Delayed metamorphosis in decapod crustaceans : evidence and consequences

Most marine invertebrate species exhibit a complex life cycle including a planktonic larval phase and a benthic juvenileadult phase. Metamorphosis and settlement are the links between these phases of development. In many species, metamorphosis is triggered by specific chemical and/or physical cues, mainly associated with the adult habitat. In the absence of such cues, competent larvae can delay their metamorphosis by a few days to several months. Most investigations on the delay of metamorphosis have been realised on sessile or sedentary species. In relation to mobile decapod crustaceans, the number of such studies is low, probably because the members of this group retain their mobility after metamorphosis, and hence, may depend less on enviromental cues for the induction of settlement and metamorphosis. Nevertheless, the larvae of some decapod species have been shown to depend on metamorphosis-stimulating cues. These include special types of substrates, physical or chemical traits of particular (e.g., estuarine) water bodies, as well as odors from conspecific or congeneric adults. The capacity for delay is, in the decapod species studied so far, limited and may normally end with spontaneous metamorphosis. An extended time of larval development presents the advantage of enhancing the probability for locating a suitable habitat, but it may imply, as a disadvantage, a reduction of juvenile growth or survival and a prolonged development time preceding benthic life. This paper reviews the available evidence for delayed metamorphosis in decapod crustaceans, indentifed cues, the importance of larval age at the time of contact with a cue, and costs of delayed metamorphosis. Additionally, we propose new frontiers for future investigations on delayed metamorphosis in decapod crustaceans, including the molecular identification of chemical cues, the identification of the stage(s) of the moulting cycle that is or are sensitive to such cues, the study of hormonally mediated effects on the moulting cycle, the quantification of energetic or other costs of delayed metamorphosis, and the analysis of relationships between the effectiveness of adult odors and phylogenetic proximity of larvae and adults.


INTRODUCTION
Within the life cycle of many marine organisms with a pelagic larval phase one of the most critical moments is the transition to the benthic habitat.Two processes are involved in this phase, settlement and metamorphosis.
In many larvae, these processes are triggered by specific chemical and/or physical cues associated with the adult habitat (Crisp 1974, Hadfield 1978, Burke 1983, Pechenik 1990, Forward et al. 1997, Lau & Qian 2001, Gebauer et al. 2002).The absence of such cues at the time when larvae are competent to metamorphose may cause a developmental delay of several hours to months (Pechenik 1990) before being concluded by spontaneous metamorphosis or death, depending of the species (Zimmerman & Pechenik 1991, Zalow & Benayahu 1996, Gebauer et al. 1998).Flexibility in the timing of metamorphosis has been considered as a selective advantage, because it should enhance the probability of locating a habitat suitable for juvenile and adult survival (Thorson 1950, Crisp 1974, Obrebski 1979, Morgan 1995).However, postponing metamorphosis might also carry costs, if the larvae with prolonged development time produce juveniles that exhibit high mortality and/or reduced size (Pechenik & Cerulli 1991, Pechenik et al. 1993, Gebauer et al. 1999, Maldonado & Young 1999, Pechenik & Rice 2001, Phillips 2002).
The goal of this work is to review available evidence of delayed metamorphosis in decapod crustaceans.We concentrate specifically on (1) cues that influence the late phase of larval development of decapod crustaceans (metamorphosis and settlement) (2) the time of metamorphic competence, and (3) the costs of an extended larval phase (delayed metamorphosis).Finally, we suggest some new frontiers for future investigations.

CUES FOR METAMORPHOSIS IN DECAPOD CRUSTACEANS
Studies on the dependence of metamorphosis on exogenous cues in mobile species such as decapod crustaceans are scarce compared to the number of studies that have been conducted on sessile species.This imbalance is probably due to the fact that the former do not lose their mobility following metamorphosis and thus maintain the capacity to find habitats with better conditions than those encountered during the initial place of settlement and metamorphosis.During recent years, increasing evidence has accumulated demonstrating flexibility also in the timing of metamorphosis in larvae of various mobile species, depending on the presence or absence of specific cues, similar to what is known from sessile species (McCormick 1999).
Larvae do not only respond to a single specific cue but also to combinations of various stimuli.In C. granulata, for example, the presence of conspecific adults in combination with a muddy substratum (typical sediment in habitats where this species lives) was found to induce metamorphosis faster than each isolated stimulus (Gebauer et al. 1998).Similar results has been reported for megalopae of U. pugilator (O'Connor 1991; Table 1).One conceivable explanation for these results could be an accumulation of substances secreted by the adults on the substratum, producing a higher concentration of this cue and hence a stronger effect of the cues on the metamorphosis of megalopae.
Recent studies demonstrated that megalopae of different species may respond not only to odors secreted by conspecific adults, but also to those from congeners.Gebauer et al. (2002) showed that megalopae of Sesarma curacaoense respond to a substance secreted by both conspecific and and congeneric adults (S. rectum), while substances secreted by other (phylogenetically more distant) Grapsidae had no such effects.A similar pattern has been reported from P. herbstii, where the time to metamorphosis was reduced when megalopae were in contact with exudates produced by members of their own species or by Dyspanopeus sayi, but no effect was observed when megalopae received stimuli from adult U. pugnax (Rodríguez & Epifanio 2000).These results show than the metamorphosis-stimulating effect tends to decrease with increasing phylogenetic distance.However, this may not be a general pattern, as the megalopae of Petrolisthes cinctipes did not show gregarious settlement in response to adults of the congeneric species P. eriomerus (Jensen 1989).Similarly, in an investigation conducted by O'Connor & Gregg (1998), megalopae of Uca pugnax maintained with adult water of U. minax showed no reduced development time.Thus, cross-species activities of exudates from closely related crab species is not universal and might depend on whether particular species live sympatrically or not.A cross effect of substances secreted by species that share the same habitat could present a disadvantage, as it may increase the interspecific competition for benthic resources.Future studies are required to test this prediction and establish the particular conditions where congeneric cues can or can not induce metamorphosis.

METAMORPHIC COMPETENCE
During the larval period it is, in general, possible to distinguish two different phases: (1) a precompetence phase, where the larva has no capacity to metamorphose and principally functions for dispersal and growth in the plankton, (2) the final phase of competence in which the larva is physiologically and morphologically prepared to respond to cues that induce metamorphosis (Crisp 1974, Pechenik 1985, 1999, Pawlik 1992, Avila 1998, Bryan et al. 1998).The onset of metamorphic competence can be identified by triggering metamorphosis with an identified cue after previous absence of morphological or behavioural patterns related to metamorphosis (Pechenik 1984, Pechenik & Heyman 1987, Miller & Hadfield 1986, Pechenik & Qian 1998).
Experiments specifically designed to determine the earliest moment and the length of time of contact with a stimulus required to stimulate a positive larval response are scarce.In most investigations, the contact with the stimulus occurs at the beginning of the megalopa stage.The suitable moment and the length of the period during which the megalopae should be in contact with the cue have been demonstrated for Sesarma curacaoense and Chasmagnathus granulata.The megalopae of these species were experimentally placed into contact with the cue (conspecific adult odors) for differential periods of time during their moulting cycle.The results of these investigations demonstrated that the megalopae of S. curacaoense had latest to come into contact with the cue when about 65% of their moulting cycle had elapsed, and this contact had to last for approximately one day.In C. granulata, this event should latest occur after 32-53% of the time of megalopal development and for a duration of about five days (Gebauer unpubl.).Later or shorter contacts with the cue did not affect the timing of metamorphosis, so that the megalopae metamorphosed at the same age as control larvae that never were in contact with the cue.A similar pattern was found in Uca pugnax, where the appropriate moment for megalopae to receive the metamorphic stimulus was observed after approximately at 67 % of their moulting cycle (O'Connor & Gregg 1998).
Although it is not possible to generalise the few available observations on the exact moment at which larvae become receptive to a metamorphic cue, some tendencies can be identified for the three species studied so far: in all cases, the megalopae appear to be most receptive to such cues during the intermoult, C (after 30-50 % of development), and premoult period, D 0 (after 45-76 % of development), when an increase in the secretion of moulting hormone occurs (Spindler & Anger 1986, Anger 1987, 2001).The cue might thus aid to the increase in hormonal (ecdysteroid) levels and/or to the synchronisation of the moult process.Contact of the megalopa with cues in a late phase of the moulting cycle (D 2 -D 4 premoult) had, in the species studied, no effect on the moment of metamorphosis, which might be explained by a determination of the developmental program before this period.This program seems to be largely independent of extrinsic factors, as suggested by Anger ( 2001) who discussed the absence of effects of starvation or the inability to regenerate mutilated limbs during the late premoult period.

COSTS OF DELAYED METAMORPHOSIS
Delayed metamorphosis has been considered as a selective advantage for species that present this capacity, because it should enhance the probability of locating a suitable habitat for their benthic phase (e.g., Thorson 1950).However, delayed metamorphosis might also produce costs such as decreased juvenile survival or size.In species with lecithotrophic (i.e.non-feeding) larvae, generally, delayed metamorphosis affects the postmetamorphic stages due to an extended period of metabolic energy losses (see Pechenik et a l . 1 9 9 8 , P e c h e n i k 1 9 9 9 f o r r e v i e w ) .I n planktotrophic larvae, in contrast, a prolonged l a r v a l p e r i o d m a y n o t n o r m a l l y a f f e c t postsettlement fitness (Pechenik & Eyster 1989), because such larvae continue feeding i.e., they do not depend on internal energy reserves.If such larvae are nutritionally stressed during their previous development, however, postmetamorphic costs can also occur (Pechenik et al. 1996, Phillips 2002).
In decapod crustaceans, the consequences of delayed metamorphosis have been little studied and there is no consistent pattern (Hunt & Scheibling 1997).Costs of delayed metamorphosis in the form of reduced growth rates of first juveniles have been observed in a hermit crab (Clibanarius longitarsus) (Harvey 1992).Enhanced mortality and reduction of size after a significant prolongation of the larval phase in absence of metamorphosis-stimulating cues have been demonstrated also in the first benthic stage of Chasmagnathus granulata.A possible explanation for such costs may be in an additional utilization of internal reserves, due to a substantial reduction of the ingestion rate during the latest phase of the megalopal moulting cycle, even under constant and optimal feeding conditions (Anger 1991(Anger , 2001)).Due to their smaller initial size, juveniles originating from megalopae with delayed metamorphosis might be more vulnerable also to benthic predation, especially to cannibalism within or between conspecific cohorts of recruits (Eggleston & Amstrong 1995, Hunt & Scheibling 1997, Moksness et al. 1998, Luppi et al. 2001).In general, intraspecific postsettlement predation is considered as one of the potential key factors regulating population density and structure in benthic decapods (e.g., Fernández et al. 1993, Hines & Ruiz 1995, Lovrich & Sainte-Marie 1997, Moksness et al. 1997, Luppi et al. 2002).In addition, smaller juveniles are probably also comparatively weak competitors for food and refuge, which should further reduce their probability of survival and growth in natural habitats (Hines 1986).

CONCLUSIONS
The ability to delay metamorphosis is not an exclusive characteristic of sessile or sedentary species.Megalopae of some mobile decapods have the capacity to respond to specific cues of the adult habitat, and to adjust the timing of settlement and metamorphosis correspondingly.The length of this possible delay is, in general, shorter than in sessile species; it may be concluded by "spontaneous" metamorphosis if no appropriate cues are provided.
These conclusions open several important frontiers for future investigations, for instance the chemical identification and specificity of cues involved in settlement and metamorphosis, or studies of the performance of juveniles and adults in relation to conditions experienced during the larval phase.
Further it should be examined during which stages of the moulting cycle and how metamorphosis-stimulating cues may act during the different phases of the moulting cycle, addressing also the question whether they enhance and/or synchronise hormonal processes.
It is further recommended that we direct more attention to fully marine species, since the available information is centered on estuarine species with a larval export strategy.Additional questions emerge also in the latter category, for example if the larvae of all -or most-species with an export strategy have the capacity to recognize estuarine characteristics as indicators of an adequate habitat likely to promote successful settlement and metamorphosis.
The potential costs linked to periods of delayed metamorphosis lead to some of the most interesting questions in this context.Is reduced size and low survival rate of juveniles originating from megalopae that have delayed metamorphosis a consistent pattern in decapod crustaceans?Are those juveniles weak competitors and/or are they particularly vulnerable to benthic predation or vulnerable for a longer period of time, so that their chances of successful growth and reproduction may be significantly reduced?If those costs indeed are a common consequence of delayed metamorphosis in decapod crustaceans, the causes of these costs should be determined.Are such costs consequences of prolonging the moult stage causing a reduced ingestion of food and an increasing utilization of internal reserves, or are there other underlying causes?The answers to these questions will help to understand the many subtle factors influencing the recruitment success and population dynamics of marine crustaceans.
Clearly, this brief review demonstrates that in crustaceans, as in other marine invertebrates, there are close links between the planktonic larval and the subsequent benthic phase of the life cycle (for a recent discussion, see Giménez 2003, Giménez & Anger 2003).Emphasizing those links, Pechenik et al. (1998) stated that the metamorphosis is not a new beginning, but rather a continuation within the life history.

TABLE 1
Summary of cues for metamorphosis in different species of Decapod CrustaceansResumen de inductores de la metamorfosis en diferentes especies de crustáceos decápodos