercial shooters) was not available. There was a significant bias in the sex ratio towards males among animals for which the sex had been determined (χ 2 = 12.6, P < 0.005). All animals examined were adults, although it was not possible to determine ages. Of the various organs examined, stomach nematodes were examined from all 170 animals, the pylorus was examined for Filarinema spp. in 152 animals, the small intestine in 116, the large intestine in 150 and the bile ducts in 137.
In total, 65 species of gastrointestinal parasites were encountered, including 61 species of strongylid nematodes, two occurring in the large intestine (Phascolostrongylinae: Hypodontus Mönnig, 1929; Macropostrongyloides Yamaguti, 1961), 55 in the tubular fore-stomach (Cloacininae) and four (Dromaeostrongylidae: Filarinema) in the pyloric antrum, together with four species of anoplocephalid cestodes, one occurring in the bile ducts (Progamotaenia Nybelin) and three in the small intestine (Progamotaenia, Triplotaenia Boas, Wallabicestus Schmidt, 1986) (table 2). Allocation of individual hosts to subspecies was based on their geographical distribution (van Dyck & Strahan, Reference Dyck van and Strahan2008) as well as on allozyme and DNA sequence data relating to distribution (Richardson & Sharman, Reference Richardson and Sharman1976; Eldridge et al., Reference Eldridge, Potter, Johnson and Ritchie2014; Richardson, Reference Richardson2019). Among the mainland subspecies of O. robustus, 52 helminth species were found in O. r. robustus, 30 species in O. r. woodwardi and 35 species in O. r. erubescens.
Using the bootstrap method of Poulin (Reference Poulin1998), it was estimated that an additional 4.6 species are likely to be found in O. robustus with more extensive sampling. The measure of diversity used, the reciprocal of Simpson's Index, was 24.6 for the host species, 20.7 for O. r. robustus, 19.9 for O. r. woodwardi and 19.3 for O. r. erubescens. Using Sorenson's Index to assess the similarity of the helminth communities in each of the subspecies, the similarity between the community in O. r. robustus and O. r. woodwardi was 53.7%, between O. r. woodwardi and O. r. erubescens 55.4%, and between O. r. robustus and O. r. erubescens 66.7%.
The frequency distribution of prevalences (fig. 2) showed an almost exponential decline from a large number of parasite species occurring at a low prevalence to a small number occurring at a high prevalence.
Fig. 2. Numbers of species of parasites of Osphranter robustus in each 10% prevalence class. ‘S’ indicates the number of host-specific species.
Of the 65 species of parasites present, 17 were specific to O. robustus. Others were shared with sympatric macropodid host species: the antilopine wallaroo, Osphranter antilopinus (21 species), the black wallaroo, O. bernardus (four species), the red kangaroo, O. rufus (16 species), the eastern grey kangaroo, Macropus giganteus (11 species), the western grey kangaroo, M. fuliginosus (11 species), the agile wallaby, Notamacropus agilis (ten species), the whiptail wallaby, N. parryi (13 species), the black-stripe wallaby, N. dorsalis (five species), the red-necked wallaby, N. rufogriseus (six species), rock wallabies, Petrogale spp. (28 species), the spectacled hare wallaby, Lagorchestes conspicillatus (six species), the northern nail-tail wallaby, Onychogalea unguifera (four species) and the swamp wallaby, Wallabia bicolor (eight species).
In instances where shared parasites occurred at a high prevalence in an alternative host (indicated in bold in table 2), the shared parasites consisted of the following number of species: O. antilopinus (17 species), O. bernardus (two species), O. rufus (12 species), M. fuliginosus (four species), M. giganteus (eight species), N. agilis (two species), N. dorsalis (one species), N. parryi (one species), L. conspicillatus (one species) and Petrogale spp. (one species).
Examining the geographical distributions of parasite species occurring at a high prevalence or those that were considered to be primarily parasites of O. robustus, several patterns were evident. A series of species occurred across the continent in all three subspecies of O. robustus: Cloacina communis, C. curta, C. echidne, C. frequens, C. johnstoni, C. macropodis, C. parva, C. phaethon, M. baylisi, H. macropi, P. festiva and P. macropodis. The distribution of C. parva is shown as an example in fig. 3. A second series was present again in all three subspecies, but in this case restricted to northern Australia; these included: Cloacina crassicaudata, C. dindymene, C. dirce, C. eileithyia, C. ixion, Mac. beveridgei and Pharyngostrongylus patriciae. The distribution of C. dindymene is shown as an example in fig. 4. A third series of species were restricted to O. r. woodwardi: Cloacina longibursata, C. polyxena, C. spearei, C. tyro, Mac. cf. comani, Pharyngostrongylus papillatus and P. sharmani. The distributions of C. polyxena and C. tyro are shown as examples in fig. 5. Species with a distribution essentially occurring along the Great Dividing Range of eastern Australia and, therefore, primarily parasitic in O. r. robustus were: Cloacina clymene, C. typhon, Labiosimplex robustus, Macropostrongylus spearei and Wallabicestus sp. The distributions of C. clymene and Ma. spearei are shown as examples in fig. 6. Several species were found only in O. r. erubescens, these being Cloacina daveyi, C. epona and C. longelabiata (fig. 7). Species with apparently localized distributions were: C. atthis, restricted to the Pilbara region of Western Australia; Cloacina dis, restricted to the Charters Towers region of north Queensland; and Popovastrongylus pluteus, restricted to O. r. robustus in north-eastern New South Wales (fig. 8). Several species did not fit precisely into any of these categories. Cloacina feronia was not found in M. r. woodwardi but was present in the remaining subspecies, while Mac. beveridgei was absent from Western Australian wallaroos. Cloacina eileithyia was prevalent in O. r. woodwardi in the Northern Territory but was found in a single wallaroo from the Charters Towers region of north Queensland (table 2). Triplotaenia undosa and Labiostrongylus grandis, while occurring in both O. r. robustus and O. r. erubescens, were found only in northern Australia.
Fig. 3. Geographical distribution of Cloacina parva in Osphranter robustus, representing species with a broad distribution across the continent in all three subspecies of O. robustus.
Fig. 4. Geographical distribution of Cloacina dindymene in Osphranter robustus, representing species occurring in all three subspecies of O. robustus, but restricted to northern Australia.
Fig. 5. Geographical distributions of Cloacina tyro (circles) and C. polyxena (squares) in Osphranter robustus, representing species restricted to O. r. woodwardi.
Fig. 6. Geographical distributions of Cloacina clymene (circles) and Macropostrongylus spearei (triangles) in Osphranter robustus, representing parasite species restricted to the east coast of the continent and, therefore, primarily parasitic in O. r. robustus.
Fig. 7. Geographical distributions of Cloacina daveyi (circles) and C. epona (squares) in Osphranter robustus, representing species restricted to O. r. erubescens.
Fig. 8. Geographical distributions of Cloacina atthis (triangles), C. dis (squares) and Popovastrongylus pluteus (circle) in Osphranter robustus, representing parasite species with highly localized geographical distributions.
Discussion
This is the first study of the gastrointestinal helminth parasites of O. robustus across its entire geographical range, the only previous study being an examination of 30 wallaroos from central and northern Queensland (Beveridge et al., Reference Beveridge, Chilton, Johnson, Smales, Speare and Spratt1998). Consequently, the present study greatly expands the numbers and ranges of parasites encountered, encompassing for the first time the three mainland subspecies of the common wallaroo. Although the collecting of specimens was entirely opportunistic and collections were significantly biased towards northern Queensland, as well as containing an obvious sex bias, they provide novel data on the structure of the helminth community of O. robustus as well as a basis for future, more detailed studies.
Three mainland subspecies of the common wallaroo are recognized (Jackson & Groves, Reference Jackson and Groves2015) and, in this study, specimens from the northern region of the Northern Territory and the Kimberley region of Western Australia were readily referable to the northern wallaroo, O. r. woodwardi, based on their distribution (Clancy & Croft, Reference Clancy, Croft, van Dyck and Strachan2008). Likewise, wallaroos from the remaining regions of Western Australia, from central Australia, South Australia, western New South Wales and western Queensland were readily attributable to the euro, O. r. erubescens, again based on geographical distribution (Clancy & Croft, Reference Clancy, Croft, van Dyck and Strachan2008). Wallaroos from south-eastern Queensland and north-eastern New South Wales were attributable to O. r. robustus based on both distribution and coat colour (Clancy & Croft, Reference Clancy, Croft, van Dyck and Strachan2008), while animals from the Dividing Range in northern Queensland could potentially have been either O. r. robustus or O. r. erubescens, as the precise boundary between the two subspecies in this area has not been delineated (Clancy & Croft, Reference Clancy, Croft, van Dyck and Strachan2008). These two subspecies are differentiable based on biochemical grounds (Richardson & Sharman, Reference Richardson and Sharman1976; Richardson, Reference Richardson2019) and there are limited DNA sequence data for distinguishing between them (Eldridge et al., Reference Eldridge, Potter, Johnson and Ritchie2014). Based on the distribution of wallaroos shown in Fig. 1 of Eldridge et al. (Reference Eldridge, Potter, Johnson and Ritchie2014) and identified using DNA sequence data, as well as data in Richardson (Reference Richardson2019) based on allozyme data, it is highly likely that all of the Queensland specimens collected from the Great Dividing Range in this study belong to O. r. robustus (fig. 1).
There are several potential biases inherent in this study. First of all, the collecting of specimens was opportunistic rather than structured. There is a clear bias towards male hosts, and the possible effects of host sex on parasites likely to be present in this species are not known. Second, the identification of nematodes from the stomach based on the examination of a relatively small sample of the (potentially) thousands present is likely to overlook the presence of species occurring at low intensities or abundances (Vendl & Beveridge, Reference Vendl and Beveridge2014) and, therefore, the data presented are likely to favour the more abundant species and to potentially overlook rarer species. Thirdly, a survey of this type took no account of seasonal variation in parasite burdens. Previous studies have shown a highly seasonal pattern of parasite infection in M. giganteus in the winter rainfall areas of southern Australia (Arundel et al., Reference Arundel, Dempster, Harrigan and Black1990), while similar studies of the parasites of O. rufus in the arid zone of Australia with non-seasonal rainfall found limited seasonal variation in parasite prevalence and abundance (Arundel et al., Reference Arundel, Beveridge and Presidente1979). One exception in the latter study was Labiosimplex longispicularis, which has a highly seasonal annual life cycle (Mykytowycz & Dudzinski, Reference Mykytowycz and Dudzinski1965). Comparable data are lacking for congeners occurring in O. robustus, and, therefore, some caution is needed in interpreting the current results in the face of a complete lack of seasonal data and the fact that O. robustus occurs in arid, non-seasonal areas of rainfall as well as summer-dominant rainfall areas (northern Australia) and temperate non-seasonal rainfall areas (south-eastern Australia).
In the current study, a total of 65 species of gastrointestinal parasites were encountered, compared with the 40 reported by Beveridge et al. (Reference Beveridge, Chilton, Johnson, Smales, Speare and Spratt1998) from Queensland. This difference is not surprising given the expanded geographical region included in the present study. More helminth species (52) were found in O. r. robustus, compared with O. r. woodwardi (30) and O. r. erubescens (35), but this may reflect sampling intensity rather than true differences between the subspecies. The difference in numbers of helminth species encountered was mirrored in the reciprocal of Simpson's Index for each of the host subspecies. The reciprocal of Simpson's Index of diversity for the entire species (O. robustus) (24.6) was similar to that (23.0) reported from Queensland by Beveridge et al. (Reference Beveridge, Chilton, Johnson, Smales, Speare and Spratt1998). The slight difference may be due to the addition of a significant number of parasite species (25) occurring at low prevalences as Simpson's Index is biased towards prevalent species (Magurran, Reference Magurran1988).
The bootstrap method of Poulin (Reference Poulin1998) used to estimate the number of species yet to be described from O. robustus suggested that an additional 4.6 species might be found, indicating that the majority of helminth species likely to occur in this host have already been encountered. Based on the data presented here, these additional and as yet unknown species are likely to be host-specific species with an extremely limited geographical range or unusual colonizations from sympatric macropodid hosts.
Sorenson's Index of similarity suggested that the helminth community of O. r. woodwardi differed from communities in the remaining host subspecies, a finding considered in more detail below but probably due to a number of helminth species occurring exclusively in this host subspecies.
Of the helminth species encountered, only 17 were specific to O. robustus, whilst the remainder of the parasites were shared with sympatric host species. Being able to differentiate shared parasites on the basis of their prevalences compared with those in O. robustus sheds some additional light on the direction of host sharing. For example, in the case of Petrogale spp., 23 parasite species were potentially shared with O. robustus, yet in all cases except one (W. obendorfi) the parasites occurred at very low prevalences in rock wallabies (Beveridge et al., Reference Beveridge, Spratt, Close, Barker and Sharman1989) compared with prevalences in O. robustus, suggesting the occasional transfer of most species from O. robustus to Petrogale spp. By contrast, in species of Notamacropus, prevalences of shared parasites were generally much higher than those in O. robustus (Beveridge et al., Reference Beveridge, Chilton, Johnson, Smales, Speare and Spratt1998), suggesting, by contrast, the occasional transfer of parasites to O. robustus from Notamacropus spp. In some instances, high prevalences of the shared parasite also occurred in the shared host, with 21 species of nematodes in O. antilopinus, 12 in O. rufus, four in M. fuliginosus and eight in M. giganteus. Smaller numbers of parasite species were shared with sympatric members of the genera Lagorchestes and Onychogalea. The known life cycles of all of the nematode parasites are terrestrial, with eggs being deposited in faeces, larval development occurring in the external environment and third-stage larvae being ingested by the host while grazing (Smales, Reference Smales1977a, Reference Smalesb; Beveridge & Presidente, Reference Beveridge and Presidente1978; Beveridge, Reference Beveridge1979). In the case of the anoplocephalid cestodes, there is a requirement for an intermediate host (most probably an oribatid mite) (Beveridge, Reference Beveridge, Khalil, Jones and Bray1994), but, again, these are ingested incidentally during grazing. Consequently, there are substantial opportunities for the transfer of parasites between co-grazing macropodid hosts. The extent to which macropodid grazing preferences may influence this transmission has been examined for four sympatric macropodid species in Victoria (Aussavy et al., Reference Aussavy, Bernardin, Corrigan, Hufschmid and Beveridge2011) and three species in southern New South Wales (Spratt et al., Reference Spratt, Walter and Haycock2017). Thus, in the present study, in the lowest prevalence class (0–10%), fewer than 30% of the parasite species found in O. robustus were host-specific, while in the higher prevalence classes (>40%), only one of five (20%) of the parasites were host-specific (fig. 2).
The high number of parasite species shared with O. antilopinus was not unexpected as these two hosts occur in sympatry across much of northern Australia (Eldridge et al., Reference Eldridge, Potter, Johnson and Ritchie2014) and are also closely related phylogenetically (Meredith et al., Reference Meredith, Westerman and Springer2008). Beveridge et al. (Reference Beveridge, Chilton, Johnson, Smales, Speare and Spratt1998) estimated that the two species shared more than 50% of their parasites in north Queensland. In other areas of Australia, wallaroos occur frequently in sympatry with O. rufus, M. giganteus and M. fuliginosus (van Dyck & Strahan, Reference Dyck van and Strahan2008) and, therefore, it is unsurprising that they share their parasites in these regions, in particular with the congeneric species O. rufus. Osphranter robustus also shares habitat with various species of Notamacropus, Lagorchestes and Petrogale (van Dyck & Strahan, Reference Dyck van and Strahan2008), thereby explaining the sharing of parasites with these host genera.
The data on numbers of species in each 10% prevalence class mirror that presented by Beveridge et al. (Reference Beveridge, Chilton, Johnson, Smales, Speare and Spratt1998) for O. r. robustus, although based on a much larger sample size. Similar patterns have been reported for additional species of macropodids from north Queensland (Beveridge et al., Reference Beveridge, Chilton, Johnson, Smales, Speare and Spratt1998) and the pattern appears to be a characteristic of the helminth communities of these marsupials with greatest number of species occurring in the low prevalence classes and fewer species in the high prevalence classes.
Recent molecular studies of various cloacinine genera of nematodes occurring in macropodids have suggested that the diversity found among the nematodes may be the result of host colonization (and possibly a limited degree of within-host speciation) rather than coevolution (Chilton et al., Reference Chilton, Huby-Chilton, Beveridge, Smales, Gasser and Andrews2011, Reference Chilton, Huby-Chilton, Koehler, Gasser and Beveridge2016a, Reference Chilton, Huby-Chilton, Koehler, Gasser and Beveridgeb, Reference Chilton, Shuttleworth, Huby-Chilton, Koehler, Jabbar, Gasser and Beveridge2017). The current data support this hypothesis to the extent that many of the parasites of O. robustus are shared with sympatric host species, a prerequisite for the colonization of a new host species and subsequent speciation. However, the details of this mechanism remain to be investigated.
Based on the currently available data, there are several patterns of geographical distribution of the parasites identified herein. A number of species (e.g. C. parva, C. phaethon, C. macropodis) (fig. 3) have a transcontinental distribution occurring in all three host subspecies, and while molecular studies of these nematode species indicate significant levels of genetic differentiation within each species, there is no clear geographical separation of the identified genotypes and, consequently, no obvious association with separate subspecies of O. robustus (Shuttleworth et al., Reference Shuttleworth, Beveridge, Koehler, Gasser and Jabbar2016b). A number of parasite species are restricted to O. r. woodwardi, although they may also occur in the sympatric host O. antilopinus, while another series of parasites is restricted to northern Australia, including northern Queensland, thereby being found in O. r. robustus, O. r. erubescens and O. r. woodwardi, again, with some also occurring in O. antilopinus. The Carpentaria Gap, a zone of floral and faunal disjunction south of the Gulf of Carpentaria has been invoked to explain the disjunct distribution of plant and animal species between northern Queensland and the Northern Territory, including genetically distinct populations of O. antilopinus (Wadley et al., Reference Wadley, Fordham, Thomson, Ritchie and Austin2016), and while this may facilitate an explanation of the parasite species restricted to O. r. woodwardi, it provides no explanation for the species found across northern Australia including Queensland other than a distribution restricted to the monsoonal tropical region of northern Australia. Additional caution is needed in assessing such current distributions of nematode parasites, as C. eileithyia is found commonly in O. r. woodwardi in the Northern Territory (Beveridge, Reference Beveridge1998), but was found in a single individual of O. r. robustus in northern Queensland, suggesting that more intensive sampling may be needed before definitive conclusions can be drawn about nematode distributions.
Apart from helminth species with a broad distribution and those with highly restricted distributions such as C. dis, C. atthis and P. pluteus, groups of species were also identifiable restricted to each of the three subspecies of O. robustus. While a significant series of species was identified, which were restricted to O. r. woodwardi, a smaller number were restricted to O. r. erubescens, with several species occurring primarily within the range of O. r. robustus along the east coast of the continent, although the division between subspecies in this region is not clear, as indicated above. Taken together, the data suggest that the process of subspeciation occurring in the wallaroos may have influenced the diversification of their parasites in particular groups of species, as well as the possibility of climatic factors limiting the distribution of other species to the monsoonal tropical regions of northern Australia, irrespective of host subspecies. The observations presented here support the genetic analysis of H. macropi by Chilton et al. (Reference Chilton, Jabbar, Huby-Chilton, Jex, Gasser and Beveridge2012) in which each genotype of the parasite was associated with a particular subspecies of O. robustus. However, a more detailed genetic analysis of the host, O. robustus, is needed before such hypotheses can be explored further.
In spite of obvious deficiencies in knowledge of the parasites, their distributions and their genetic characteristics, and equally obvious deficiencies in knowledge of the phylogeography of the host O. robustus, as well as its close congener O. antilopinus, it appears that a more detailed study of this host–parasite system could be useful in understanding the evolution of the complex parasite communities in O. robustus and its closely related congeners.
Acknowledgements
Thanks are due to the many colleagues and friends who helped in the collection of the data presented here, including (the late) David Banks, Neil Chilton, Brian Coman, Gregg Curran, Andrew Doube, Peter Johnson, Shane Middleton, (the late) Rick Speare and David Spratt. Chris Beveridge is thanked for his assistance in calculating indices. David Spratt is also thanked for his comments on drafts of the manuscript.
Financial support
This work was supported by the Australian Research Council and the Australian Biological Resources Study (grant number 217-06).
Conflicts of interest
None.
Ethical standards
Specimens were collected under the following permits: Queensland National Parks and Wildlife Service T00436, T00759, T00943, T1131, WA0006125; New South Wales National Parks and Wildlife Service A681, A756; South Australian Department of Environment and Heritage E07358, G24351; Northern Territory Department of Primary Industry 15747; and the Western Australian Department of Environment and Conservation SF007407.
