Several parasites induce phenotypic alterations in their hosts, which can enhance the parasites’ fitness at the expense of the hosts (see references in Fayard et al., 2020). These parasite-induced alterations can affect various traits, including physiology, reproduction, and appearance (Poulin, 1995; Moore, 2002; Thomas et al., 2005; Cézilly and Perrot-Minnot, 2010). Among the parasites in coastal ecosystems, digenean trematodes are highly abundant and often utilize mollusks as their first and second intermediate hosts (Bush et al., 1997; Poulin, 1998; Kuris et al., 2008). In first intermediate hosts, trematode infections can lead to reduced reproductive output or complete reproductive impairment (Sorensen and Minchella, 2001; Hall et al., 2007; Granovitch et al., 2009). Hosts, in response to the fitness loss, may evolve non-immunological defenses such as fecundity compensation, by adjusting their reproductive traits (Agnew et al., 2000; Hurd, 2009; Parker et al., 2011). However, the impact of trematodes on the reproduction of second intermediate hosts, as well as the existence of this type of non-immunological defenses, remains understudied (but see Fostera, 1958; Serbina, 2014; Parietti et al., 2020). The mechanisms underlying fecundity compensation are not fully understood, but previous research suggests that non-immunological defenses against parasites involve energetic, hormonal, and physiological pathways that involve a molecular dialogue between hosts and parasites (Hall et al., 2007; Pike et al., 2019; Gomez de Leon et al., 2021). In this study, we examined the histology, biochemistry, and sex hormone levels in the female gonads of Fissurella crassa, a limpet species parasitized by the digenean Proctoeces humboldti in the metacercaria stage. This host-parasite system is one of the most extensively studied systems on the eastern Pacific coast, in which previous research has demonstrated greater gonadal biomass of parasitized limpets compared to non-parasitized ones (Aldana et al., 2014; Aldana et al., 2020; García-Huidobro et al., 2020).

Mollusk-trematode interactions have been used as a model to investigate energy allocation strategies in parasitized organisms (Sandland and Minchella, 2003). When hosts are unable to resist infection through immune defense, they may compensate for the damage caused by parasites, adjusting their life history traits (Lefévre et al., 2008). Fecundity compensation and gigantism are two strategies that involve changes in life history traits and have been widely studied in mollusks serving as the first intermediate hosts for trematodes (Sorensen and Minchella, 2001; Ebert et al., 2004). Fecundity compensation involves a greater allocation of resources to reproduction or an earlier sexual maturity of adult or juvenile host individuals, respectively (Minchella and LoVerde, 1981; Thornhill et al., 1986). However, the recognition of these phenotypic alterations in second intermediate hosts is unclear, as the nature and outcome of host-parasite interactions may depend on species, life cycles, and contextual factors (Aldana et al., 2020; Liu and Gaines, 2022). From the parasite’s perspective, fecundity compensation benefits mechanisms that enhance their own fitness, such as prolonging its life within the host or increasing its transmission rates (Schmid-Hempel, 2021). Compensatory host responses to parasitism, including changes in fitness-related traits like survival, growth, fecundity, and competitiveness, can align with the transmission routes of parasites (Lefévre et al., 2008). Parasites can exploit selected compensatory responses in other ecological contexts, mimicking the causes that induce them and thus becoming the causal agents of the host’s compensatory response (Thomas et al., 2012).

In the coastal regions of Chile and Peru, the interaction between the fellodistomid trematode Protoeces humboldti George-Nascimento and Quiroga, 1983, and its second intermediate host, the keyhole limpets Fissurella spp. (Vetigastropoda), has received considerable attention (see Aldana et al., 2009; Aldana et al., 2014). The Fissurella limpets comprise a diverse group of 13 herbivorous species inhabiting intertidal and subtidal zones (McLean, 1984; Oliva and Castilla, 1992). The different species of Fissurella are dioecious and do not present evidence of sexual reversal or hermaphroditism (but see Olivares et al., 2009), they lack external sexual dimorphism and the sex is recognized by the color of the gonad, pale yellow to white in males and green in females (after dissection of the limpet) (Bretos, 1978; Bretos et al., 1983; Bretos et al., 1988; Bretos and Chihuailaf, 1990).

Proctoeces humboldti has a three-host life cycle (Oliva, 1984; Oliva and Zegers, 1988; George-Nascimento et al., 1998; Balboa et al., 2001; Oliva et al., 2018). In mussels, its first intermediate host, the parasite produces cercariae that then infect the gonad of keyhole limpets Fissurella spp., its second intermediate host, in which they develop into metacercariae. Infected limpets are subsequently ingested by the definitive host, the clingfish Syciases sanguineus, where the parasite reaches sexual maturity. In Fissurella species in general and in F. crassa in particular, the non-encysted metacercaria of P. humboldti is only found in the gonad, and is considered a progenetic metacercariae with precocious development that would facilitate the transmission of the parasite in intertidal conditions where the probability of consumption by S. sanguineus is low (Oliva and Huaquín, 2000; Balboa et al., 2001; Aldana, 2007; Oliva and Álvarez, 2011; Valdivia et al., 2014).

Previous studies on Proctoeces humboldti have suggested a positive effect of the parasite on its host Fissurella crassa (Aldana et al., 2014; Aldana et al., 2020; García-Huidobro et al., 2020). Parasitized individuals show larger shell length, muscle foot biomass, and gonadosomatic index (GSI) compared to non-parasitized individuals (Aldana et al., 2014). The increased body size of parasitized limpets is maintained across three localities in the southern Pacific of Chile, separated by approximately 1500 km (García-Huidobro et al., 2020). Moreover, particularly in areas of high biological productivity, parasitized individuals exhibited lower oxygen consumption than non-parasitized ones, suggesting lower metabolic demand in infected individuals (Aldana et al., 2020). Despite the positive impact of P. humboldti on body size and reproduction (GSI) in F. crassa, there is still a knowledge gap regarding the reproductive and biochemical aspects of parasitized limpets that allow this phenomenon to be described as the fecundity compensation. In addition, few examples exist of fecundity compensation in mollusks serving as second intermediate hosts for trematodes, and no studies have explored the underlying mechanisms of this phenomenon (but see Fostera, 1958; Serbina, 2014). In this study, we compared the histology, biochemistry, and sex hormone levels in the female gonads of F. crassa parasitized and non-parasitized by P. humboldti. Based on the positive effect of P. humboldti on the gonadosomatic index of F. crassa (Aldana et al., 2014; Aldana et al., 2020; García-Huidobro et al., 2020), our hypothesis is that the size and density of oocytes, the amount of glucose, proteins, and lipids, and the levels of estradiol and progesterone will be higher in parasitized individuals than in non-parasitized individuals, suggesting a potential mechanism of fecundity compensation in this host-parasite system.

Our results indicate that 59.5% of F. crassa individuals harbored between 1 and 32 individuals of the P. humboldti parasite. The gonadosomatic index of limpets increases from two to four growth rings (Figure 1). However, independent of age and consistent with previous reports (Aldana et al., 2014; Aldana et al., 2020; García-Huidobro et al., 2020), parasitized limpets show a higher GSI than non-parasitized limpets (Figure 1, Table 1). The log-normal fit regression analysis (GLM) shows that the GSI does not vary significantly with increasing parasite:host biomass ratio (χ²_{(1, 23)}= 0.15, P= 0.696, R²_pseudo= 0.03. Figure 2).

Gonadal histology of parasitized and non-parasitized F. crassa

In general, histological characterization of the limpet gonad reveals an epithelium covering the ovary with cubic cells and underlying connective tissue. A septum, composed of muscular fibers and other fibrillar elements, extends into the gonad (Figure 3A). Most oocytes display a nucleus and a gelatinous layer that covers the oocyte (Figure 3B). However, parasitized gonads, show tissue damage around the oocytes and a collagen-like fibrillar element around the parasite; while more distant from the site of parasite insertion, near the periphery of the gonad, normal tissue with epithelium and connective tissue is observed (Figures 3C, D).

Quantitatively, although the jelly coat area was similar between parasitized and non-parasitized gonads (Table 2, Figure 4A), the area and density of oocytes presented significant differences with opposite trends, associated with the parasitism condition. While oocyte area was greater in non-parasitized compared to parasitized gonads (Table 2, Figure 4B), oocyte density was greater in parasitized than in non-parasitized gonads (Table 2, Figure 4C). These results considered the effect of host length on the variation of each response variable (Table 2, Figure 4).

Glucose, lipids, and proteins content in gonads of parasitized and non-parasitized F. crassa

While the amount of glucose was greater in non-parasitized than in parasitized gonads (Table 2, Figure 5A), parasitized limpets presented around 44% more lipids in their gonad than non-parasitized limpets (Table 2, Figure 5B). The amount of proteins, on the other hand, was similar between gonads of different infection conditions (Table 2, Figure 5C). These results also considered the effect of host length on the variation of each response variable (Table 2, Figure 5).

Quantification of sex steroids in gonads of parasitized and non-parasitized F. crassa

Neither host length nor parasitism condition were significantly important on the variation in hormone levels (estradiol and progesterone, Table 2, Figures 6A, B). However, the level progesterone increased significantly with parasite intensity (Estradiol: F_{(1, 8)} = 0.64, R² = 0.07, P = 0.445. Progesterone: F_{(1, 6)} = 6.74, R² = 0.52, P = 0.040. Figures 6C, D).

The increase in the gonadosomatic index (GSI) associated with the presence of P. humboldti (Figure 1) is consistent with previous reports for F. crassa from the central zone of Chile (Aldana et al., 2014; Aldana et al., 2020; García-Huidobro et al., 2020). In the present study we also show that parasitized limpets have gonads with a higher oocyte density (with a smaller area, Figures 4B, C), lower amount of glucose (Figure 5A), and greater amount of lipids (Figure 5B) than non-parasitized limpets. Additionally, a positive association between the number of parasites and the amount of progesterone in gonadal tissue was observed (Figure 6D). A host may respond to the presence of some trematode infections by varying its investment into current reproduction, in quantity and/or quality of offspring (Duffield et al., 2017), to make up for lost future reproduction (Minchella and LoVerde, 1981; Minchella, 1985; Lafferty, 1993; Krist, 2001). This fecundity compensation could be associated with instances where the host cannot effectively prevent the establishment of an infection or immunologically suppress it (Parker et al., 2011) and has been widely reported in invertebrate first intermediate hosts of trematodes (Minchella and LoVerde, 1981; Thornhill et al., 1986; Lafferty, 1993; Krist, 2001). Consequently, infected hosts should attempt to achieve maximum reproductive success before the cumulative costs of infection become detrimental to their fitness (Agnew et al., 2000). In this sense, our results suggest that second intermediate host F. crassa can also compensate for the effect of P. humboldti metacercaria, increasing the GSI, mediated by biochemical and hormonal pathways.

Data on the effects of trematodes on the reproduction of their second mollusk intermediate hosts are scarce and isolated. The response of the snail Bithynia troscheli to the presence of trematode metacercariae constitutes, to our knowledge, the first evidence of fecundity compensation in a second intermediate host (Serbina, 2014). The increase in the percentage of reproductively active host females, at low infection levels, was considered as a mechanism that could compensate for the subsequent reduction in snail fecundity, as a result of the increase in trematode intensity (Serbina, 2014). Studies carried out on the shrimp Palaemonetes argentinus, the second intermediate host for the digenean Microphalus szidati, demonstrated that parasitized females produced more embryos, but had more egg loss throughout embryonic development than non-parasitized females (Merlo et al., 2016; Parietti et al., 2020). In this sense, female shrimp should produce a greater number of offspring at the early season to compensate for egg losses as a result of being parasitized. Similarly, the higher density (Figure 4C) and smaller area (Figure 4B) of oocytes in parasitized F. crassa, i.e. greater number of oocytes, would also be a compensatory response for the posterior reduction in fecundity due to the parasite.

In a study carried out on limpets F. crassa from Perú, Oliva (1992) reported structural and functional disorganization of the gonadal tissue in parasitized individuals by Proctoeces humboldti, preventing sexual identification of the gonad and inhibiting the gametogenesis process in advanced infections. However, the gonad of specimens collected from the northern and central-southern areas of Chile does not present evidence of damage of this type associated with P. humboldti, and even in particular conditions, parasitized limpets present more gonad with respect to the body (GSI) than those that are not parasitized (Aldana et al., 2014; Aldana et al., 2020; García-Huidobro et al., 2020). On the other hand, induced spawning in F. crassa from northern Chile showed that the fecundity of limpets, estimated as the number of eggs/mm of shell, decreased significantly with increasing parasite intensity of P. humboldti (Oliva and Vega, 1994). In this sense, the increase in the IGS and ultimately, the number of parasitized F. crassa oocytes could compensate for the subsequent reduction in the number of spawned eggs. Although it is unknown whether this decrease in spawning is a widespread phenomenon in Chile, its association with the increase in parasite intensity, which according to our results is positively related to the level of progesterone (Figure 6D), suggests a compensatory response of parasitized F. crassa against the oocyte losses during spawning. In this sense, increased fecundity would reflect a trade-off between the number and size of eggs (Einum and Fleming, 2000; Rollingson and Hutchings, 2013), with eggs from infected females having a smaller area, as has been observed in other host-parasite systems (Heins, 2012; Parietti et al., 2020).

Previous studies have shown that parasitized limpets F. crassa have more parasite intensity in sites where the GSI is also higher, suggesting that the increase in gonadal biomass could provide more space for the accumulation of metacercariae (Aldana et al., 2014; Aldana et al., 2020; García-Huidobro et al., 2020). In fact, our results show that the proportion of F. crassa gonad (per unit of body size) occupied by P. humboldti increases, although the gonadal mass/body mass ratio of the limpet (GSI) remains relatively stable (Figure 2). Shell alterations of gastropods induced by parasite trematodes have been reported in several systems as a strategy that would allow the parasites to increase the volume within the host and provide more space for the accumulation of greater numbers of individuals (McCarthy et al., 2004; Hay et al., 2005; Levri et al., 2005; Lagrue et al., 2007; Thieltges, 2009). For example, changes in the shell shape of the periwinkle Littorina saxatilis caused by the trematode Microphallus piriformes resulted in a 12% increase in the total volume inside the shell (McCarthy et al., 2004). The resulting increase in space is particularly advantageous for trematodes with abbreviated life-cycles, allowing parasite growth and reproduction without further reduction of host tissue. Consistently, it has been proposed that P. humboldti metacercariae would abbreviate their life cycle through progenesis, developing precociously into an adult due to the low probability of consumption by the definitive host (Oliva and Huaquín, 2000; Balboa et al., 2001; Aldana, 2007; Oliva and Álvarez, 2011). However, regardless of transmission opportunities, the long residence time of P. humboldti in Fissurella (George-Nascimento et al., 1998; Balboa et al., 2001; Aldana, 2007), would promote precocious maturity and egg production by the parasite in response to some internal developmental clock (Poulin & Cribb, 2002; Lagrue and Poulin, 2009).

Until now, the greater fecundity of parasitized limpets compared to non-parasitized limpets of the same age had been attributed to the increase in body size of F. crassa associated with P. humboldti (Aldana et al., 2014; Aldana et al., 2020). Here, we present evidence that the increase in reproductive performance of parasitized limpets could also be mediated by an increase in oocyte density or, equivalently, gonadal biomass per unit volume. That is, assuming that a parasitized gonad (minus the parasites) and a non-parasitized one occupy the same space/volume, the parasitized gonad will have a greater biomass. If this effect of P. humboldti on the oocyte density of F. crassa is coupled to the increase in body size and therefore the greater gonadal space/volume of limpets, an even greater reproductive performance of parasitized F. crassa is expected. From the host’s perspective, however, the reproductive performance evaluated through the GSI shows astonishing results. Firstly, it is greater in parasitized than in non-parasitized limpets; and secondly, it remains relatively stable as the number of parasites increases, instead of decreasing as would be expected given the energetic costs of parasite maintenance (Price, 1980; Poulin, 2007; Goater et al., 2014) and expression of host phenotypic alterations (Maure et al., 2013). In this context, parasitized F. crassa limpets appear to be showing greater reproductive effort than their non-parasitized counterparts. However, this capacity, as well as a lower energy demand associated with infection, occur in highly productive sites close to upwelling centers such as Quintay where this study was carried out (Aldana et al., 2020; García-Huidobro et al., 2020). The availability and quality of food in these environments would benefit the energy budget of F. crassa (Pulgar et al., 2013) and attenuate the conflict for resources with P. humboldti, promoting the welfare of the parasite-host system.

In general, elevated levels of glycogen and/or lipids have been identified in several host-helminth systems (Amat et al., 1991; Jokela et al., 1993; Plaistow et al., 2001; Ponton et al., 2005; Korkofigas et al., 2016) associated with the numerous energetic demands experienced. Likewise, the gonad of parasitized F. crassa presented a higher level of lipids, but a lower level of glucose (Figures 5A, B). Particularly, non-immunological defenses such as the compensatory responses in fecundity described for F. crassa, are costly to activate and maintain (Parker et al., 2011) and the lower glucose content in infected gonad supports this proposal. Similarly, Gammarus insensibilis amphipods that are more active when infected with metacercariae of the trematode Microphallus papillorobustus, had a lower glucose level than the uninfected ones (Ponton et al., 2005). In addition, the higher lipid concentration in parasitized limpets compared to non-parasitized ones may be related to the increase in oocyte density, as lipids are considered to be the nutritional source for gonad development (e.g., Soudant et al., 1996; Ke and Li, 2013). On the other hand, alterations in protein levels in gastropods during trematode infection suggest that their occurrence is associated with the identity of the host-parasite systems, food availability, and environmental conditions (Thompson, 1983). Our findings did not reveal any difference in protein quantity between parasitized and non-parasitized gonads (Figure 5C), suggesting that specific enzymatic processes may be functioning adequately even in parasitized individuals. Unfortunately, unlike previous studies (e.g., Jones, 1989; Santarém et al., 1994), we did not assess protein, amino acid and/or lipids composition in relation to parasitism condition. Information on the structure and composition of macromolecules could be useful when comparing parasitized and non-parasitized individuals.

Knowledge of the endocrine system’s role in the maturation of sex organs and the induction of sexual behavior is relatively limited in invertebrates. Although the information obtained from different species strongly suggests the involvement of vertebrate-type steroids in the control of mollusk reproduction (Lafont and Mathieu, 2007; Fernandes et al., 2011), its endogenic biosynthesis and its role in endocrine regulation has been questioned (Scott, 2012; Scott, 2013; Scott, 2018). Recent studies on possible pathways for sex steroid synthesis and regulation, have provided evidence for the hypothesis of endogenous production of sex steroids and gonadal maturity regulation by invertebrates (Thitiphuree et al., 2018; Zhang et al., 2020). Thus, considering that there is evidence against but also in support of the involvement of steroidal factors in the reproduction of mollusks (see review in Pandian, 2017), it is appropriate to evaluate the relationship between these factors and parasitism by P. humboldti, which influences the GSI of F. crassa. In this regard, although estradiol and progesterone levels showed a similar magnitude between parasitized and non-parasitized gonads (Figures 6A, B), progesterone levels increased with parasite intensity (Figure 6D). Fluctuations in testosterone, estradiol and progesterone levels in the hemolymph of F. cumingi limpets were related to the reproductive state, with a greater presence of estradiol and testosterone associated with periods of vitellogenesis and gonadal renewal (increase in GSI), while a greater progesterone level during spawning events or lower reproductivity activity (Honour, 2002). Similarly, the ovarian concentration of progesterone in F. crassa also presented variations around the year, with a peak prior to the period of maximum IGS and another peak after a secondary increased IGS (Del Campo et al., 2005). Thus, the increase in steroid hormones, including estradiol and progesterone, may explain the increase in the number of oocytes in infected limpets, although only progesterone increased with parasite intensity.

Parasite-mediated host phenotypic alterations have traditionally been examined as host adaptations that reduce the harmful consequences of infection, parasite adaptations that facilitate its transmission, or secondary outcomes of infection without adaptive value (Minchella, 1985; Moore, 2002; Thomas et al., 2005). However, from an evolutionary perspective, host phenotypic changes following infection could be trade-offs between host and parasite strategies, where natural selection should favor parasites that are able to initiate compensatory host responses that benefit parasite transmission and host fitness (Lefévre et al., 2008). In this scenario, such interactions could move a continuum between parasitism and mutualism where the environmental context would have a key role (Leung and Poulin, 2008; Fellous and Salvaudon, 2009). In the interaction F. crassa-P. humboldti, fecundity compensation could be a mechanism used by both the host and the parasite (Lefévre et al., 2008), which would not only mitigate the negative impact if the parasite on the host’s spawning, but also provide greater space to the progenetic metacercariae. The release of P. humboldti eggs into the environment could be synchronized with host spawning. Empirical data from different systems are required to elucidate the mechanistic processes behind these phenomena, as well as an integrated consideration from different disciplines (e.g. parasitology, physiology, biochemistry and ecology), in the performance of closely interacting species.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was approved by Daniel Valenzuela Universidad Santo Tomás. The study was conducted in accordance with the local legislation and institutional requirements.

MG-H: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Supervision, Visualization, Writing – original draft, Writing – review & editing. MR: Investigation, Methodology, Writing – original draft. NF: Investigation, Methodology, Resources, Writing – review & editing. TB: Investigation, Methodology, Resources, Writing – review & editing. FG-R: Investigation, Methodology, Writing – original draft. ÁU: Investigation, Methodology, Resources, Writing – review & editing. JP: Conceptualization, Funding acquisition, Investigation, Resources, Writing – original draft, Writing – review & editing. MA: Conceptualization, Funding acquisition, Investigation, Methodology, Resources, Writing – original draft, Writing – review & editing.