Behaviour is one of the most critical components in the success of introduced species to become invasive (Holway and Suarez 1999; Chapple et al. 2012). More specifically, behavioural flexibility is thought to facilitate the invasion process, particularly during its early stages (Wright et al. 2010). In the last two decades, numerous studies have documented the importance of behavioural traits in the invasion success of both invertebrates and vertebrates in both terrestrial and aquatic ecosystems; see Chapple et al. (2012) and Chapple and Wong (2016) for reviews. In aquatic ecosystems, crustaceans are considered as one of the most successful groups of invasive species (Hänfling et al. 2011), and both their feeding behaviour and the nature of the subsequent predator-prey interactions appear particularly relevant to their success (Weis 2010, 2016). The predatory behaviour of invasive predators is well documented (Sih et al. 2010a; Grosholz and Wells 2016), and this holds for crustaceans (Weis 2010, 2016). In contrast, the role behaviour may have in mediating the success of invasive prey species is still a relatively untapped area of research, despite evidence that the role of behaviour in the success of invasive prey may be more important than the behaviour of native predators (Pintor and Byers 2015; Mennen and Laskowski 2018; Papacostas and Freestone 2019).

Based on the lack of co-evolution history between introduced species (either predator or prey) and their native counterparts, a range of evolutionary hypotheses has been introduced to describe novel predator-prey interactions; see e.g. Grosholz and Wells (2016) for a review. These hypotheses are essentially based on the dichotomy between the preference of both native and non-native predators towards evolutionary familiar or unfamiliar prey. Native predators may consume more of a native prey species than an invasive one, because they fail to recognize an invader (Enemy Release Hypothesis; Torchin and Mitchell 2004) and/or they are unable to overcome their defense (Novel Defense Hypothesis; Elton 1958) or because of ineffective antipredator behaviour by naive prey (Naive Prey Hypothesis; Sih et al. 2010a). Invasive predators may also favour evolutionarily familiar invasive prey species over native or unfamiliar invasive prey species (Facilitation Between Invaders Hypothesis; Grosholz and Wells 2016). Alternatively, native predators may prefer evolutionary novel non-native prey over familiar ones (Biotic Resistance Hypothesis; Elton 1958), and invasive predators may consume more of an evolutionarily novel prey (either native of from another non-native region) than evolutionary familiar ones, hence be at competitive advantage (Novel Weapons Hypothesis; Callaway and Ridenour 2004). Other hypotheses related to the Optimal Foraging Theory (Pyke 1984) and Optimal Diet Theory (Sih and Christensen 2001) respectively state that a predator consumes the prey that is the easiest to capture and the most profitable, regardless of evolutionary or historical familiarity.

The nature of the abovementioned predator-prey interactions is, however, noticeably both species- and location-dependent, and may also vary along the invasion process due to a potential loss of naiveté (Grosholz and Wells 2016). It is, however, essential to the understanding of the impact of invasions on ecosystem structure and function to examine the behaviour of invasive species in their different introduction areas, since behavioural flexibility may lead a given species to express different behavioural variants in distinct ecological conditions (Wright et al. 2010). In this context, the present work takes advantage of the recently documented (Spilmont et al. 2018) co-occurrence of the Asian shore crab Hemigrapsus sanguineus and the North American amphipod Ptilohyale littoralis along the French coast of the eastern English Channel. Studies on predation and prey selection by the Asian shore crab showed that this species is omnivorous, though it exhibits a preference towards animal prey (Brousseau and Baglivo 2005). In particular, juveniles of the blue mussel Mytilus edulis are actively selected (Bouwmeester et al. 2020), including over other invertebrates both in field and laboratory experiments (Lohrer and Whitlatch 2002). In addition, amphipods are known to be a recurrent component of the diet of H. sanguineus (McDermott 1999; Lohrer et al. 2000; Blasi and O’Connor 2016). The present study assesses, based on experiments run under controlled laboratory conditions, (i) if Hemigrapsus sanguineus is actually able to prey on Ptilohyale littoralis from the same introduced area, and (ii) if H. sanguineus exhibits a preference between P. littoralis and the blue mussel Mytilus edulis, known as one of its preferred prey.

In no-choice experiments (Stage 1; Fig. 1A), male H. sanguineus exhibited a higher consumption rate on M. edulis (prey consumed per individual: 3.7 ± 1.3 prey ind^-1; mean ± standard deviation) than on P. littoralis (2.1 ± 1.5 prey ind^-1; WMW test: P = 0.004). The reverse situation was observed for females (WMW test: P = 2.1 10^-4) with a higher consumption of amphipods (2.5 ± 1.6 prey ind^-1) than mussels (0.3 ± 0.8 prey ind^-1).

The consumption rate of M. edulis significantly differed between male and female H. sanguineus (WMW test: P = 4.4 10^-6). Specifically, twelve females did not prey on M. edulis and the maximum number of mussels eaten by a single individual was 3. In sharp contrast, all males consumed at least one mussel and five of them did prey on the 5 individuals presented. The consumption rate of P. littoralis was non-significantly different between males and females (WMW test: P = 0.445).

The results of the choice experiments (Stage 2; Fig. 1B) showed that the consumption rate for each type of prey did not significantly differ from the no-choice experiment (WMW tests, P > 0.148). In addition, the lack of significant difference between the expected and observed consumption rates indicates the absence of preference strictosensu for any type of prey (Table 2), either for males (Chi-square test: χ²₁ = 0.76, P = 0.385) or for females (Chi-square test: χ²₁ = 0.01, P = 0.904). Thirteen out of the 15 individual χ² tests comparisons for males showed no significant differences, and all 15 tests were not significant for females (Table 2); this confirmed that the observed consumption rates were not influenced by inter-individual variability. These results were further supported by those from the complementary choice experiment (Stage 2bis, i.e. prey presented in unbalanced proportions, Fig. 2) which showed than neither M. edulis nor P. littoralis was systematically consumed when a single individual was offered (none of the prey was systematically consumed first), i.e. none of the prey is preferred strictosensu following Taplin (2007).

The present study provides evidence, under laboratory-controlled conditions, for the existence of predator-prey interactions between two introduced crustaceans, the Asian shore crab Hemigrapsus sanguineus and the North American amphipod Ptilohyale littoralis, co-occurring outside their distinct native ranges in the intertidal zone of the eastern English Channel. Specifically, in no-choice experiments, both male and female H. sanguineus actively selected both M. edulis and P. littoralis, with females exhibiting a significantly lower consumption rate (i) on M. edulis than males, but not on P. littoralis, and (ii) on the native M. edulis than on the non-native P. littoralis. These results are consistent with previous studies showing a stronger predation pressure of males H. sanguineus on M. edulis than females due to differences in claw morphology and strength (Brousseau et al. 2001; Bourdeau and O’Connor 2003), and similar predation pressure on the amphipod Hyale plumulosa from male and female H. sanguineus (Blasi and O’Connor, 2016). Our observations are consistent with studies demonstrating the ability of crabs of the genus Hemigrapsus to feed on amphipods in laboratory experiments (McDermott 1999; Griffen and Byers 2006; Cornelius et al. 2021; Bleile and Thieltges 2021), with males exerting a higher predation pressure on mussels than amphipods (Cornelius et al. 2021; Bleile and Thieltges 2021), contrary to females (Cornelius et al. 2021).

In two-prey choice experiments, H. sanguineus consistently did not exhibit any preference strictosensu for either native or non-native prey though significant differences in the consumption rates of both types of prey were observed. This result seems to diverge from previous evidence that the Asian shore crab preferred mussels over other animal prey (Tyrrell and Harris 2000; Brousseau and Baglivo 2005), including in experiments with amphipods (Bleile and Thieltges 2021). However, the outcome from Bleile and Thieltges (2021) regarding preference are not directly comparable to the present study due to experiments restricted to males, choice experiments conducted with 4 types of prey and differences in the definition of “preference” itself and the subsequent data analysis. We, however, showed that male H. sanguineus actually consumed more M. edulis individuals than P. littoralis, which is consistent with most studies defining preference as a higher consumption rate. Terminological considerations aside, this recurrent observation in laboratory experiments on H. sanguineus is of prime importance regarding its ecological impact. Though not always preferred strictosensu, juvenile mussels may constitute a large part of the crab’s diet which may ultimately have a critical influence on the composition of natural benthic intertidal communities relying on M. edulis as engineer species.

We, however, unambiguously showed that H. sanguineus is able to detect and consume P. littoralis although the latter was putatively introduced a decade later (i.e. circa 2016; Spilmont et al. 2018) than the former (i.e. circa 2005; Dauvin et al. 2009) at our study site. This suggests that the Asian shore crab cannot be considered as a naive predator when confronted to a newly introduced amphipod prey. This apparent lack of naiveté is consistent with the ability of H. sanguineus to rapidly sense, pursue and consume different amphipod preys (McDermott 1999). Despite a lack of co-evolutionary history and historical familiarity between H. sanguineus and P. littoralis in their introduced eastern English Channel range, handling skills learned by H. sanguineus either in their native or non-native ranges in the presence of similar prey (e.g. Apohyale prevostii; Spilmont et al. 2018) could have been transferred to novel prey, as shown in the shore crab Carcinus maenas (Hughes and O’Brien 2001). Note that the indigenous M. edulis was not considered as a naive prey towards H. sanguineus; indeed, the present experiments were conducted about 10 years after the introduction of H. sanguineus at our study site and it can be considered that M. edulis is now able to detect the presence of the Asian shore crab’s cues, as recently demonstrated by Uguen et al. (2022) for mussels and H. sanguineus from the same study site.

Since H. sanguineus did not exhibit any preference strictosensu for either the indigenous (M. edulis) or the introduced (P. littoralis) prey, none of the evolutionary hypotheses of novel prey-predator interactions (see Grosholz and Wells 2016) applies to H. sanguineus in its introduced range along the French coast of the eastern English Channel. Instead, following Grosholz and Wells (2016), H. sanguineus would use a combination of optimal foraging strategy and optimal diet strategy and thus consume the prey that is the easiest to catch and the most profitable. However, the optimal diet theory is usually not supported when predators attack motile prey (Sih and Christensen 2001). Indeed, consuming a motile small amphipod would unlikely maximize the net energy gain per unit time foraging compared to the consumption of the larger non- or slow-moving mussel. It is actually much more probable that H. sanguineus is an opportunistic predator and that its consumption of P. littoralis mainly depends on their encounter rate. This hypothesis in consistent with the fact that invasive prey usually exhibit various degrees of activity, sheltering and exploratory behaviours (Chapple and Wong 2016) which, when the species is considered as a prey, would either mitigate or enhance the predation pressure. This was demonstrated in studies with the invasive freshwater amphipod Dikerogammarus villosus exhibiting less exploratory behaviour and greater sheltering than native species, which made the former less vulnerable to fish predation (Kobak et al. 2014; Mennen and Laskowski 2018). In our case, the degree of activity of P. littoralis could be estimated only by complementary experiments with continuous monitoring.

The fact that a motile prey such as P. littoralis was consumed in high proportion, especially by female H. sanguineus, implicitly suggests, that P. littoralis either did not exhibit any, or expressed inappropriate/uneffective, antipredator response, which would support the ‘naive prey’ hypothesis (Banks and Dickman 2007; Sih et al. 2010b). Though, to our knowledge, anti-predator responses have never been investigated in Hyalid amphipods, this has been demonstrated in other families (e.g. Mennen and Laskowski 2018). The consumption rates we recorded on P. littoralis are consistent with those observed by Blasi and O’Connor (2016) – note that Hyale plumulosa’s accepted name is actually Ptilohyale plumulosa (WoRMS 2020), which has been categorized as P. littoralis for the Atlantic ; see pp. 102–103 in Bousfield and Hendrycks 2002 and p. 51 in Lo Brutto and Iaciofano 2018 – in experimental enclosures without any shelter (consumption of ca. 60% of prey available). Our estimates are, however, clearly higher than the consumption rates they recorded with a refuge (sediment and rocks; consumption of ca. 15%). Despite differences in the experimental setup, these discrepancies likely reflect different behaviours and/or antipredator responses of P. littoralis between native and invaded ranges. For the eastern English Channel area, similar laboratory observations using native amphipods, such as the sibling species Apohyale prevostii or the MelitidaeMelita palmata would help to more thoroughly elucidate interactions between H. sanguineus and amphipods by comparing native and introduced prey.

Our results are relevant in an experimental context and, though controlled experiments remain a stepping stone in the understanding of prey-predator interactions, the extrapolation of laboratory observations to field events remains uncertain (Brousseau and Baglivo 2005). Our experiments represented idealized foraging conditions due to the small size of the experimental container that minimized search time of the predatory crab and overestimate ingestion under natural conditions (Lohrer and Whitlach 2002; Epifanio 2013). They were also conducted at night when H. sanguineus is known to be more actively feeding (Spilmont et al. 2015) and without sediment in the containers which added to the optimal foraging conditions (Blasi and O’Connor 2016). Feeding patterns of the Asian shore crab on mussels and amphipods have been shown to be different in the field and in the laboratory (Brousseau et al. 2014; Blasi and O’Connor 2016); possible explanations include limited foraging opportunities due to abiotic factors (diurnal cycle, wave action, etc), increased prey choice and variable densities.

Though being part of the diet of the Asian shore crab in both its native and north-American invaded ranges, amphipods usually constitute only a small proportion of the overall food use (Lohrer et al. 2000; Griffen et al. 2008). However, a potential impact of the sibling H. takanoi on natural amphipod populations has recently been underlined (Cornelius et al. 2021) and our results suggest that the predatory pressure on P. littoralis would mainly be exerted by female H. sanguineus when several prey are available. At our study site, H. sanguineus was found concomitantly with P. littoralis, but also with a range of potential other prey (e.g. the amphipods Apohyale prevostii and Melita palmata, the annelid Eulalia clavigera and the gastropod Littorina littorea), as well as the potentially competitive green crab Carcinus maenas (Spilmont et al. 2018). Furthermore, the Asian shore crab is itself a potential prey for fish (Heinonen and Auster 2012; Savaria and O’Connor 2013), birds (Dumoulin and van Outryve 2009) and other crab species (Papacostas and Freestone 2019), against which H. sanguineus could also behave as a naive prey. The potential benefit for the crab to forage on naive prey could therefore be dampened by its own naiveté (Papacostas and Freestone 2019). As long as these abiotic factors and consumptive and non-consumptive biotic interactions are not taken into account, neither the ecological role of H. sanguineus, which as a large benthic decapod could play a role in structuring communities (Boudreau and Worm 2012), nor the potential synergistic interaction between the two introduce species, which could accelerate impacts on native communities (‘invasional meltdown’: Simberloff and Von Holle 1999), can be elucidated.

This work is a contribution to the CPER research project CLIMIBIO. The authors thank the French Ministère de l’Enseignement Supérieur et de la Recherche, the Hauts de France Région and the European Funds for Regional Economical Development for their financial support for this project. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

N.S.: research conceptualization, sample design and methodology, investigation and data collection, data analysis and interpretation, original draft; writing

L.S.: research conceptualization, sample design and methodology, data interpretation, original draft; review & editing