Journal of Ethology

, Volume 29, Issue 1, pp 115–120

Can the parthenogenetic marbled crayfish Marmorkrebs compete with other crayfish species in fights?


  • Stephanie A. Jimenez
    • Department of BiologyThe University of Texas-Pan American
    • Department of BiologyThe University of Texas-Pan American

DOI: 10.1007/s10164-010-0232-2

Cite this article as:
Jimenez, S.A. & Faulkes, Z. J Ethol (2011) 29: 115. doi:10.1007/s10164-010-0232-2


The parthenogenetic marbled crayfish, Marmorkrebs, has no known wild population, but has been introduced into natural ecosystems in two continents. Interactions with native crayfish, particularly through fighting, could affect the ecological impact of Marmorkrebs introductions. Marmorkrebs have been characterized anecdotally as having low levels of aggression, which could mitigate their potential to compete with native species. We isolated Marmorkrebs and Louisiana red swamp crayfish (Procambarus clarkii), then conducted size-matched intra and interspecific pairings. Marmorkrebs were as likely to win a fight as P. clarkii, although contests between P. clarkii and Marmorkrebs began significantly faster than contests between two Marmorkrebs. These results suggest that Marmorkrebs have the potential to compete with other species on the same level as P. clarkii, which is itself a highly successful introduced species around the world.


AggressionCrayfishCompetitionInvasive speciesMarbled crayfishMarmorkrebsLouisiana red swamp crayfishProcambarus clarkii


Marmorkrebs are marbled crayfish of unknown origin and are the only known parthenogenetic decapod crustaceans (Vogt 2008). They were discovered in the German pet trade in the 1990s (Scholtz et al. 2003; Vogt et al. 2004), but information concerning their origin was deemed “totally confusing and unreliable” (Vogt et al. 2004). Marmorkrebs have no formal species name or description, although they belong to the American genus Procambarus (Scholtz et al. 2003; Braband et al. 2006). There is no known wild population, but Marmorkrebs have been introduced into natural ecosystems in at least four countries on two continents (Germany: Blanke and Schulz 2003; Marten et al. 2004; The Netherlands: Holdich and Pöckl 2007; Italy: Marzano et al. 2009; Madagascar: Jones et al. 2009; Kawai et al. 2009). Although the ecological outcomes of these introductions are still not clear, there have been persistent concerns about the potential for Marmorkrebs to be a pest species (Vogt et al. 2004), given that many other introduced crayfish species have become invasive (Gherardi and Holdich 1999) and the high reproductive potential that parthenogenesis permits. Jones et al. (2009) went so far as to call Marmorkrebs “the perfect invader”.

Competition between introduced and native species is one factor affecting whether an introduced species becomes a truly invasive species. In crayfish, some of this competition is mediated through aggressive behavior, which can confer direct benefits to the winner (Herberholz et al. 2007). Crayfish will fight different species as freely as their own (Guiaşu and Dunham 1999; Tierney et al. 2000; Klocker and Strayer 2004; Lynas et al. 2007; Hazlett et al. 2008), and dominance probably contributes to the success of some crayfish species as invaders (Back 1995; Tierney et al. 2000; Usio et al. 2001; Gherardi and Daniels 2004; Klocker and Strayer 2004; Lynas et al. 2007). Thus, understanding how Marmorkrebs interact with other crayfish can help in assessing the risk posed by accidental introductions. Vogt and colleagues wrote, “(N)othing is known to date about the interaction of the marbled crayfish with other crayfish species, invertebrates, or fish in nature. In the aquarium, this species is apparently less aggressive than other crayfish and prefers a hiding and conflict-avoiding strategy” (Vogt et al. 2004). Although this anecdote suggests Marmorkrebs have low aggression, Marmorkrebs do fight among each other, as evidenced by their use in a study of circadian rhythms and aggression (Farca Luna et al. 2009).

In this study, we pit Marmorkrebs both against each other and against Louisiana red swamp crayfish (Procambarus clarkii) to determine if Marmorkrebs can compete with other species of crayfish in aggressive interactions. An abstract of this work has been published (Jimenez and Faulkes 2009).

Materials and methods

Marmorkrebs, Procambarus sp., and Louisiana red swamp crayfish, Procambarus clarkii (Girard, 1852), were hatched and reared in the laboratory, so that animals of both species were approximately the same age and had similar social experience. The parents of the Marmorkrebs used in the experiment were from the existing laboratory colony described elsewhere (Jimenez and Faulkes 2010). The parents of the P. clarkii used were purchased from Carolina Biological Supply Company.

Procambarus clarkii of both sexes were used, because sex has not been shown to have a significant effect on fight outcomes or dominance status in juveniles (Bovbjerg 1956; Pavey and Fielder 1996; Figler et al. 1999). Animals were fed a mixed diet of commercial tropical fish foods, peas, carrots, and frozen bloodworms. Under these conditions, P. clarkii grew noticeably faster than Marmorkrebs. Fights were staged between intermolt juveniles about 8–10 months old, ranging from 10.16 to 22.23 mm carapace length (measured with a digital caliper).

Memory of past encounters affects subsequent encounters (Rubenstein and Hazlett 1974), so we isolated crayfish for 3 weeks before testing to remove memory of previous social encounters (Hemsworth et al. 2007). Similarly, crayfish observing interactions also affects later fights (Zulandt et al. 2008), so we placed crayfish in small (170 mm length × 90 mm width × 100 mm height) tanks with opaque dividers between them to remove visual cues.

The relative size of two animals affects fight outcomes (Bovbjerg 1956; Rubenstein and Hazlett 1974; Pavey and Fielder 1996), so we size-matched animals as closely as possible. The average difference between individuals was 2.08 mm. The juvenile P. clarkii had not begun to exhibit sexual dimorphism in claw size at the time of the experiments.

Animals were placed simultaneously in an aquarium (280 mm length × 170 mm width × 150 mm height). We recorded contests using a digital video camera and a personal computer for later analysis. The fight was considered to begin when animals were within one inch of each other, turned to face each other, and then engaged in physical interaction.

Because agonistic interactions between crayfish are ritualized (Rubenstein and Hazlett 1974), we measured fight intensity on a four point scale (Huber et al. 1997; Huber and Delago 1998):
  1. 1.

    no fighting;

  2. 2.

    threat displays with no claw contact;

  3. 3.

    claw lock, where both animals contested the encounter and at least one animal used its claws to grab its opponent; and

  4. 4.

    strike and rip, where both animals contested the encounter and at least one animal made unrestrained use of the claws.


Each fight was watched live to assess the maximum intensity reached. The recorded fights were also reviewed at least once to confirm the maximum intensity reached between the animals. Only the maximum intensity reached is included. The fight was concluded, and the trial stopped, when one individual consistently retreated from the other, usually by a tailflip. In total, there were 17 interspecific pairings between Marmorkrebs and P. clarkii, and 19 intraspecific pairings between two Marmorkrebs.

We hypothesized that differences in levels of aggression between P. clarkii and Marmorkrebs might be reflected in different dynamics of the fights. We hypothesized that “mismatched” fights containing one highly aggressive individual and one less aggressive individual (ostensibly Marmorkrebs in interspecific pairings; see “Introduction”) would:
  1. 1.

    be initiated faster than a fight containing two low aggression individuals, because which individual initiates a fight affects its probability of winning (Rubenstein and Hazlett 1974; Guiaşu and Dunham 1997);

  2. 2.

    determine a winner, and thus be concluded, in a shorter amount of time; and

  3. 3.

    be less intense fights, because the low aggression animal would be expected to retreat rather than escalate.


Statistical analyses were run on SPSS v. 12 for Windows. Chi-squared tests were used to test expected frequencies of fight outcomes, with the “expected” values set at 50%. Unpaired t tests were used to compare latencies to fight and fight durations. Mann–Whitney U tests were used to compare ranked data of fight intensity.


During fights between Procambarus clarkii and Marmorkrebs, there was no significant difference in which species initiated (n = 17, χ2 = 2.88, df = 1, P = 0.09) or won (n = 17, χ2 = 0.059, df = 1, P = 0.81) a fight, as shown in Fig. 1.
Fig. 1

Outcomes of interspecific contests between Marmorkrebs and P. clarkii. Dotted line shows proportion expected by chance

The dynamics of intra and intraspecific fights differed in one way. Interspecific fights between Marmorkrebs and P. clarkii were significantly faster to start than fights between two Marmorkrebs (t34 = 3.33, P = 0.0021), shown in Fig. 2. Once begun, however, there were no significant differences in the durations (t34 = 1.6424, P = 0.11) of inter and intraspecific fights, as shown in Fig. 2, or their maximum intensity (Mann–Whitney U test, Z = −0.073, P = 0.96), as shown in Fig. 3.
Fig. 2

Comparisons of fight dynamics in intraspecific contests between two Marmorkrebs and interspecific contests between Marmorkrebs and P. clarkii. a Latency to start of fight. b Durations of fights. Dot, mean; linedividing boxes, median; box, 50% of data; whiskers, 95% of data; asterisks, minimum and maximum values
Fig. 3

Histograms of maximum fight intensity occurring during contests. 1 no fight, 2 threat displays, 3 locking claws, 4 unrestrained claw use

Although we size-matched animals as closely as possible, perfectly matching them is difficult, so we examined whether these small size differences were correlated with fight dynamics. If animals are equally matched, there should be no significant difference in fight dynamics according to size. In contests between two Marmorkrebs, there was no difference in which animal initiated (n = 19, χ2 = 1.316, df = 1, P = 0.2513) or won (n = 19, χ2 = 1.316, df = 1, P = 0.2513) fights. Interspecific fights were not analyzed in this way because the P. clarkii individuals were always the larger animal of the pair (albeit only slightly), so size and species were confounded.


Juvenile Marmorkrebs can compete with Procambarus clarkii in aggressive interactions. Marmorkrebs win as many fights as P. clarkii, and there is no significant difference in which species initiates fights. The length and intensity of the ensuing fights are not significantly different for inter or intraspecific fights. The delay from the start of the trial to the start of the fight is shorter for fights between P. clarkii and Marmorkrebs than for fights between Marmorkrebs, which is the only evidence supporting the hypothesis that Marmorkrebs are less aggressive than P. clarkii. Chemical cues (i.e., scents) play a large role in signaling and individual recognition during fights (Zulandt Schneider et al. 1999, 2001; Breithaupt and Eger 2002; Bergman et al. 2005; Bergman and Moore 2005; Moore and Bergman 2005). We hypothesize that an individual sensing an unfamiliar scent dissimilar to its own may be more likely to investigate it, and perhaps initiate a fight. Two Marmorkrebs would probably release very similar chemical cues because of their genetic similarity (Martin et al. 2007), which may not be recognized as being from another individual. Scents released by P. clarkii would be expected to be distinct from Marmorkrebs, and from each other P. clarkii, and thus cause the shorter delay to the start of the fight.

Clearly, a fight in a bare tank between socially isolated juvenile animals is a much simpler situation than competition in a natural habitat. Crayfish aggression is context dependent, with the social history of the animals involved (Daws et al. 2002; Graham and Herberholz 2008), habitat complexity (Baird et al. 2006), prior residence in a shelter (Figler et al. 1999), and food availability (Stocker and Huber 2001), to name a few examples, playing roles. It may be that placing Marmorkrebs with other species in more complex environments may allow Marmorkrebs to use behavior to evade conflict rather than engage in it. Claw size may also play a larger factor in a complex natural setting than shown by these experiments. Large claw size is advantageous to winning fights for animals of similar sizes (Snedden 1990; Garvey et al. 1994; Rutherford et al. 1995), and claws often grow at disproportionate rates to the rest of the body (Schroeder and Huber 2001). Many crayfish are sexually dimorphic, and males have larger claws than females. Thus, even if two crayfish species are evenly matched as juveniles, as we show here, one species may gain a competitive advantage in aggressive encounters over time due to allometric claw growth. Marmorkrebs, being all female and with moderately sized claws, may be at a disadvantage in contests with species of equal body size but larger claws.

A similar situation may occur because of overall growth rate. Although we kept both species under the same conditions, Marmorkrebs lagged behind P. clarkii in growth somewhat. Because size is the major factor in determining contests, a slower growing species, for example Marmorkrebs, may be at a competitive disadvantage to faster-growing species over an individual’s lifetime, even if slow-growing individuals can win fights with size-matched individuals (Vorgurger and Ribi 1999). Nevertheless, fecundity is a major factor contributing to the success of introduced crayfish (Lindqvist and Huner 1999), so the huge boost to fecundity provided by parthenogenesis in Marmorkrebs may be large enough to swamp such disadvantages.

While it is not possible to stage contests between Marmorkrebs against all other crayfish species, that Marmorkrebs appear evenly matched with P. clarkii in contests suggests that, to the extent that aggression affects the invasive potential of a species, Marmorkrebs probably pose a level of threat similar to P. clarkii. Given that P. clarkii is a very successful introduced species around the world (Gherardi 2006), the degree of impact could be substantial in regions where Marmorkrebs are introduced.

To date, Marmorkrebs have only been introduced into habitats far from the regions where Procambarus species are most abundant, namely southern North American and Central America (Hobbs 1984). Given that this species is actively promoted in the North American pet trade, however (Robbins 2009), and the pet trade’s unimpressive record of keeping species captive (Padilla and Williams 2004), introduction of Marmorkrebs into the Americas seems inevitable. Because the genus Procambarus evolved in and is adapted to the Americas, Marmorkrebs may pose a greater threat to the Americas than elsewhere if introduced. That Marmorkrebs have not been given a scientific name or description makes the situation complex for policy makers, should they wish to create legislation and guidelines proactively to limit distribution of Marmorkrebs.


The project was supported by the National Science Foundation Small Grant for Exploratory Research (award IOS-0813581) and Research Experience for Undergraduates program (award DBI-0649273, supporting S.A.J.). We thank Steffen Harzsch and Silvia Sintoni (Department of Evolutionary Neuroethology, Max Planck Institute for Chemical Ecology) for the gift of Marmorkrebs that started the Faulkes laboratory Marmorkrebs research colony.

Copyright information

© Japan Ethological Society and Springer 2010