Specialized host exploitation constrains the evolution of generalism in microsporidian parasites of Artemia

Get Complete Project Material File(s) Now! »

Experimental test of sex ratio-biasing in A. franciscana

To produce our experimental animals, we hatched native-range A. franciscana from diapausing eggs sampled in the Great Salt Lake, USA (41°10′ N, 112°35′ W; 2007), and invasive-range A. franciscana and A. parthenogenetica from cysts sampled in Aigues-Mortes, France (43°34′ N, 4°11′ E; 2013). All cysts were stored in dry condition at 4°C; they were hatched following a protocol modified from Bengtson et al. [S17]. Cysts were rehydrated with deionized water for 2 hours, decapsulated by a brief exposure (< 10 min) to 2% sodium hypochlorite, and rinsed with fresh water. Decapsulated cysts were incubated until emergence in an aerated saline medium (salinity 5 g/L) at 28°C and under constant light. We produced the saline medium by mixing concentrated brine (Camargue Pêche, France) with deionized tap water. All brine was autoclaved before use to ensure it contained no horizontally transmitted parasites; there are no (known) vertically transmitted parasites in these populations, so individuals hatched from dormant eggs are in principle parasite-free. After emergence, the experimental Artemia were transferred to large tanks of non-aerated medium at 23°C under natural light, and salinity was gradually increased to 90 g/L. They were fed ad libitum with a mixed solution of two parts Tetraselmis chuii (6.8*109 cells/L, Fitoplancton marino, Spain) to one part powdered yeast solution (0.4 g/L, Gayelord Hauser, France). Shortly before sexual maturity, we separated the individuals by species and sex [S18] and formed the parental groups (see above). All A. franciscana females used in the experiment were virgins.
Parental groups were kept at a constant density (40 individuals/L), light (36 W), temperature (21°C) and salinity (90 g/L). They were fed a standardized volume of T. chuii/yeast solution daily (5 mL or 20 mL for groups of 10 or 40 individuals, respectively; solution concentration: 3.4*109 T. chuii cells + 0.2 g yeast/L), and their mortality was monitored. The nauplii harvested from the parental groups were reared in separate jars under the same temperature, light and salinity conditions as the parental groups; their density was fixed (200 nauplii/liter, based on the initial count) and they were fed a standardized volume of the T. chuii/yeast solution daily (0.1 mL/nauplius, based on the initial count).

Statistical analysis

We analyzed the offspring sex ratios produced in the experiment using generalized linear mixed-effects models with a binomial error distribution [package lme4 in R 3.1.0; S14, S15]. The response variable was the number of A. franciscana males (successes) and females (failures) counted per parental group and per clutch. All models included Parental group as a random effect and Treatment as a fixed effect. We tested the scenarios described above by contrast manipulation, forcing the factor Treatment to have only two levels and attributing these levels to the different treatments (see Table 2). We then selected the best model by AIC comparison [S16]. The effect of absolute number of parents was tested by taking the best model and attributing a third factor level to treatment “invasive range*”. We also tested models that took parental mortality and time of collection into account, but these did not improve the AIC score and were discarded.

Duration of female reproductive stages

When females are sampled randomly in a population, the proportion of females in a certain reproductive stage can also be interpreted as the relative duration of that stage. In accordance with previous reports, we found that the stages E-A and E-B were much longer than stages C and D in A. franciscana females (Table 2) (Bowen 1962, Metalli and Ballardin 1970). There were some differences between samples, most notably the extended stage E-A at Site 9 and the extended stage C at Puit Romain. Our data did not allow us to see if the longer E-A stage at Site 9 extended the reproductive cycle, or if it was prolonged at the expense of another stage. The extended stage C at Puit Romain may have been a sampling artefact: many of the amplexing females from Puit Romain aborted after sampling, passing from stage E-B to C artificially (this did not occur in the other amplexing samples). This was detectable for amplexing females because they were stored separately, but would not have been for randomly sampled females. If the latter also aborted broods, we may have overestimated the proportion in stage C.

Discrimination by female reproductive stage, species, and size

A. franciscana females were more likely to be part of an amplexing pair when they were close to sexual receptivity, which occurs in stage D (Fig. 3) (cf. Pastorino et al. 2002). The relative risk of amplexus dropped sharply immediately afterwards (stage E-A), then increased as the next batch of oocytes began to mature (stage E-B). The high probability of amplexus during stage E-B, which lasts several days, fits the ‘hours to days’ duration of amplexus reported in the literature (Lent 1971, Wolfe 1973, Belk 1991, Rogers 2002). Oddly, stage C saw a dip in amplexus probability, relative to stage E-B, in all samples (Fig. 3). It is unclear why this should be, since females in this stage are mere hours away from becoming receptive (in stage D). One possibility, alluded to by Browne & Halanych (1989), is that males become dislodged as females give birth to their previous clutch (between stages E-B and C), or during the ensuing molt (during stage C). If so, dislodged males risk the loss of their invested effort if they cannot maintain their claim on the female in some other way during this period.
In addition, females of the asexual species A. parthenogenetica were much less likely to be amplexed. Previous experiments in our lab indicate that A. parthenogenetica females are tolerant of amplexus (T. Lenormand, unpublished data), suggesting that A. franciscana males are able to discriminate between con- and heterospecific females. This is the first report of species discrimination in Artemia, and it raises an intriguing question. The sexual species A. franciscana was introduced to France in 1970 (Rode et al. 2013c), from a native range in which there are no sympatric congeners (although A. franciscana does amplex other sympatric Anostraca occasionally, Belk and Serpa 1992). Has its ability to discriminate between conspecific and asexual females evolved since its invasion? Future work should provide an answer.
A. franciscana pairs were also size-assortative in two of the three samples (Fig. 4), confirming previous results (Forbes et al. 1992). The absence of size assortment in Fangouse may have been caused by the limited variation in body size at that site.
Finally, despite the wealth of discrimination demonstrated by A. franciscana, they remained in amplexus when the female partner was castrated by the cestode F. liguloides. This is obviously counterintuitive, but it may be explained by the very recent emergence of the parasite in this host: until 2013, F. liguloides was never seen infecting A. franciscana in Aigues-Mortes (Sánchez et al. 2012, Rode et al. 2013b, personal observation). The selection to discriminate against females castrated by this parasite must therefore be very recent; given time, A. franciscana males should evolve to reject these females (e.g. Bollache et al. 2002). It is unknown whether any of the cestodes infecting A. franciscana in its native range castrate female hosts (Redón et al. 2015b); if so, it would be particularly interesting to test whether A. franciscana males could identify and reject such females.

Male mate choice in Artemia: clasp first, think later

Our results fit well with previous authors’ conclusions about male mate choice in Artemia, which appears to be a ‘clasp first, think later’ affair. In experimental settings, male Artemia typically attempt amplexus with any female they encounter, without discriminating by female size or species (Belk and Serpa 1992, Forbes et al. 1992, Rogers 2002, T. Lenormand upublished data). Indeed, Rogers (2002) reports that “male Artemia franciscana […] were also willing to amplex bits of grass shaped like females, and the tip of a pencil”. Clearly, Artemia males are indiscriminate when amplexus is first attempted and established. When Artemia pairs are sampled from a stable population, however, they do show size- and genotype-assortative mating (this study, Browne et al. 1991, Forbes et al. 1992). A striking example of this distinction was given by Forbes et al. (1992), who showed that in experimental tanks, pairs sampled soon after they were formed were not size-assorted, while pairs sampled one week later were. Thus, mate choice emerges if individuals are given time to assess, and possibly separate from, their amplexed partner. In principle, pairs can separate due to male choice (releasing the female) or female choice (dislodging the male). However, females seem to primarily reject males in the first moments of amplexus (Forbes et al. 1992, Rogers 2002, Tapia et al. 2015, pers. obs.), so it is probable that any difference in discrimination between recently-amplexing and established pairs is primarily due to male choice. Clasping first and thinking later could be an optimized mate selection strategy for male Artemia. Artemia mate by scramble-competition polygyny (Alcock 1980), in which males avoid aggressive encounters and concentrate their reproductive energy into finding and fertilizing receptive females (Belk 1991). Such a strategy would favor an opportunistic approach to amplexus, with little mate evaluation beforehand (Belk 1991). However, Artemia males also invest heavily into mate guarding, and this high level of male investment should favor a considerable capacity for male mate choice (Edward and Chapman 2011). An ability to assess amplexed females, therefore, would be strongly selected for (Forbes et al. 1992). Indeed, similar behaviors have been reported for other crustaceans with reproductive scramble competition (Hunte et al. 1985, Galipaud et al. 2015), suggesting that indiscriminate amplexus coupled with post-amplexus choice is a robust male strategy in these conditions.

READ  A microscopic interpretation for adaptive dynamics trait substitution sequence models 

Assessment of amplexed females and the overall female population

How do males assess a female’s species and reproductive stage? As discussed above, male A. franciscana judge their partner’s quality during amplexus, ruling out the soluble ‘distance’ pheromones produced by many aquatic crustaceans for these purposes (e.g. Lonsdale et al. 1998, Zhang et al. 2010). Instead, Artemia males may recognize ‘contact’ pheromones bound to the female’s exoskeleton. These are used by many crustacean species (e.g. Thompson and Manning 1981, Borowsky and Borowsky 1987, Lonsdale et al. 1998, Ting et al. 2000, Goetze and Kiørboe 2008, Weeks and Benvenuto 2008), including cladocerans, some of Artemia’s closer relatives within the Crustacea (Van Damme and Dumont 2006, La et al. 2014). Alternatively, males may judge a female’s reproductive stage based on morphological cues, or on the intensity of her resistance to amplexus.
To decide whether to release an amplexed female, however, males must also know how she compares to other females in the population (Jormalainen 1998). Forbes et al. (1992) speculated that A. franciscana males continue to assess the availability of other females while in amplexus, releasing their amplexed partner if she is of comparatively low quality. A recent study of the amphipod Gammarus pulex, however, has shown that males do not base the decision to continue amplexus on the relative quality of their partner (compared to a passing female), but rather on her absolute quality (Galipaud et al. 2015). The authors attribute the lack of comparative decision making – which would be optimal – to the male Gammarus’ inability to assess female quality at a distance. Instead, through repeated contact with many females, they may judge the overall quality and availability of the female population, and use this information to establish an absolute threshold for amplexus continuation. Given that Artemia males also need contact to evaluate females, we favor this mechanism for A. franciscana.

Evolution of discrimination against hetero-specific females

Currently, an experiment is underway to test whether A. franciscana males from populations in the native range are pre-adapted to discriminate against A. parthenogenetica females when mate guarding. If not, we will have demonstrated that the ability to do so evolved after A. franciscana’s invasion of Southern France. Such a result would suggest that the mechanisms used by males to evaluate an amplexed female are more flexible, evolutionarily speaking, than those used by females to estimate the population sex ratio.

Sex-specific estimation of population parameters

The Artemia system in Aigues-Mortes may provide a unique opportunity to study the way males and females perceive and react to population cues. In Chapter 1, we showed that female A. franciscana can judge the sex ratio of a population. Based on the results of Chapter 2, we speculate that A. franciscana males may have a similar ability. Mate guarding theory expects an ability to judge the receptivity of individual females (which A. franciscana males have) to be coupled with an ability to assess the overall availability and receptivity of females in the population (Jormalainen 1998), so that males may adjust their willingness to guard non-receptive females as a function of the population sex ratio. This has been confirmed for various crustacean species (e.g. Manning 1980, Iribarne et al. 1995, Dick and Elwood 1996, Benvenuto et al. 2009), and we therefore expect A. franciscana males to have this ability as well.
Experiments testing whether A. franciscana males adjust their guarding criteria based on the overall sex ratio of the population could have three potential outcomes. First, if males can assess the population sex ratio without recognizing asexuals (like A. franciscana females), we could conclude that the two A. franciscana sexes are likely to use the same population cues. This would be the most parsimonious use of signaling in the species. Second, if males can assess the population sex ratio correctly (distinguishing between conspecifics and asexuals), we could conclude that A. franciscana males and females use a separate set of cues to evaluate the population (e.g. accumulated information from repeated contacts, Galipaud et al. 2015, vs.

Amplexus in the time of F. liguloides

One particularly interesting result of Chapter 2 is that A. franciscana males do not discriminate against F. liguloides-infected, castrated females when mate guarding. Until 2013, F. liguloides were never recorded infecting A. franciscana in Aigues-Mortes, so it is not unreasonable that males sampled in 2015 were unable to detect infections with this cestode (Sánchez et al. 2012, Rode et al. 2013b). Now, however, evidence indicates that F. liguloides is colonizing A. franciscana at a rapid pace (see Appendix 2), so selection to discriminate against castrated females should be increasing equally rapidly (Bollache et al. 2002). In the future, it would be particularly interesting to use resurrection studies (reviving dormant Artemia cysts of different ages, cf. Rode et al. 2011) to investigate whether males evolve this ability as F. liguloides establishes itself in their population.

Characteristics of the Aigues-Mortes saltern

We studied A. rigaudi and E. artemiae infecting A. franciscana and A. parthenogenetica in the saltern of Aigues-Mortes in Southern France (43.53°N, 4.21°E). This is a seasonal system, where temperature, salinity, and species composition vary throughout the year. Average monthly temperature ranges from 5-10°C in winter (December to February) to 20-25°C in summer (June to September, Fig. 1B). The salinity is highly variable, ranging from roughly 50 to roughly 250 g salt/L, but is generally higher from May to November than from January to April (Fig. 1A; we have no data for December). Artemia are present year-round in large quantities (the population size in estimated to be on the order of 109–1010 by commercial exploiter F. Gout, Camargue Pêche, Grau-du-Roi, France; personal communication), but their density is typically low in late winter and early spring (personal observation). Finally, the species composition of the Artemia population varies: A. parthenogenetica are entirely absent in winter, but form the majority of the population in summer (Fig. 1C).
From a microsporidian perspective, the Aigues-Mortes saltern is a metacommunity. It is made up of a network of large basins, between which water is allowed to flow or not as a function of the salt production process. This causes environmental factors such as salinity and food quality to vary, leading to variation in the outcomes of inter-host competition: A. franciscana or A. parthenogenetica can outcompete one another or coexist (Browne 1980, Browne and Halanych 1989, Barata et al. 1996b). At any given time, therefore, adjoining basins can contain different host communities, though water movement between the basins is regular enough that there is no spatial structure in the Artemia population (Nougué et al. 2015). The sites Fangouse, Puit Romain, and Site 9 are isolated from the general flow of water (though other types of dispersal are possible for Artemia, e.g. Brendonck and Riddoch 1999, Green et al. 2005). Within basins, Artemia communities are also well-mixed (Lenz and Browne 1991), so that we can assume that all the spore pools of A. rigaudi and E. artemiae spores are shared among the host species (Fels 2006).

Table of contents :

Section 1: Reproductive interference
Chapter 1: Maladaptive sex ratio adjustment in the invasive brine shrimp Artemia franciscana
Chapter 2: Discrimination against heterospecific and non-receptive females during precopulatory mate guarding in Artemia
Discussion & perspectives
Section 2: Parasite specialization
Chapter 3: Cryptic and overt host specificity shapes the epidemiology of two sympatric microsporidian species
Chapter 4: Infectivity, virulence and transmission in a two-host, two-parasite system
Chapter 5: Specialized host exploitation constrains the evolution of generalism in microsporidian parasites of Artemia
Discussion & perspectives


Related Posts