The rediscovery of the putative ant social parasite Manica parasitica syn. nov. (Hymenoptera: Formicidae) reveals an unexpected endoparasite syndrome

1. Introduction

Parasitism is characterized by one organism exploiting another organism by feeding on it, showing adaptations to it and harming it [1]. Parasitism is ubiquitous across the tree of life, and parasites can drive biological diversification [2]. Approximately half of all animal species are parasites, and parasitism has evolved more than 200 times in animals [3,4]. Social insects provide resources for parasites and ideal conditions for parasite transmission because they occur abundantly in terrestrial ecosystems, and live in large, densely packed colonies which often persist for multiple years [5,6]. Accordingly, colonies of social insects are exploited by many pathogens, endo- and ectoparasites, as well as parasitoids [5].

Ant colonies in particular are inhabited by a menagerie of guests, the so-called myrmecophiles, which are distantly related invertebrates that steal resources or prey on the ants' brood [7]. By contrast, social parasitism is a phenomenon that is common between ant species. Here, a parasitic species exploits the social behaviour of another, often closely-related free-living species to survive and reproduce [8]. Of the approximately 14 000 recognized ant species [9], more than 400 social parasite species are known, which evolved at least 91 times independently across the ant tree of life [10,11]. Furthermore, ants are hosts of many endoparasites, including nematodes, cestodes, trematodes, fungi, bacteria and viruses [5,12,13].

Among these host–parasite pairs, strong selection acts on both partners, generating morphological, physiological and behavioural syndromes, which indicate host–parasite co-evolution [13,14]. One well-studied ant endoparasite syndrome is the mermithogenic syndrome, where mermithid nematodes develop for parts of their life cycles inside ant hosts [15]. Upon maturity, the nematode alters the host's behaviour, and the infected ant drowns itself, releasing the parasite [16]. Mermithid infections can cause modifications to the host's morphology, including mosaic phenotypes in females and intercastes [17]. Some morphological alterations are so extreme that parasitized individuals were erroneously described as new species, causing taxonomic confusion [15,18,19]. Another established syndrome in ants is the social parasite syndrome. Here, the morphology, behaviour, and reproductive patterns of ant social parasites are adapted to the parasitic life history. For example, queens of dulotic species kill the host queen during colony founding, taking over the host colony; the worker caste bears saber-shaped mandibles specialized for assaulting host colonies and stealing their pupae [20]. In contrast, inquiline parasites often coexist with the host queen and are characterized by reduced size, loss of the worker caste, reduced mouthparts and sib-mating inside the nest [14,21–24].

The endoparasite and social parasite syndromes present overlapping shifts in morphology and behaviour (table 1). Therefore, disentangling these syndromes is challenging when presented with morphology alone. Here, we examine the biology of the putative social parasite Manica parasitica (Creighton, 1934). Manica parasitica (figure 1a–c) was first collected in the nest of Manica bradleyi (Wheeler, 1909) (figure 1d–f) in Yosemite National Park in California [25]. Only two additional collections have been documented, both in the Sierra Nevada. Creighton classified M. parasitica as a social parasite of M. bradleyi due to its co-nesting behaviour and morphology, which overlaps with the social parasite syndrome, such as shiny integument, smaller workers, reduced propodeum and a lower petiole height compared to M. bradleyi [25]. Creighton [25] assumed temporary social parasitism because putative parasite and host workers occupy the same nest, and lack of mandibular specialization did not indicate dulosis. Later, Wheeler & Wheeler [26] noted M. parasitica ambulated differently than M. bradleyi, but their conclusion was based upon a single worker, and they could not rule out that it was injured [26]. We revisited two historical M. parasitica collecting localities and excavated nests containing M. bradleyi and M. parasitica. To determine the relationship between the two, we critically re-evaluate the status of M. parasitica as a symbiont and as a species using multiple lines of evidence.

• Download figure
• Open in new tab
• Download PowerPoint

2. Material and methods

For a full description of materials and methods, see electronic supplementary material, file S1.

We located two populations of M. parasitica in the Sierra Nevada (figure 1g). Nests were excavated via shovel; workers and brood were transferred to nest-boxes for observation. To document interactions between M. parasitica and M. bradleyi, we recorded videos with a Sony AX53 4K Handycam.

To test the hypothesis that M. parasitica is an independently evolving species, we gathered multiple specimens of M. parasitica and M. bradleyi from the two populations (figure 1g), also sampling from the larger range of M. bradleyi. We extracted DNA, prepared genomic libraries, and performed targeted enrichment of ultraconserved elements (UCEs) [27]. We processed raw reads with the PHYLUCE and SWSC-EN pipelines [28,29], incorporating sequences from previous datasets [27,30]. We used two analytical approaches: concatenate and partition via IQTREE v. 2.1.2 [31], and summary coalescent analysis via ASTRAL-III [32]. Collection data are in electronic supplementary material, table S1. Raw data are on the NCBI Sequence Read Archive (BioProject PRJNA1013280).

To re-evaluate the morphological observations made by Creighton [25], we took morphometric data from M. parasitica and M. bradleyi using a Leica M205C microscope equipped with a Leica DFC450 digital camera and the Leica Application Suite v.4.5 (electronic supplementary material, file S1, figure S1 and table S3). Our morphometric dataset included 82 specimens (M. parasitica, n = 40; M. bradleyi, n = 42) from both populations. Raw data and scripts are available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.9s4mw6mnz [33].

Because we observed no reproductives of M. parasitica nor M. bradleyi during excavations, we dissected workers of both species to determine reproductive status, following Dolezal & Brent [34]. To compare the internal anatomy of M. parasitica and M. bradleyi, we scanned three specimens each using synchrotron X-ray micro-computed tomography (microCT), processed the tomograms by segmenting every 10^th slice with Slicer v. 5.0.3 [35], and performed semi-automated segmentation with Biomedisa [36]. Original scans and segmentations also on the Dryad Digital Repository [33].

Metacestodes (larval cestodes) extracted from ants were fixed in 95% ethanol. Specimens were mounted on microscope slides in Berlese's medium [37]. They were examined and photographed using a Zeiss Axio Imager M2 light microscope.

3. Results

In this study, we located two populations of Manica parasitica, a putative social parasite of M. bradleyi, in the Sierra Nevada. To test if M. parasitica is an independently evolving species, we inferred a phylogeny for our target taxa (n = 34) using broad sampling (see electronic supplementary material table S4 for sequencing statistics). We recovered a statistically well-supported clade consisting of M. bradleyi and M. parasitica, which were extensively interdigitated (figure 2a; electronic supplementary material, file S3: figure S1). Additionally, we recovered population structure with individuals of M. parasitica and M. bradleyi generally clustering together by population.

• Download figure
• Open in new tab
• Download PowerPoint

To evaluate the diagnostic, species-specific characters proposed by Creighton [25], we collected morphometric data. We found that M. parasitica is significantly smaller than M. bradleyi (figure 2d: WL), and M. parasitica differs significantly from M. bradleyi in multiple measurements (electronic supplementary material, file S3: figure S2 for full results). However, when corrected for size, many of these differences became insignificant (figure 2d: PEHI, PrdHI). Instead, the postpetiole was significantly broader, and the antennal scapes were significantly longer in M. parasitica versus M. bradleyi, both of which have been proposed as morphological traits similarly affected by the social parasite and endoparasite syndromes. Additionally, hind femur length was significantly longer in M. parasitica versus M. bradleyi (figure 2d: HFLI and SI2). Significant differences in hind femur length and antennal scape length, regardless of size correction, suggest differential effects of parasite infection on the development of the head, thorax and abdomen versus extremities, which remain less affected in parasitized individuals.

We did not recover reproductives of either species from excavations. We dissected workers of both species to determine reproductive status. We found no evidence of reproductive activity in M. parasitica or M. bradleyi workers, but the gaster of every M. parasitica worker (n = 32) contained at least one, and up to 31 metacestodes. Workers of M. bradleyi (n = 33) contained none. From our microCT reconstructions, we conclude that the metacestodes are localized in the haemocoel, without any apparent disruption to major organ systems (figure 2c; electronic supplementary material, table S2), although the venom gland of M. parasitica appeared atrophied compared to M. bradleyi (electronic supplementary material, file S3: figure S3).

Examination of the metacestodes revealed a pattern of body organization corresponding to cysticercoids [38] with a fully developed scolex (identical to that of adults) retracted into a solid cyst (figure 2b, top). They possess a rostellum armed with 194–198 hooks arranged in two regular rows in a simple circle (figure 2b, bottom). Hooks are shaped specifically (hammer-shaped or T-shaped), with short blade, short handle and elongate guard. Hook length measured along the axis of the guard is 14–16 μm. Suckers are armed 14–16 diagonal rows of spines situated peripherally; spines 12–13 μm long (figure 2b, bottom). The shape and arrangement of the rostellar hooks, in combination with armed suckers, identify this species as a member of the genus Raillietina Fuhrmann, 1920 (Cestoda: Davaineidae).

Following morphometric analysis, we compared M. parasitica against the endo- and social parasite syndromes (table 1). Although some characters of the endoparasite syndrome remain unknown, M. parasitica overlaps with both syndromes, e.g. reduced body size, limited geographical distribution, smooth integument, elongated antennal scapes, altered integument colour and broadened postpetiole (figure 2d; electronic supplementary material, table S3). However, more overlap with the endoparasite syndrome is apparent: the worker caste is not lost, and pilosity, mandibular dentition and propodeum are not reduced. Manica parasitica also exhibits some characteristics that defy the endoparasite syndrome, i.e. the head is not ovular, ocelli are absent and the gaster is not swollen, despite carrying a sometimes-heavy load of parasites.

4. Discussion

Above, we tested the hypothesis that Manica parasitica is an independently evolving social parasite species by (i) extensively sampling two localities, (ii) inferring the phylogeny of M. parasitica and M. bradleyi using dense population sampling, (iii) evaluating morphological and life-history characters, and (iv) using the combined evidence to evaluate M. parasitica in the context of the social parasite and endoparasite syndromes.

Because M. parasitica has an unusual combination of characters that variously overlap with (or defy) the endo- and social parasite syndromes (table 1), we initially gave credulity to Creighton's hypothesis that M. parasitica is a social parasite. Therefore, we had initiated the phylogenetic analysis before we realized that M. parasitica was infected with cestodes. Our phylogeny is consistent with the result that M. parasitica represents cestode-infected M. bradleyi workers instead of an independently evolving species. Manica parasitica and M. bradleyi samples are interdigitated within a single clade, rendering each other polyphyletic (figure 2a). Previous studies have identified paraphyletic host groups, with social parasite species nested within the host clade. However, in those studies, parasite taxa were monophyletic, and host paraphyly was interpreted as a signature of incomplete lineage sorting following recent speciation [39–41]. By contrast, the interdigitated pattern of M. bradleyi and M. parasitica suggests close relatedness between individuals and the absence of genetic divergence.

The morphological alterations of M. parasitica are apparently caused by infection with metacestodes. Remarkably, we found that a single metacestode was sufficient to induce the full suite of morphological changes in an individual (9 of 40 dissections; electronic supplementary material, table 2). We have identified the cestode genus as Raillietina (sensu lato), a cosmopolitan genus consisting of approximately 290 species that uses insects of different orders (rarely, gastropods) as intermediate hosts and non-aquatic bird and mammal species as definitive hosts [42–44]. Four subgenera have been erected [42,43] based on two binary characters of the adult morphology (genital pores unilateral/alternating; uterine capsules with a single egg/multiple eggs), which are sometimes recognized as the full genera Raillietina (sensu sricto), Fuhrmanneta Stiles & Orleman, 1926, Paroniella Fuhrmann, 1920 and Skjabinia Fuhrmann, 1920 [44,45]. However, these are species assemblages formed to facilitate the identification of adult cestodes rather than monophyletic groups. Of the genus Raillietina (sensu lato), there are at least 21 species recorded from North America [43]; their definitive hosts are birds of the family Phasianidae, Odontophoridae, Picidae, Icteridae and Columbidae as well as small mammals (rodents, rabbits and hares). Exact species identification of the metacestode in this study may continue with comparative studies of the scolex armaments of the known North-American Raillietina spp.

We remain uncertain of the details of parasite transmission, but speculate that M. bradleyi collects bird faeces containing cestode eggs while foraging and feeds these to its larvae via trophallaxis, which has been observed in other myrmecine ants [46]. Development of infected individuals is likely disrupted by nutrient deficiency, potentially causing the morphological modifications observed in M. parasitica.

While we did not conduct a quantitative analysis of behaviour in this study, we confirmed Wheeler & Wheeler's [26] observation of M. parasitica's unusual gait via casual observations of multiple individuals, finding that infected workers move slowly and often fall when scaling small objects. In both populations, we observed M. parasitica using the same nest entrances as M. bradleyi (figure 1h). Workers of M. parasitica interacted normally with M. bradleyi nest-mates, accepting grooming from uninfected M. bradleyi workers (electronic supplementary material, file S2). Manica parasitica workers often shake as they rest or groom (electronic supplementary material, file S2). Endoparasites are known to manipulate the behaviour of their hosts, and some ‘adaptive manipulations’ are interpreted as extended phenotypes of the parasite, increasing parasite transmission [47]. The gait of M. parasitica may be an adaptive manipulation by the parasite, making workers easier prey for foraging birds. Alternatively, the gait of M. parasitica workers could be a pathological reaction or an adaptive host response [48]. The potential adaptive significance of this behaviour must be determined experimentally.

We remain uncertain of the significance, if any, of the reduced integument sculpture and altered coloration of parasitized individuals. Reduced sculpturing may be a product of thinner cuticle, which may be caused by the parasite reappropriating resources during larval development. Darker coloration of parasitized individuals has also been observed in Myrmica infected with davaineid cestodes [49]. Dark cuticle could be caused by an immune response to the parasite infection because melanization is one of many responses that arthropod immune systems use to combat infections [50,51]. If infection occurs before pupation, melanin may affect the adult tanning process, resulting in darker individuals.

Considering the combined evidence, we conclude that M. parasitica represents cestode-infected M. bradleyi workers rather than an interspecific social parasite of M. bradleyi. Accordingly, we propose the taxonomic synonymy of M. parasitica (Creighton, 1934) (syn. n.) under M. bradleyi (Wheeler, 1909). We would like to reiterate the findings of Csősz et al. [19] by emphasizing that, whenever possible, multiple lines of evidence should be used when describing new taxa. This especially applies to putative social parasite species given their rarity and the overlaps between the social parasite and endoparasite syndromes (table 1). While there are multiple hypotheses for the adaptive significance of traits listed in table 1 for the endo- and social parasite syndromes, to our knowledge few have been formally tested. Future studies are needed to shed light on the identity and life cycle of Raillietina cestodes infecting M. bradleyi, and to test the adaptive significance of the morphological and behavioural modifications exhibited by ants infected with endoparasites, as well as by ant social parasites.

Ethics

This work did not require ethical approval from a human subject or animal welfare committee.

Data accessibility

The sequence reads generated in this study are available from the NCBI Sequence Read Archive (BioProject PRJNA1013280). The data and scripts that support the findings of this study are available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.9s4mw6mnz [33].

Supplementary material is available online [52].

Declaration of AI use

We have used AI-assisted technologies in creating this article for processing microCT scans.

Authors' contributions

M.P.: conceptualization, data curation, formal analysis, investigation, methodology, project administration, software, validation, visualization, writing—original draft, writing—review and editing; B.B.G.: investigation, writing—original draft, writing—review and editing; T.v.d.K.: visualization, writing—original draft, writing—review and editing; E.H.: visualization, writing—original draft, writing—review and editing; I.B.: data curation, writing—review and editing; C.R.: conceptualization, data curation, funding acquisition, investigation, resources, supervision, writing—original draft, writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

This project was supported by the US National Science Foundation (NSF CAREER DEB-1943626), the Social Insect Research Group (SIRG) at Arizona State University, the University of Hohenheim and the Carl-Zeiss-Foundation.