Toxoplasma gondii is a major cause of congenital brain disease; however, the underlying mechanism of neuropathogenesis in brain toxoplasmosis remains elusive. To explore the role of T. gondii in the development of neural stem cells (NSCs), NSCs were isolated from GD14 embryos of ICR mice and were co-cultured with tachyzoites of T. gondii RH strain. We found that apoptosis levels of the NSCs co-cultured with 1×106 RH tachyzoites for 24 and 48 h significantly increased in a dose-dependent manner, as compared with the control. Western blotting analysis displayed that the protein level of C/EBP homologous protein (CHOP) was up-regulated, and caspase-12 and c-Jun N-terminal kinase (JNK) were activated in the NSCs co-cultured with the parasites. Pretreatment with endoplasmic reticulum stress (ERS) inhibitor (TUDCA) and caspase-12 inhibitor (Z-ATAD-FMK) inhibited the expression or activation of the key molecules involved in the ERS-mediated apoptotic pathway, and subsequently decreased the apoptosis levels of the NSCs induced by the T. gondii. The findings here highlight that T. gondii induced apoptosis of the NSCs through the ERS signal pathway via activation of CHOP, caspase-12 and JNK, which may constitute a potential molecular mechanism responsible for the cognitive disturbance in neurological disorders of T. gondii.

Malaria parasites often manage to maintain an infection for several months or years in their vertebrate hosts. In humans, rodents and birds, most of the fitness costs associated with malaria infections are in the short initial primary (high parasitaemia) phase of the infection, whereas the chronic phase (low parasitaemia) is more benign to the host. In wild birds, malaria parasites have mainly been studied during the chronic phase of the infection. This is because the initial primary phase of infection is short in duration and infected birds with severe disease symptoms tend to hide in sheltered places and are thus rarely caught and sampled. We therefore wanted to investigate the relationship between the parasitaemia during the primary and chronic phases of the infection using an experimental infection approach. We found a significant positive correlation between parasitaemia in the primary peak and the subsequent chronic phase of infection when we experimentally infected great reed warblers (Acrocephalus arundinaceus) with Plasmodium ashfordi. The reason for this association remains to be understood, but might arise from individual variation in exoerythrocytic parasite reservoirs in hosts, parasite antigenic diversity and/or host genetics. Our results suggest that the chronic phase parasitaemia can be used to qualitatively infer the parasitaemia of the preceding and more severe primary phase, which is a very important finding for studies of avian malaria in wild populations.

Trophically-transmitted parasites frequently alter multiple aspects of their host's phenotype. Correlations between modified characteristics may suggest how different traits are mechanistically related, but these potential relationships remain unexplored. We recorded 5 traits from individual isopods infected with an acanthocephalan (Acanthocephalus lucii): hiding, activity, substrate colour preference, body (pereon) coloration, and abdominal (pleon) coloration. Infected isopods hid less and had darker abdominal coloration than uninfected isopods. However, in 3 different experiments measuring hiding behaviour (time-scales of observation: 1 h, 8 h, 8 weeks), these two modified traits were not correlated, suggesting they may arise via independent mechanisms. For the shorter experiments (1 h and 8 h), confidence in this null correlation was undermined by low experimental repeatability, i.e. individuals did not behave similarly in repeated trials of the experiment. However, in the 8-week experiment, hiding behaviour was relatively consistent within individuals, so the null correlation at this scale indicates, less equivocally, that hiding and coloration are unrelated. Furthermore, the difference between the hiding behaviour of infected and uninfected isopods varied over 8 weeks, suggesting that the effect of A. lucii infection on host behaviour changes over time. We emphasize the importance of carefully designed protocols for investigating multidimensionality in host manipulation.

Parasitic infection is often associated with changes in host life-history traits, such as host development. Many of these life-history changes are ultimately thought to be the result of a depletion or reallocation of the host's resources driven either by the host (to minimize the effects of infection) or by the parasite (to maximize its growth rate). In this paper we investigate the energetic budget of Aedes aegypti mosquito larvae infected by Vavraia culicis, a microsporidian parasite that transmits horizontally between larvae, and which has been previously shown to reduce the probability of pupation of its host. Our results show that infected larvae have significantly less lipids, sugars and glycogen than uninfected larvae. These differences in resources were not due to differences in larval energy intake (feeding rate) or expenditure (metabolic rate). We conclude that the lower energetic resources of infected mosquitoes are the result of the high metabolic demands that microsporidian parasites impose on their hosts. Given the fitness advantages for the parasite of maintaining the host in a larval stage, we discuss whether resource depletion may also be a parasite mechanism to prevent the pupation of the larvae and thus maximize its own transmission.

Current theories assume that adaptive parasite evolution explains variation in the level of virulence and parasite success. In particular, mixed-genotype infections by parasites should generally be more virulent, and faster multiplying strains more successful, either because fixed strategies have evolved or because parasites facultatively alter virulence in response to co-infecting competitors. We compared several measures of parasite success and virulence between single-clone and mixed-clone infections of 2 strains of the trypanosome Crithidia bombi in its bumblebee host, Bombus terrestris. Contrary to expectation, we could not find differences between single-clone and mixed-clone infections in parasite prevalence, infection success, duration and clearance rate. However, a clearly significant effect of colony on infection intensity was present, and the colony effect emerged in virtually all other measures. We thus conclude that host environment as defined by the family (colony) genotype and thus host heterogeneity are more important in determining parasite virulence than the parasite characteristics. This does not invalidate modern theories of parasite evolution but suggests that variation in both hosts and parasites must be taken into account in more detail.

Genome sequence information in combination with new technologies has allowed researchers to approach genetic problems in new ways. High-density oligonucleotide arrays were used to probe the genome content of the yeast Saccharomyces cerevisiae. We show that these arrays, containing oligonucleotides complementary to the sequenced strain of S. cerevisiae, can be used to identify open reading frames that are missing or present in higher or lower copy number in related isolates of S. cerevisiae. We apply this method to the characterization of the genome of a strain derived from a clinical isolate of S. cerevisiae. Our results show that the telomeres are the regions with the most variability between the two strains.

It is only within the last half-dozen years that any serious attempt has been made to deal systematically with the parasitic worms of Australia. In 1909 Dr T. Harvey Johnston commenced what has proved to be a most extensive and successful survey of the cestode parasites. Dealing at first chiefly with Southern forms, he has latterly (1912) continued his investigations in Queensland and has extended them to include Nematodes and Echinorhynchs. Some years previously (1901) Dr S. J. Johnston of Sydney had started a study of the Trematode parasites, but this did not lead to any outstanding results until last year, when his important paper on the Trematode parasites of Australian frogs appeared. He has followed this up by another important work on Trematodes from North Queensland, which has appeared during the present year.

Trypanosoma (Megatrypanum) incertum Pittaluga 1905 was found in 33 out of 206 Pipistrellus pipistrellus caught at various sites in Britain. The trypanosome is described from blood smears. Development took place in laboratory-reared Cimex pipistrelli and Cimex lectularius. Epimastigote forms initially multiplied rapidly in the ventriculus and midgut of Cimex. Metacyclic trypanosomes were found in the rectum of both species of Cimex after 8 days when bugs were maintained at 20 °C and as early as 3 days at 30 °C. Electron microscopy of infected bugs revealed that there was no attachment to epithelial cells of the ventriculus or midgut, but within the rectum epimastigotes were attached by their flagella to the cuticle of the rectum by hemidesmosomes. Transmission was achieved by feeding experimentally infected bugs to bats kept in the laboratory. These bats were negative as judged by xenodiagnosis using laboratory-reared Cimex. Bats which had been caught in the wild demonstrated low-grade or sub-patent parasitaemias (positive in xenodiagnosis) for up to 400 days after the day of capture. Despite an extensive search of impression smears of tissues immediately after trypanosomes first appeared in the blood of experimentally infected bats no multiplicative stages were found.

General comparative note on the two species.

Pediculus humanus (vestimenti) is a larger, more robust and less active insect than P. capitis,—the ♀♀ having a relatively greater egg-carrying capacity than those of the head louse. The eggs are larger and the number laid (under the conditions of these experiments) is greater, while the habits associated with egg laying differ, although placing the ♀♀ of humanus under conditions applicable to capitis or vice versa may induce a considerable degree of uniformity. Cross pairings between the insects are easily brought about and the offspring are fertile inter se. Hybrid strains were maintained until the F. 3 generation and there seemed no reason, judging from breeding results, why such strains should not be continued indefinitely. Nevertheless the marked disparity in the sexes of the F. 1 generation of some of the crosses between P. capitis ♂ and P. humanus ♀ suggests that the parents are specifically distinct.

No such obvious disparity occurred between the sexes of the F. 2 and F. 3 hybrid generation, or of either of the pure stocks.

Habits. The body louse exhibits some of the habits of a gregarious animal especially during the moulting phases, also a preference for returning to the same spot for oviposition, which leads to the clustering of its eggs. These habits are shown, though in a less marked degree, by P. capitis, and it is possible therefore that they are to some extent the outcome of confinement. Pairing within both species took place at any time during day or night, and was very frequently observed after feeding. ♂♂ with but little food in their alimentary tract were, however, often seen in coitus. The period during which the insects remained paired was frequently observed to be over an hour, but no upper limit was defined.

A ♂ of P. humanus fertilized 18 out of 21 ♀♀ placed with him in succession. Four attempts with P. capitis were less successful; one ♂ fertilized ten ♀♀ and very possibly might have equalled the P. humanus record but for a scarcity of virgin ♀♀ while the experiment was in progress. The longest period during which a ♀ of P. humanus retained the power to lay fertile eggs in the absence of a ♂ was 20 days, usually it would seem to be from 16 to 18 days. In the case of P. capitis the period was shorter; 12 days being the longest ascertained period, while it was more usually from seven to eleven days.

The greatest number of eggs laid by any one ♀ of P. humanus was 295, an average of 6·4 per day—the daily average of a number of ♀♀ being 5·1. P. capitis ♀♀ showed a lower fecundity, the highest record being 141 with a daily average of 4—the general average being 3·7. These figures are probably exceeded under natural conditions. An experiment in differential feeding with P. humanus (Table VIII) shows clearly that fecundity is dependent on feeding. When extra feeding time over and above seven hours per day was given the average for four ♀♀ was eight per day. It is reasonable to suppose that the average for P. capitis would also be increased by unrestricted feeding.

The fertility of the eggs laid was not affected by increased feeding. The greatest number of fertilized eggs laid by a ♀ P. humanus after the removal of the ♂ was 115 (♀ No. 9), with a ♀ showing a higher daily laying average this might well be exceeded. With P. capitis the parallel figure is 70 (♀ No. 9). The ♀♀ of both species, after arriving at maturity, started oviposition irrespective of their having paired or not, but eggs laid by virgin ♀♀ were invariably infertile.

Length of life. The life of the ♂ P. humanus used in the experiment recorded in Table I was 32 days; the longest ♀ life was 46 days, with an average of 34. For P. capitis the figures were: ♂ life 30 days; ♀ life 38 days, with an average of 27 days. Whether or not the average lives of the insects would be extended by unrestricted feeding is an open question.

The life of the hybrid insects was not noticeably shorter than that usual for P. humanus, and they seemed to thrive better than P. capitis.

Tests made with unfed P. humanus showed that the longest lives were at a medium temperature of 16° to 18° C., many of the insects living from three to four days, while two lived five and one lived seven days. At 24·5° C. all died within five days. At 36·1° C. all died within three days.

Newly-hatched larvae, unless fed, lived less than 24 hours at 36·1° C., and when kept in a box in the vest pocket they lived but little more than a day; none survived a second day.

Adults kept in a box in the side pocket of a coat lived five days without food; this was in March.

Moulting. 40 young lice were reared in a box carried in a vest pocket and particulars of their moulting recorded.

1st moult: 3% moulted on the 3rd day; 42% on the 4th and 55% on the 5th day.

2nd moult: 15% moulted on the 7th day; 72% on the 8th and 13% on the 9th day.

3rd moult: 5% moulted on the JOth day; 3% on the 11th, 55% on the 12th, 32% on the 13th day, while 5% took 14 days to reach maturity.

The ♂♂ usually mature rather earlier than the ♀♀

Cold. Active specimens of P. humanus survived two days at a temperature of −2·3° C. to −1·1° C., but none recovered after exposure to these conditions for a week.

Hatching of eggs. Table IX shows that under humid conditions at 31° C. 3% of the 1300 eggs tested hatched on the 7th day; 56% on the 8th; 33% on the 9th; 8% on the 10th and ·2% later on the same day or on the 11th.

A test of batches of eggs taken from a stock box, some of which must have been laid several days previously, showed that none hatched at 15·6°−18·4° C., while at 24·5° C. there was considerable egg mortality, and the hatching period was spread over a longer period than usual, though not to the extent mentioned by Warburtou (1909); at 36·l° C. hatching was spread over five days and the mortality was not excessive.

To give some idea of the possible rate of multiplication of P. humanus we may estimate the egg period as 12 days and a further 12 days to the maturity of the ♀♀. Allowing an average of eight eggs per day, spread over a fertility period of 40 days, we find that, during her life, a single ♀ may have 4160 offspring.

Two simple methods of measuring the levels of inequality in reproductive success of different individuals in parasite populations are presented. These techniques are then applied to a number of sets of data for cestodes and acanthocephalans. The analysis suggests that both population density and host nutrition are important in determining the observed degree of inequality in reproductive success and body size. Cestodes, with a more flexible growth form, are shown to exhibit higher levels of inequality than acanthocephalans. The discussion outlines the evolutionary importance of considering variation in the reproductive success of different individuals.

A wide variety of nematode species have been observed to invade the central nervous system. They may be located in the meningeal spaces or may penetrate into the tissues of the brain and spinal cord.

The pathological changes resulting from invasion of the central nervous system are influenced by the route of entry, the size and the mobility of the parasite. They may be diffuse or focal and may include haemorrhage, degenerative changes, cellular infiltration and glial proliferation. Such changes may or may not be observed in close association with the parasite.

Symptoms indicating involvement of the central nervous system have long been associated with nematode infections outside the central nervous system. The pathogenesis of these symptoms is obscure, but they may possibly be of allergic origin.

The direct pathological effects on the central nervous system are mainly the result of trauma and are directly proportional to the size and activity of the parasite. The possibility that nematodes may transport viruses into the central nervous system is briefly discussed.

Chemical cues are so important in the recognition mechanism of social insects that most social parasites (which rely on hosts to rear their brood) have been documented as overcoming the mechanism by which colony residents recognize non-nestmates, by mimicking the odour of the usurped colony. We simulated in the laboratory the process by which the obligate social parasite, Polistes semenowi, invades nests of the host species, Polistes dominulus, in the field and analysed the epicuticular lipid layer before and after host nest usurpation. The experiment documents that P. semenowi social parasites have an epicuticular hydrocarbon pattern which is very similar to that of their host but, after entering host colonies, parasites mimic the odour of the colonies they invade, to the point that they perfectly match the hydrocarbon profile peculiar to the colony they entered. However, both before and after host nest invasion, parasites show a tendency to possess diluted recognition cues with respect to their hosts.

The effects of dietary protein and zinc restrictions on Heligmosomoides polygyrus were compared following primary and challenge infection in female BALB/c mice fed either control (24%), marginal (7%) or low (3%) protein combined with either high or low zinc (60 or 3 mg Zn/kg diet). Dietary protein restriction (3%) resulted in significantly lower body weight gain. As well, blood urea nitrogen (BUN) significantly decreased with decreasing dietary protein level. However, neither plasma albumin concentration nor relative thymus or spleen weights were reduced. Marginal zinc deficiency was confirmed by significantly lower tibia and liver zinc concentration, but food intake, body weight gain, relative thymus and spleen weights, and alkaline phosphatase activity were not altered. On day 29 post-primary infection, worm burdens were significantly higher in mice fed either marginal or low protein and in mice fed a low zinc diet, while parasite egg output was significantly higher in mice fed both low protein and low zinc diets. Immune status was compromised in mice fed low protein (significantly lower serum IgG1 and lower eosinophilia), and in mice fed low zinc diet (significantly lower eosinophilia). Early in the infection, IgE titres were elevated in mice fed low protein or low zinc, but IgE titres declined to levels lower than the control diet groups after 14–21 days. On day 29 post-challenge infection, worm burdens and parasite egg output were significantly higher in mice fed low protein, whereas the other groups had expelled almost all parasites. Dietary restriction had no effect on serum IgE. Significantly reduced serum IgG1 titres and eosinophilia in mice fed 3% protein supported the view that low dietary protein but not low zinc increased host susceptibility to H. polygyrus by compromising host immune function following reinfection in immunized mice.

We describe the new microsporidium Hamiltosporidium tvaerminnensis gen. et sp. nov. with an emphasis on its ultrastructural characteristics and phylogenetic position as inferred from the sequence data of SSU rDNA, alpha- and beta-tubulin. This parasite was previously identified as Octosporea bayeri Jírovec, 1936 and has become a model system to study the ecology, epidemiology, evolution and genomics of microsporidia - host interactions. Here, we present evidence that shows its differences from O. bayeri. Hamiltosporidium tvaerminnensis exclusively infects the adipose tissue, the ovaries and the hypodermis of Daphnia magna and is found only in host populations located in coastal rock pool populations in Finland and Sweden. Merogonial stages of H. tvaerminnensis have isolated nuclei; merozoites are formed by binary fission or by the cleaving of a plasmodium with a small number of nuclei. A sporogonial plasmodium with isolated nuclei yields 8 sporoblasts. Elongated spores are generated by the most finger-like plasmodia. The mature spores are polymorphic in shape and size. Most spores are pyriform (4·9–5·6×2·2–2·3 μm) and have their polar filament arranged in 12–13 coils. A second, elongated spore type (6·8–12·0×1·6–2·1 μm) is rod-shaped with blunt ends and measures 6·8–12·0×1·6–2·1 μm. The envelope of the sporophorous vesicle is thin and fragile, formed at the beginning of the sporogony. Cytological and molecular comparisons with Flabelliforma magnivora, a parasite infecting the same tissues in the same host species, reveal that these two species are very closely related, yet distinct. Moreover, both cytological and molecular data indicate that these species are quite distant from F. montana, the type species of the genus Flabelliforma. We therefore propose that F. magnivora also be placed in Hamiltosporidium gen. nov.

The high level of protection which can be induced by vaccination of a range of hosts, from rodents to primates, with live radiation-attenuated schistosome larvae offers great promise for development of a human schistosome vaccine. Studies of the irradiated vaccine models benefitted from significant funding during the 1970–90s and much was learned concerning the inducers, targets and mechanisms of immunity. Less progress was made in definition of the protective antigens involved. The application of new techniques for identifying membrane and secreted antigens has recently provided new vaccine candidates and a new impetus for schistosome vaccine development. This article is intended as an overview of some of the main lessons learned from the studies of the irradiated vaccines as a backdrop to renewed interest in schistosome vaccine development.

This study reports that the cytopathic effect of Trichomonas vaginalis, an important human parasite of the urogenital tract, occurs due to mechanical stress and subsequent phagocytosis of the necrotic cells. The investigation was done using a primary culture of bovine oviduct epithelial cells (BOECs), grown either in monolayers or as floating cells. Trophozoites displaying different virulence levels were co-incubated with BOECs for times varying between 1 min and 48 h. Analyses were performed using videomicroscopy, scanning and transmission electron microscopy, colourimetric assays and cytochemistry. Injury was observed as early as 1 h after incubation, while after 12 h the host cells were severely damaged when a fresh trichomonad isolate was used. Trichomonads attack the host cells by clustering around them. Mechanical stress on the microvilli of the host cells was observed and appeared to induce plasma membrane damage and cell death. After membrane injury and lysis, fragments of the necrotic cells were ingested by trichomonads. Phagocytosis occurred by trichomonads avidly eating large portions of epithelial cells containing the nucleus and other organelles, but living or intact cells were not ingested. Necrotic fragments were rapidly digested in lysosomes, as shown by acid phosphatase and ruthenium red assays where only the BOECs were labelled. The lytic capacity of the trichomonads was more pronounced in host cell suspensions.

Using spleen cells of mice infected or immunized respectively with cercariae or antigen preparations of Schistosoma mansoni, S. haematobium or S. japonicum monoclonal antibodies (mAbs) were produced against the schistosome gut-associated antigens CAA (circulating anodic antigen) and CCA (circulating cathodic antigen). Fusions nearly exclusively produced either anti-CAA (n = 25) or anti-CCA mAbs (n = 55) with a strong isotype restriction (IgM, IgG1 and IgG3) against both antigens, the majority of anti-CAA mAbs being IgG1 and the majority of anti-CCA mAbs being IgM. The mAbs, which on the basis of their selection were reactive with multiple carbohydrate epitopes of CAA or CCA, were applied in different immunological techniques including immunofluorescence, a dot immunobinding assay and immuno-electrophoresis to study the epitope repertoire. Anti-CAA mAbs were found to be reactive with 5 different epitopes, none of which occurred as multiple epitopes on eggs. Anti-CCA mAbs, on the other hand, recognized at least 10 different epitopes, while 44% of anti-CCA mAbs recognized epitopes common to the adult worm and the egg. Both CAA-and CCA-epitopes were found to be developmentally expressed at the level of the tegument in cercariae, schistosomula and 5-day-old lung worms, but in the adult worm were primarily found in the gut. Thus, the production of panels of mAbs has not only resulted in the selection of reagents optimally performing in diagnostic immunoassays, but also allowed a more detailed study of the epitope repertoire of these important schistosome antigens.

Nitric oxide (NO) has been demonstrated to be the principal effector molecule mediating intracellular killing of Leishmania, both in vitro and in vivo. We investigated the type of cell death process induced by NO for the intracellular amastigote stage of the protozoa Leishmania. Specific detection methods revealed a rapid and extensive cell death with morphological features of apoptosis in axenic amastigotes exposed to NO donors, in intracellular amastigotes inside in vitro – activated mouse macrophages and also in activated macrophages of regressive lesions in a leishmaniasis-resistant mouse model. We extended our investigations to the dog, a natural host-reservoir of Leishmania parasites, by demonstrating that co-incubation of infected macrophages with autologous lymphocytes derived from dogs immunised with purified excreted-secreted antigens of Leishmania resulted in a significant NO-mediated apoptotic cell death of intracellular amastigotes. From the biochemical point of view, NO-mediated Leishmania amastigotes apoptosis did not seem to be controlled by caspase activity as indicated by the lack of effect of cell permeable inhibitors of caspases and cysteine proteases, in contrast to specific proteasome inhibitors, such as lactacystin or calpain inhibitor I. Moreover, addition of the products of two NO molecular targets, cis-aconitase and glyceraldehyde-3-phosphate dehydrogenase, also had an inhibitory effect on the cell death induced by NO. Interestingly, activities of these two enzymes plus 6-phosphogluconate dehydrogenase, parasitic enzymes involved in both glycolysis and respiration processes, are overexpressed in amastigotes selected for their NO resistance. This review focuses on cell death of the intracellular stage of the pathogen Leishmania induced by nitrogen oxides and gives particular attention to the biochemical pathways and the molecular targets potentially involved. Questions about the role of Leishmania amastigotes NO-mediated apoptosis in the overall infection process are raised and discussed.

This study quantifies the rate and intensity of re-infection with human hookworm and Schistosoma mansoni infection 12 months following successful treatment, and investigates the influence of socio-economic, geographical and environmental factors. A longitudinal study of 642 individuals aged over 5 years was conducted in Minas Gerais State, Brazil from June 2004 to March 2006. Risk factors were assessed using interval censored regression for the rate and negative binomial regression for intensity. The crude rate and intensity of hookworm re-infection was 0·21 per year (95% confidence interval (CI) 0·15–0·29) and 70·9 epg (95% CI 47·2–106·6). For S. mansoni the rate was 0·06 per year (95% CI 0·03–0·10) and intensity 6·51 epg (95% CI 3·82–11·11). Rate and intensity of re-infection with hookworm were highest among males and positively associated with previous infection status, absence of a toilet and house structure. Rate and intensity of S. mansoni re-infection were associated with previous infection status as well as geographical, environmental and socio-economic factors. The implications of findings for the design of anti-helminth vaccine trials are discussed.

Three species of poeciliids (Gambusia holbrooki, Xiphophorus helleri and X. maculatus) and 15 species of ecologically similar native freshwater fishes (mainly eleotrids, ambassids, melanotaeniids and retropinnids) were examined for parasite richness to investigate parasite flux, qualitative differences, quantitative differences and the structuring factors in parasite communities in the 2 fish types in Queensland, Australia. Theory suggests that poeciliids would harbour depauperate parasite communities. Results supported this hypothesis; poeciliids harboured more species-poor parasite infracommunities and regional faunas than natives (P < 0·0001), despite greater sampling effort for the former. Cluster analysis of presence/absence data for poeciliids and the 6 most-sampled native fishes revealed that parasite communities of the 2 fish groups are qualitatively distinct; the proportion of parasite species with complex life-cycles was lower in poeciliids than in native species, and Myxosporea, Microspora, Coccidia and parasitic Crustacea were all absent from poeciliids. Limited exchange of parasite species has occurred between natives and poeciliids. Logistic ordinal regression analysis revealed that fish origin (exotic or native), environmental disturbance and host sex were all significant determinants of parasite community richness (P < 0·05). Theoretical modelling suggests that poeciliids are at a competitive advantage over native fishes because of their lack of parasites.