Content area
The precise transmission mode(s) of acephaline gregarines in their earthworm hosts has long been questioned, yet a rigorous experimental evaluation of sexual transmission is currently lacking. That Monocystis sp., a common gregarine parasite of the earthworm Lumbricus terrestris, infects the sexual organs of its host is suggestive of sexual transmission. Considering the divergent evolutionary consequences of various modes of transmission, excluding or proving sexual transmission in this host– parasite system is critical to fully understanding it. We cultured uninfected earthworms from cocoons and subsequently mated them to either an infected or uninfected partner (from the wild). We then compared these individuals with an orally infected group, which were infected using a newly developed gavage (oral injection) method. Our data have unambiguously established that (1) horizontal sexual transmission does not play a significant role in the transmission of Monocystis sp., and (2) oral transmission through the soil is likely the principal mode of transmission between earthworms. This finding is important to models of mate-choice because infection avoidance does not appear to drive mating decisions. Finally, we further report a successful and relatively simple method to obtain infection-free individuals, which can subsequently be infected via oral gavage and used in empirical studies.
Although acephaline gregarine parasites have been studied for most of the past half-century, the precise transmission modes of gregarines within the earthworm Lumbricus terrestris remain to be rigorously tested and clarified (J. Janovy, Jr., pers. comm.). Monocystis sp. (Fig. 1) is perhaps the most common and best-known parasite of L. terrestris. The majority of authors support the notion that Monocystis sp. is transmitted through the ingestion of soil, which has been contaminated with sporocysts (Bhatia, 1924; Troisi, 1933; Miles, 1962; Olsen, 1986; Bush et al., 2001; Roberts and Janovy, 2005). However, various other modes of transmission have been suggested for quite some time. One-hundred and sixty years ago, Dujardin (1845), after discovering intact sporocysts in the gut of a mole, suggested that infection of a vertebrate vector played an essential role in the life cycle. Other authors have since reported the presence of sporocysts in the gut of vertebrates (Bull et al., 1998). However, because Monocystis sp. is solely an invertebrate parasite, it is likely that its presence was merely the result of earthworm predation, rather than an essential step in the parasite life cycle. Such predation, although not involving an intermediate host per se, could of course facilitate sporocyst dispersal if sporocysts remain infective following gut passage (Röttger, 1995).
[Image Omitted. See PDF.]
Furthermore, Keilin (1925) suggested that sporocysts are released into the soil by parasite-induced autotomy, Hahn (1928) claimed a ‘reverse pathway’ in which parasite stages enter the gut from the coelom and are passed in the fecal matter, whereas Loubatières (1955) and Röttger (1995) suggested that sporocysts exit via the dorsal pores of earthworms. Further, sporocyst passage via the vas deferens to the environment has been proposed, whereby sporocysts exit the body through the male pores (Olsen, 1986; Roberts and Janovy, 2005). Although published before 3 of the aforementioned studies, Miles (1962) ostensibly rejected all of these hypotheses by successfully growing parasite-free earthworms in soil that had been sterilized and previously inhabited by parasitized earthworms.
Finally, Schmidt (1854) was the first to suggest that sporocysts were passed among earthworms during copulation, a notion that has been supported by Bhatia (1924) and Troisi (1933), who found small quantities of sporocysts in the spermathecae of adult earthworms. At this point, a clear distinction between sporocyst transfer between copulating partners and sexual transmission must be made. If sporocysts are transferred during copulation into the spermathecae, there is the possibility that offspring may become infected as the spermathecal contents are distributed to cocoons, i.e., vertical transmission. However, sexual transmission between mating partners implies that sporocysts are capable of hatching within the spermathecae, sporozoites migrating to the seminal vesicles, and subsequently developing an infection, i.e., horizontal transmission. The reproductive, evolutionary, and epizootiological consequences of these 2 modes of transmission are drastically different (Lipsitch et al., 1996; Thrall and Antonovics, 1997; Kover and Clay, 1998; Thrall et al., 1998; McLachlan, 1999; Day and Proulx, 2004; Knell and Webberley, 2004). Furthermore, various transmission modes are critical to questions of mate-choice. ‘Good genes’ (Hamilton and Zuk, 1982) and ‘contagion indicator’ (Able, 1996) models differ primarily in that good gene hypotheses maintain indirect benefits of superior resistance to parasites in offspring, whereas contagion indicator hypotheses maintain direct benefits of avoiding self-infection.
Miles (1962) has contributed considerably to our understanding of the transmission of acephaline gregarines. These experiments, however, were not conducted in the L. terrestris–Monocystis sp. system, but in the taxonomically related Eisenia foetida–Apolocystis elongata system. Apart from model organisms, our study differs 2 additional ways. First, experimental infections were previously induced by inoculating a food source and making it available for consumption. This has the complication that dose effects, i.e., variation in feeding rate, are not controlled, which may have affected parasite prevalence. Second, horizontal sexual transmission was not experimentally assessed. That vertical transmission was excluded suggests that sporocysts are not transferred during copulation; it does not, however, exclude the possibility of sexual transmission between adults.
The current study experimentally evaluates whether horizontal sexual transmission between mating partners is a mode of transmission used by acephaline gregarine parasites (Monocystis sp.) of the earthworm L. terrestris.
MATERIALS AND METHODS
Juvenile earthworms
Lumbricus terrestris were obtained by growing hatchlings from cocoons that were produced by earthworms purchased from a local sport angler shop, which provides earthworms captured from the wild. Seventy-nine juvenile earthworms were grown for more than 18 mo in a climate chamber at 12–14 C and complete darkness. Upon hatching, earthworms were immediately isolated in plastic containers (500 ml) and placed in a mixture of sterilized garden soil (Torfwerk Breesen Peatmoss Blumenerde, Breesen, Germany): Salt < 3.0 g/L, pH = 5– 6.5, nitrogen 100–500 mg/L, phosphate 100–500 mg/L, potassium 100– 600 mg/L, and dried horse manure. The horse manure was not sterilized, i.e., autoclaved, to provide a source of nutrients (trials without this nutrient source resulted in severely impeded growth and survivorship; S. Field, unpubl. obs.). Because vertical transmission in acephaline gregarines had been previously excluded as a major transmission pathway (Miles, 1962), this process should allow us to obtain virgin, parasite-free earthworms for transmission experiments (Miles, 1962; Olsen, 1986). Growing earthworms from cocoons was necessary because Monocystis sp. has an exceedingly high prevalence in wild populations, and the determination of infection status is destructive. Additionally, in this way, we were able to control for variation in life history strategy between individuals.
Sexual transmission
When earthworms reached sexual maturity (indicated by the presence of a full clitellum), 8 randomly chosen earthworms were killed and checked for sporocysts (referred to as C1). Once the absence of Monocystis sp. was confirmed, earthworms were placed in an isolation setup (see Field and Michiels, 2005, for setup design) for 2 wk so that individuals could acclimate to their new burrows. To recreate the typical vertical burrows of L. terrestris, worms were placed in PVC cable boxes (50 × 3 × 1.5 cm) filled with soil. The bottom ends of the tubes were secured with duct tape punctured with 10 small holes to allow the entrance of water while preventing escape.
Upon beginning the experiment, 44 earthworms were weighed, paired with an individual of approximate size, and placed into the mating setup by simply exchanging the PVC ‘burrow’ from the isolation to the mating setup. The mating setup consisted of a flat Plexiglas sheet divided into compartments by walls (15 cm), which prevented extra pair copulations and escape (Fig. 2). Rectangular holes were made in the Plexiglas floor into which the artificial burrow openings were fitted; the bottom rested in a square plastic container containing approximately 1 cm of water. Individuals were paired with either an infected partner purchased from the sport angler shop (T1 = 30) or an uninfected partner that was similarly raised in the climate chamber (C2 = 14), i.e., 7 pairs.
[Image Omitted. See PDF.]
Earthworms were observed by infrared camera with frame-grabber software (CyberOptics Semiconductor, Beaverton, Oregon), which took digital pictures at 30-sec intervals. In this way, we observed and timed all copulations for all individuals. The climate chamber was set to a 12: 12 hr (light:dark) cycle, and a dimmed light bulb was placed in a far corner to further mimic natural conditions experienced during copulation in the wild. To ensure sufficient sperm transfer during mating, individuals in the mating setup were separated only after 2 complete matings. On some occasions, matings were interrupted by the lights being turned on in the climate chamber; thus, the additional restriction of at least 400 min of mating was imposed (an average uninterrupted mating is typically 200 min; S. Field, pers. obs.). During the video observation phase of the experiment middens, i.e., earthworm casts, were removed daily to exclude the possibility of infection via the consumption of contaminated middens from an infected partner (T1).
While in the mating setup, earthworms were fed strips of frozen lettuce. In preparing the lettuce, the outer leaves and the bottom (head lettuce) were removed, and the strips were thoroughly rinsed with water to further eliminate the risk of Monocystis sp. contamination via soil particles on the lettuce. Following successful mating, the artificial burrows (containing its earthworms) were removed from the mating setup and returned to the isolation setup. Thereafter, each earthworm remained in the isolation setup for exactly 84 days (12 wk) to allow a putative infection to develop, after which they were weighed and killed by placing them into Eppendorf tubes and freezing at −80 C, where they remained until parasite evaluation. A third control group (C3) was also maintained; it had neither access to the mating setup nor interaction with any earthworm at any time. This group (n = 12) remained in the isolation setup at all times and were killed on the final day of the experiment to demonstrate that animals remained parasite-free for the course of the experimental infections. For all individuals that mated (C2 and T1), we determined the cocoon production during the experiment by sieving the burrow contents following earthworm demise.
Oral transmission
To provide a relative group to which putative infections via the sexual pathway could be compared, we also investigated the classic transmission mode of Monocystis (Miles, 1962; Bush et al., 2001; Roberts and Janovy, 2005), in which a further 15 earthworms (T2) were infected with Monocystis sp. via oral gavage (injection of liquid directly into the esophagus). Five parasitized earthworms (wild/purchased) were killed and their seminal vesicles harvested and pooled. The seminal vesicles were then placed in an Eppendorf tube and homogenized by hand with a plastic piston. The sporocyst concentration of this pooled sample, i.e., Monocystis sp. source, was then determined using a counting chamber (Field et al., 2003) and, thereafter, diluted to a final working concentration of approximately 5,000 sporocysts/μl. Of this working concentration, 4 μl were introduced deep into the esophagus with a positive displacement pipette (Gilson Microman model M10; pipette tip: capillary piston model CP10; Mandel Scientific, Guelph, Canada). Earthworms were first placed on ice for 5 min to reduce activity, after which they were removed and placed on a moist paper towel. The rounded plastic pipette tip was then gently inserted into the mouth and held such that the worm could not retreat. The posterior end of the earthworm was then agitated, which tends to induce the individual to crawl up the tip (3.5 cm), causing the opening to insert deep into the esophagus. At this point, the pipette contents were expelled and the tip gently removed.
Putative infections were then allowed to develop for 84 days after which individuals were killed and parasite presence was evaluated (see below). This experimental group, although described as a treatment group, can be viewed as a positive control group for the sexual transmission treatment and served to ensure that the duration of the experiment was long enough to allow an infection to develop. As such, because of the likely discrepancy in dose between this and the sexual treatment, we do not carry out direct statistical comparisons among these groups.
Parasite evaluation
Individuals were thawed, dissected via a dorsal incision, and the seminal vesicles removed and placed into fresh Eppendorf tubes. They were then weighed and homogenized with a sonicator (Bandelin sonopuls; Model UW 2070, Burladingen, Germany). Analysis of seminal vesicle contents was carried out following the ‘feathering’ smear technique used in analyzing human semen samples (WHO, 1999). Ten microliters of the homogenized seminal vesicle fluid were transferred to a microscope slide, carefully smeared, and allowed to air-dry, then embedded with Kaiser's glycerol gelatine, and examined for the presence of sporocysts. To determine whether Monocystis sp. is transferred to a receiving partner during copulation, the spermathecae, where received sperm is stored, were also removed and examined using the same technique, with the exception that the entire spermatheca was placed on the slide, i.e., not 10 μl. Slides were observed under dark field microscopy (×100). Detection began with a thorough left-to-right sweep across the length of the cover slip (24 × 40 mm), followed by an identical reverse sweep. If no sporocysts were observed, 25 random fields of view were selected from the entire slide area and checked for sporocysts. If there were still no observable sporocysts, the seminal vesicles were considered infection-free. Finally, for those individuals in which sporocysts were detected, the sporocyst concentration was determined using a counting chamber as per Field et al. (2003).
Analysis
After freezing, but before parasite evaluation, individuals from the C2, C3, T1, and T2 groups were randomly recoded (by an external observer) to ensure that the investigator was blind with reference to treatment group; the C1 group was not evaluated blindly because it was crucial to know a priori whether these individuals were infected. Because parasite load was right-skewed, for statistical analysis, parasite load was natural log-transformed to achieve a normal distribution (Sokal and Rohlf, 1995), and parametric tests were used whenever underlying assumptions were satisfied. Box-and-whisker plots were used, with the median as the middle bar, the whiskers representing the upper and lower quartiles, and the box representing the interquartile range of data points (with outliers as the points lying more than 1.5 interquartile ranges from the top or bottom of the box). Statistical analyses were carried out using SPSS software (version 11.5, Chicago, Illinois). It should also be noted that the occasional sporocyst was detected in the C2 and C3 groups by the more sensitive smear technique. Smears, compared with counting chambers, are generally more sensitive because they examine nondiluted samples of seminal vesicle fluid, whereas calculating sporocyst concentration using a counting chamber (hemocytometer) uses a diluted sample and sporocysts are only counted if they fall within a counting grid. However, the detection of 5–10 sporocysts within 10 μl of seminal vesicle fluid represents an extreme divergence from wild infections (4– 5 orders of magnitude; Figs. 3, 4).
[Image Omitted. See PDF.]
[Image Omitted. See PDF.]
RESULTS
Source worms
All individuals purchased from the angler shop were infected with Monocystis sp. Parasite load varied greatly ranging from 2.12 × 105 to 2.91 × 107 parasite sporocysts/individual (x̄ = 7.09 × 106 ± 5.92 × 106, n = 30).
Sexual infections
None of the individuals that mated with an infected partner developed an infection in their seminal vesicles. Furthermore, none of the spermathecae in these individuals (T1) was found to contain sporocysts, although they were thoroughly screened. Lastly, infection was also not apparent in earthworms that mated with uninfected partners (C2, Fig. 3).
Oral infections
Experimental oral infections of approximately 20,000 sporocysts via gavage resulted in 100% infection. Parasite load ranged from 3.80 × 104 to 5.89 × 105 sporocysts/individual (x̄ = 2.39 × 105 ± 1.86 × 105, n = 15), yet was significantly less than in the source group (wild earthworms, t = 10.64, df = 43, P < 0.0001, Fig. 3; Fig. 4 a vs b).
Paternity costs
To investigate a possible paternity cost of infection (sperm concentration), we analyzed the relationship between parasite load (of the infected/wild group) and cocoon production in their partner (T1). There was no correlation (Pearson r = −0.24, n = 30, P = 0.22; Fig. 4). Furthermore, we compared cocoon production between individuals that received either a parasitized or a parasite-free partner (Fig. 4) and found no significant difference between treatments (t = 0.52, df = 42, P = 0.61).
DISCUSSION
To our knowledge this is the first study to directly and rigorously assess horizontal sexual transmission in the Lumbricus terrestris–Monocystis sp. system. In our experimental infections (T1), no individuals became infected with Monocystis sp. following a minimum of 400 min of mating with a highly infected partner from the wild. This result is surprising considering the following: (1) the proximity of sporocysts to the male reproductive system would suggest the possibility of simultaneous transfer of sporocysts through the vas deferens (Olsen, 1986; Roberts and Janovy, 2005) during copulation; (2) sporocysts are found in such high numbers within the seminal vesicles; and (3) sporocysts are found so ubiquitously in wild populations. We also found no sporocysts within the spermathecae of individuals that mated with highly infected partners. This additionally suggests that sporocysts are not even transferred among mating partners and further supports the absence of horizontal sexual transmission. However, some authors have claimed to witness sporocysts in the spermathecae (Bhatia, 1924; Troisi, 1933). Miles (1962), and confirmed in the current study, has shown by the production of infection-free earthworms from cocoons, that vertical transmission is not a transmission mode of acephaline parasites in earthworms. Successfully raising infection-free earthworms from cocoons further suggests that any sporocysts that do reach the spermathecae are presented with a ‘cul-de-sac’ (until perhaps host death).
Why are sporocysts, present in high concentrations in the seminal vesicles, not expelled through the vas deferens? We hypothesize that spatial separation and physical barriers might prevent this from occurring. In our experience, we have yet to discover sporocysts in significant concentrations in the testis sacs, the ventral-lying region of the seminal vesicle that leads to the sperm funnel and eventually the vas deferens. Future studies should investigate whether the sperm funnels are physical barriers that prevent sporocysts from entering the vasa deferentia.
Our results are in accordance with Miles (1962) and support the general view that Monocystis sp., and in general, acephaline gregarines, are orally transmitted via ingestion of contaminated soil particles. In the oral infections (T2), we unmistakably detected significant numbers of sporocysts, indicating successful migration of sporozoites through the gut wall, migration to the seminal vesicles, and multiplication within the seminal vesicles.
The length of time required to develop a detectable infection is somewhat surprising. Orally infected earthworms were killed a minimum of 84 days (3 mo) following exposure to approximately 20,000 sporocysts and, nevertheless, had a mean parasite load 30 times lower than wild earthworms! This suggests that either infections develop very slowly or only very few sporozoites are actually successful in reaching and entering the seminal vesicles. If infections develop very slowly, earthworm longevity may account, at least in part, for the discrepancy in parasite load between the wild populations and the orally infected group because wild earthworms are generally assumed to have been infected at a younger age and, based on size, appeared to be older.
Because trophozoites are known to digest and consume developing spermatozoa, it is plausible that infection represents a paternity cost, especially under conditions of sperm competition; although highly infected individuals were not associated with reduced cocoon production in their partners (Figs. 5, 6). Under multiple mating, however, such a difference may become evident. Paternal success, sperm quality, and parasitism under conditions of sperm competition should represent a valuable future avenue of research.
[Image Omitted. See PDF.]
[Image Omitted. See PDF.]
Mülsow (1911) and Troisi (1933) also claimed to have induced experimental infections orally; however, no controls were applied, and there was no evidence that the experimental worms were initially infection-free (Miles, 1962). Miles' own experimental oral infections were ‘sporadic and unpredictable’, with only 30% of orally exposed individuals developing full infections. This is likely for reasons previously mentioned, namely, variation in feeding rate. He further mentions that there was no way to predict which individuals would eventually become infected and hypothesized that natural host resistance is involved. The current method represents a reliable procedure for experimentally infecting individuals for parasitology research.
As previously mentioned, the occasional sporocyst was discovered among the various controls. It is unlikely that this presence is the result of a natural infection, as the level of infection was so low, it is difficult to imagine its biological significance. In such cases, the sporocyst concentration would be considerably less than 10 sporocysts/μl, nearly 5 orders of magnitude less than the mean sporocyst concentration in wild earthworms. We suspect, although every effort was made to avoid it, contamination of dissection tools explains these sporadic incidents.
Our data have unambiguously established that horizontal sexual transmission does not play a significant role in the transmission of Monocystis sp., and oral transmission through the soil is likely the principal mode of transmission between earthworms. This finding is important to models of mate-choice because infection avoidance does not appear to drive mating decisions. Previous authors have expressed doubt about whether it is possible to obtain Monocystis sp.-free L. terrestris (Meier, 1956; Miles, 1962), yet our study further reports a relatively simple, although time-consuming, method of obtaining infection-free individuals for empirical studies in evolutionary ecology.
Acknowledgments
We thank Gregor Schulte for valuable help building the mating setup, for programming the digital cameras, and for coding individuals to avoid observer bias. We also thank Jan Heuschele and Gregor Schulte for assistance in introducing earthworms to their burrows and for caring for them during absences. Finally, we thank the 2 anonymous reviewers for their constructive and positive comments on previous versions of the manuscript. This study was funded by a grant from the German Science Foundation (DFG Mi 482/6–2).
Copyright Allen Press Inc. Apr 2006