Differentiated phenotypic plasticity in larvae of the cannibalistic salamander Hynobius retardatus

Alternative phenotypes in natural populations can arise from either genetic polymorphism or an environmentally induced phenotype, that is, polyphenism. Evolutionary models of polyphenism developed by theoretical studies predict that polyphenism is favored when there are environment-dependent fitness trade-offs between alternatives and that the threshold frequency for a facultative switch between alternative phenotypes is adjusted in accordance with different selection regimes. The broad-headed (alternative) larval morph of Hynobius retardatus, which is induced by crowding with conspecifics or heterospecific anuran (Rana pirica) larvae, is a representative example of cannibalistic polyphenism. Morph induction by such proximate factors must reflect evolutionary (conditional frequency-dependent) processes. To clarify the role of frequency-dependent processes in polyphenism, I investigated the occurrence rate of the broad-headed morph under experimental crowding conditions (low conspecific, high conspecific, and high heterospecific densities) using larvae from eight natural populations with different larval densities of conspecifics and heterospecifics, and found interpopulational differences in the expression of the morph. Thus, there is a larval density-dependent equilibrium frequency of the morph in each pond, suggesting that the local switch point for morph induction was modified by selection to produce evolved differences between ponds. The evolution of such interpond differences has three necessary conditions: (1) There are pond-dependent fitness trade-offs between alternatives, (2) The maintenance of the morph is costly, and (3) The presence of conspecific or, especially, heterospecific larvae provides a reliable cue to the receiver.     

A trade-off between prey- and predator-induced polyphenisms in larvae of the salamander Hynobius retardatus

Organisms in natural habitats participate in complex ecological interactions that include competition, predation, and foraging. Under natural aquatic environmental conditions, amphibian larvae can simultaneously receive multiple signals from conspecifics, predators, and prey, implying that predator-induced morphological defenses can occur in prey and that prey-induced offensive morphological traits may develop in predators. Although multiple adaptive plasticity, such as inducible defenses and inducible offensive traits, can be expected to have not only ecological but also evolutionary implications, few empirical studies report on species having such plasticity. The broad-headed larval morph of Hynobius retardatus, which is induced by crowding with heterospecific anuran (Rana pirica) larvae, is a representative example of prey-induced polyphenism. The morph is one of two distinct morphs that have been identified in this species; the other is the typical morph. In this paper, we report that typical larval morphs of Hynobius can respond rapidly to a predatory environment and show conspicuous predator-induced plasticity of larval tail depth, but that broad-headed morphs cannot respond similarly to a predation threat. Our findings support the hypothesis that induction or maintenance of adaptive plasticity (e.g., predator-induced polyphenism) trades off against other adaptive plastic responses (e.g., prey-induced polyphenism). For a species to retain both an ability to forage for larger prey and an ability to more effectively resist predation makes sense in light of the range of environments that many salamander larvae experience in nature. Our results suggest that the salamander larvae clearly discriminate between cues from prey and those from predators and accurately respond to each cue; that is, they adjust their phenotype to the current environment.     

Widespread contribution of methane-cycle bacteria to the diets of lake profundal chironomid larvae

Reports of unexpectedly 13C-depleted chironomid larvae in lakes have led to an hypothesis that significant transfer of detrital organic matter to chironomid larvae may occur via methane-cycle bacteria. However, to date little is known of how such transfer might vary across species and lakes. We gathered data from 87 lakes to determine how widespread this phenomenon might be and to define boundaries for its likely magnitude. Carbon stable isotope values of chironomid larvae varied greatly between taxa. Very marked 13C-depletion was evident only in certain taxa, especially Chironomus plumosus, C. anthracinus, and C. tenuistylus, all characteristic of eutrophic or dystrophic lakes and known to be tolerant of low oxygen conditions. Furthermore, marked 13C-depletion was only found in larvae from lakes in which late-summer hypolimnetic oxygen depletion near the sediment surface was below an apparent threshold concentration of 2-4 mg O2/L. Similarly, application of a two-source mixing model suggested that methanotrophic bacteria made the greatest contribution to profundal chironomid growth (0-70% of larval carbon) when the late-summer oxygen concentration dropped below approximately 2 mg O2/L. Our study demonstrates that methane-derived carbon is an important, but often neglected, contribution to the flux of carbon through the food webs of many productive or dystrophic lakes.     

Factors which affect the occurrence of cannibalism and the broad-headed "cannibal" morph in larvae of the salamander Hynobius retardatus

Cannibalism in amphibian larvae may be suppressed among siblings in comparison to distant relatives or nonkin, even when a "cannibal" morph that can consume conspecifics shows adaptive advantages. Two experimental studies were undertaken to analyze factors which affect the frequency of cannibalism and the occurrence of the broad-headed "cannibal" morph in larvae of the salamander Hynobius retardatus. The first experiment investigated whether or not the broad-headed "cannibal" morph is only induced after performing cannibalism. Because the broad-headed "cannibal" morph was observed in larval populations that showed no cannibalism, it was concluded that the "cannibal" morph could be induced without actual cannibalism. Second, possible factors affecting the occurrence of the broad-headed "cannibal" morph were analyzed with respect to the density of larvae, level of food supply, and kinship among larvae, alone or in combination. Appearance of the broad-headed "cannibal" morph was affected by interactive effects of density2kinship: although the morph was induced significantly more often at a higher density if the conspecifics were distantly or not related, it was strongly suppressed even at high density if larval kinship was very close or among siblings. In contrast, the frequency of cannibalism was independently affected by larval density, level of food supply, and kinship among larvae: it was significantly larger at high than at low larval densities, at low than at high food supply, and in mixed-sibling groups than in pure sibships. These results suggest that the "cannibal" morphs in H. retardatus are induced even without cannibalism at a high density of conspecifics in mixed-sibling groups, but the occurrence of this morph, which continuously consumes conspecifics, is suppressed among siblings.     

A copula-based inference to piecewise exponential models under dependent censoring,with application to time to metamorphosis of salamander larvae

In ecology and evolutionary biology, controlled animal experiments are often conducted to measure time to metamorphosis which is possibly censored by the competing risk of death and the follow-up end. This paper considers the problem of estimating the survival function of time-to-event when it is subject to dependent censoring. When the censorship is due to competing risks, the traditional assumption of independent censorship may not be satisfied, and hence, the usual application of the Kaplan–Meier estimator yields a biased estimation for the survival function of the event time. This paper follows an assumed copula approach (Zheng and Klein in Biometrika 82(1):127–138, 1995) to adjust for dependence between the event time of interest and the competing event time. While the literature on an assumed copula approach has mostly focused on semiparametric settings, we alternatively consider a parametric approach with piecewise exponential models for fitting the survival function. We develop maximum likelihood estimation under the piecewise exponential models with an assumed copula. A goodness-of-fit procedure is also developed, which touches upon the identifiability issue of the copula. We conduct simulations to examine the performance of the proposed method and compare it with an existing semiparametric method. The method is applied to real data analysis on time to metamorphosis for salamander larvae living in Hokkaido, Japan (Michimae et al. in Evol Ecol Res 16:617–629, 2014).     

Identification of sex and parasitism via pheromones by the Ozark zigzag salamander

Chemical communication appears to be the primary mode of information transfer among woodland salamanders of the genus Plethodon. These pheromones have been shown to convey a wide variety of information, and are used to establish territorial boundaries. Responses to pheromones can include changes in both behavior and energy consumption. The ability of Ozark zigzag salamanders, Plethodon angusticlavius, to detect the sex of a conspecific through chemical substrate markings has not been assessed. In addition, some populations of P. angusticlavius are naturally infected with the ectoparasitic mite Hannemania eltoni. Parasitism can impact both the behavior and metabolism of the host as well as the behavior of others towards the parasitized individual. We examined whether sex and parasitism of the pheromone donor and receiver influenced behavioral and metabolic responses in P. angusticlavius. In laboratory trials, behavior of P. angusticlavius was influenced by the sex and parasite load of the pheromone donors and receivers. Behavioral responses were also influenced by whether the receiver was in its own territory or in the territory of the pheromone donor. Metabolic responses were not affected by the sex or parasite load of the pheromone donor, but males had higher metabolic rates than females, even after adjusting for differences in body size. Overall, our study indicates that both sex and parasitism can be detected via pheromones in this species, and the nature of the response depends on residency status, sex, and parasite load of the receiver.     

Organization of Plethodon salamander communities: guild-based community assembly

A long-standing goal in evolutionary ecology is to determine whether the organization of communities is reflective of underlying deterministic processes. In this study, I examined patterns of species co-occurrence among eastern Plethodon salamanders and determined whether they were consistent with predictions from a guild model of competition-based community assembly. Using a database of 45 species and 4540 geographic sites, I found that patterns of co-occurrence were significantly nonrandom at both a regional and continental scale, and species of different size guilds were distributed more evenly in sites than was expected by chance. Sites with the highest species richness had consistent patterns of community composition, and with few exceptions, the same five species were present at all sites. Taken together, these results imply that larger Plethodon communities are assembled from simpler communities in a manner consistent with what is predicted through competitive mechanisms and suggest that stable species combinations are possible to achieve at various levels of species richness. These results also provide strong evidence consistent with the hypothesis that competitive-based community assembly is a general phenomenon in Plethodon and that interspecific competition is prevalent among the eastern species of this group.     

Linking direct and indirect data on dispersal: isolation by slope in a headwater stream salamander

There is growing recognition of the need to incorporate information on movement behavior in landscape-scale studies of dispersal. One way to do this is by using indirect indices of dispersal (e.g., genetic differentiation) to test predictions derived from direct data on movement behavior. Mark-recapture studies documented upstream-biased movement in the salamander Gyrinophilus porphyriticus (Plethodontidae). Based on this information, we hypothesized that gene flow in G. porphyriticus is affected by the slope of the stream. Specifically, because the energy required for upstream dispersal is positively related to slope, we predicted gene flow to be negatively related to change in elevation between sampling sites. Using amplified DNA fragment length polymorphisms among tissue samples from paired sites in nine streams in the Hubbard Brook Watershed, New Hampshire, USA, we found that genetic distances between downstream and upstream sites were positively related to change in elevation over standardized 1-km distances. This pattern of isolation by slope elucidates controls on population connectivity in stream networks and underscores the potential for specific behaviors to affect the genetic structure of species at the landscape scale. More broadly, our results show the value of combining direct data on movement behavior and indirect indices to assess patterns and consequences of dispersal in spatially complex ecosystems.     

Effects of monospecific and mixed-algae diets on survival, development and fatty acid composition of penaeid prawn (Penaeus spp.) larvae

Four species of microalgae (Chaetoceros muelleri, Tetraselmis suecica, Tahitian Isochrysis sp. (T-iso) and Dunaliella tertiolecta) with distinctly different fatty acid profiles were grown in continuous culture and fed to prawn larvae (Penaeus japonicus, P. semisulcatus and P. monodon) as monospecific diets. The best two diets (C. muelleri and T. suecica) were also fed as a mixed diet. Experiments were run until the larvae fed the control diet of C. muelleri metamorphosed to Mysis 1. The survival and development (i.e. performance) of the larvae were affected by algal diet, and the diets were ranked in the order of decreasing nutritional value: C. muelleri ≥ T. suecica > T-iso > D. tertiolecta. Larvae fed a mixed diet of C. muelleri and T. suecica (2:3 by dry weight) performed as well or better than those fed C. muelleri, and the performance of both these groups of larvae was better than those fed T. suecica. The lipid and carbohydrate compositions of the algae had little or no effect on the lipid and carbohydrate compositions of the larvae or their performance. However, the larvae that performed best (i.e. those fed C. muelleri) had significantly more lipid and carbohydrate than those that performed worst (i.e. those fed D. tertiolecta). Larvae fed C. muelleri or the mixed-algae diet had higher proportions of the essential fatty acids eicosapentaenoic acid [EPA, 20:5(n-3)] and arachidonic acid [ARA, 20:4(n-6)] than the larvae fed on other diets. Furthermore, the larvae fed T. suecica, which showed intermediate performance between larvae fed C. muelleri and T-iso or D. tertiolecta, also had higher proportions of EPA and ARA. Both C. muelleri and T. suecica contained EPA and ARA, but T-iso and D. tertiolecta did not, except for trace amounts of EPA in T-iso. The fatty acid ARA appears to be much more important in the diet of larval prawns than has so far been considered. The level of the essential fatty acid docosahexaenoic acid [DHA, 22:6(n-3)] in the algal diet and the larvae was not related to the performance of the larvae; only C. muelleri and T-iso contained DHA. However, the nauplii contained large proportions of DHA, suggesting that these were sufficient to meet the larval requirements for DHA during their development to Mysis 1. Mixed-algae diets could improve the performance of larvae by providing a more comprehensive range of fatty acids. Received: 22 April 1998 / Accepted: 3 December 1998     

Effects of monospecific algal diets of varying biochemical composition on the growth and survival of Pacific oyster (Crassostrea gigas) larvae

Using monospecific diets of Thalassiosira pseudonana cells grown under different steady-state conditions, it was determined that higher growth rates of larval Crassostrea gigas Thunberg were obtained when fed T. pseudonana cells grown under high light. High light grown T. pseudonana cells consistently contained relatively more of the saturated fatty acids 14:0 and 16:0. Considered over three independent experiments, high light grown T. pseudonana cells were lower in protein and higher in carbohydrate than low light grown cells. Higher growth rates of larval C. gigas were obtained on diets with more of the essential fatty acid (EFA) 22:63, and less of the other EFA, 20:53. The relative requirements of C. gigas larvae for the essential fatty acids 20:53 and 22:63 are discussed. Faster growing larvae contained higher percentages of the fatty acids 14:0 and 16:0, and lower percentages of 22:2j. Oyster growth rates were correlated with their content of the fatty acids: 14:0, 16:0 and 22:2j in two experiments utilizing separately spawned batches of larvae. Fatty acid profiles are proposed as a technique for assessing larval condition. C. gigas larvae contained ten times the percent composition of the FAs 16:43, 18:17, 20:17 and 22:2j compared with their diet. Correlation analysis suggests that the dietary source of 18:17, 20:17 and 22:2j was 16:17. It is concluded that T. pseudonana cells grown under high light are a superior diet for C. gigas larvae in comparison with low light grown cells of the same species.     

Selection of substrata by seaweeds: Optimal surface relief

Artificial substrata allow one to test a single factor in a complex environment. We wanted to determine whether surface particle size could regulate settlement and growth in marine macroalgae. Three grades of discrete monolayers of hard particles differing only in diameter (0.1 to 0.5 mm, 0.5 to 1.0 mm, 1.0 to 2.0 mm) were cemented to the surfaces of three quadrants on acrylic discs while the fourth was left smooth. All surfaces were painted with dissolved plastic to ensure chemical uniformity. At different times of the year, 25 plates were bolted directly to basement rock in the low intertidal zone on an exposed coast at the mouth of Narragansett Bay in Rhode Island (Fig. 1). Thirteen conspicuous species of macroalgae colonized the coated acrylic plates during the course of this investigation. Of these, the most abundant were Chondrus crispus, Corallina officinalis, Polysiphonia harveyi and Ulva lactuca, which were observed throughout the program. Initial settlement did not differ significantly among the experimental surfaces, but with time patterns in the distribution of these algal populations correlated with the surfaces beneath them. These differences were visually striking and not the same for each species. Most of the Chondrus crispus (79.5%) and U. lactuca (85.2%) appeared on the two largest particles, but only for C. crispus was the difference in populations within these grades significant (P<0.05). Populations were considerably smaller on the smallest grade (20.1 and 13.8% respectively), and on the smooth quadrants only a few individuals of these two species appeared (0.5 and 1.0%). The number of plants of P. harveyi, on the other hand, did not differ on the three particle sizes tested, but on the smooth surfaces only 2.9% appeared. Corallina officinalis showed optimal development on the smallest grade (44.8%). On smooth surfaces, where 8.4% of its population was observed, no upright articulated portions developed. The season during which the substrata were introduced altered the patterns of development. C. officinalis was the first macroalga to be seen in late spring, but on plates set out in the fall this species did not appear until the following late spring. P. harveyi covered some discs in fall, spring and summer, and on these surfaces Chondrus crispus was delayed up to 1 year. By the end of the experimental period (15 months), C. crispus was the most conspicuous macroalgal species on most of the plates.     

Reducing complex diets to simple rules: food selection by olive baboons

Determinants of diet in free-ranging baboons Papio anubis were investigated. Foods and non-foods differed significantly in chemical composition, the former being higher in protein and lower in fibre and phenolics. Within the range of items selected, biomass was found to be the single most important factor affecting percentage of total intake and time spent feeding (Figure 1). When this effect of biomass was statistically controlled, independent effects of chemical composition (primarily protein content) and harvesting rate were found. However, while the harvesting rates of foods were positively related to their percentage contribution to total intake, they were either uncorrelated, or, in one case, actually negatively correlated with the percentage of feeding time. A post-hoc model based on Charnov's (1976) marginal value theorem is developed to account for these latter results, and direct evidence for rate-maximising patch use is then presented. It is argued that the apparent complexity of primate diets may in large part be reducible to relatively simple optimization criteria.     

Diel burying by the tropical sea cucumber Holothuria scabra: effects of environmental stimuli, handling and ontogeny

Understanding concealment behaviour of marine animals is vital for population surveys and captive-release programmes. The commercially valuable sea cucumber Holothuria scabra Jaeger 1833 (Holothuroidea) can display a diel burying cycle, but is it widely predictable? Circadian burying of captive H. scabra juveniles, and both juveniles and adults in the wild, was examined in New Caledonia. Groups of ten cultured juveniles in mesh chambers in a tank were monitored for 24 h. Small juveniles (1–5 g) displayed an expected diel cycle of epibenthic foraging in the afternoon and night then burial in sediments in the morning. Burial was related significantly to both light and temperature in combination. Similar groups of juveniles were handled once or three times a day for 1 week then frequency of emergence during another week was compared to unhandled controls. Handling stress, whether occasional or frequent, significantly suppressed the frequency of their afternoon emergence from sediments for 4 days. In a coastal seagrass bed, burial and emergence of H. scabra were monitored during days of opposing tidal cycles in three seasons. Adults seldom buried during the day except in the cool season. At that site, most small hatchery-produced H. scabra juveniles were buried during most of the day, while larger juveniles showed little diurnal burying. This study underscores that the circadian behaviours of marine animals can exhibit substantial spatial variation, may be absent at certain sites or seasons, and can be mediated by a complexity of factors that vary over short timescales.     

Feeding by larvae of intertidal invertebrates: assessing their position in pelagic food webs

One of the leading determinants of the structure and dynamics of marine populations is the rate of arrival of new individuals to local sites. While physical transport processes play major roles in delivering larvae to the shore, these processes become most important after larvae have survived the perils of life in the plankton, where they usually suffer great mortality. The lack of information regarding larval feeding makes it difficult to assess the effects of food supply on larval survival, or the role larvae may play in nearshore food webs. Here, we examine the spectrum of food sizes and food types consumed by the larvae of two intertidal barnacle species and of the predatory gastropod Concholepas concholepas. We conducted replicated experiments in which larvae were exposed to the food size spectrum (phytoplankton, microprotozoan and autotrophic picoplankton) found in nearshore waters in central Chile. Results show that barnacle nauplii and gastropod veligers are omnivorous grazers, incorporating significant fractions of heterotrophs in their diets. In accordance with their feeding mechanisms and body size, barnacle nauplii were able to feed on autotrophic picoplankton (<5 microm) and did not consume the largest phytoplankton cells, which made the bulk of phytoplankton biomass in spring-summer blooms. Balanoid nauplii exhibited higher ingestion rates than the smaller-bodied chthamaloid larvae. Newly hatched C. concholepas larvae also consumed picoplankton cells, while competent larvae of this species ingested mostly the largest phytoplankton cells and heterotrophic protozoans. Results suggest that persistent changes in the structure of pelagic food webs can have important effects on the species-specific food availability for invertebrate larvae, which can result in large-scale differences in recruitment rates of a given species, and in the relative recruitment success of the different species that make up benthic communities.     

Response to light by trochophore larvae of Spirobranchus giganteus

Trochophore larvae of Spirobranchus giganteus (Pallas) respond positively to white light at levels of illumination from 1 to 2 168×10¹⁴ quanta cm^-2 s^-1. In this range the strength of the response is not correlated with irradiance level. The response is increased by dark adaptation. At low levels of irradiance (0.1-2.0×10¹⁴ quanta cm^-2 s^-1) larvae respond positively to blue (360-510 nm, max. 430 nm) and green (475–620 nm, max. 530 nm) light but not to wavelengths of 590 nm or over. The light response develops gradually during the 12 h following the appearance of the eyespot and is maintained throughout the remainder of the planktonic phase.     

Rates of energy consumption and acquisition by lecithotrophic larvae of Bugula neritina (Bryozoa: Cheilostomata)

Lecithotrophic larvae of the cheilostome bryozoan, Bugula neritina (L.), lose metamorphic competence 12 to 24 h after release from the maternal zooid. The high respiration rate of newly released larvae (mean=306.3 pmol O₂ larva^-1 h^-1, range=149.3 to 466.6, n=18 trials, 22.5°C) from adults collected at Link Port, Fort Pierce, Florida during the winter/spring of 1990–1991 reflects their active swimming behavior. The average energy content per larva was 15.24 mJ (range: 13.35 to 20.17 mJ ind^-1, n=5 groups). If all cells have an identical energy content and metabolic rate, then 2 and 20% of the total energy content would be consumed by the onset (2 h post-release) and the loss (24 h post-release) of metamorphic competence. Larvae of B. neritina are a composite of both larval and juvenile tissues and the loss of metamorphic competence may be due to regional depletion of labile energy stores in transitory larval cells, particularly the ciliated cells that comprise the locomotory organ, the corona. Although nonfeeding, B. neritina larvae can acquire nutrients from the environment in the form of dissolved organic materials (DOM) in seawater. Both the amino acid alanine and the fatty acid palmitic acid can be transported from seawater ([S]=1 M, 22.5°C). The rates of alanine influx (appearance of label in tissue) averaged 0.366 pmol larva^-1 h^-1 and, based on comparisons between rates of solute transport and metabolism, would contribute little (<1% of required energy) to offset the metabolic demand. The average rate of palmitic acid influx was 4.668 pmol larva^-1 h^-1 and assuming that the measured influx equals the net solute flux, could account for 21 to 72% of energy requirements. These data suggest that the duration of planktonic life of B. neritina larvae is principally regulated by the amount of endogenous energy stores, but may be modulated by available DOM in seawater.     

Dieldrin uptake by larvae of the crab Leptodius floridanus

The rate of uptake of ¹⁴C-dieldrin by crab larvae (Leptodius floridanus) from 0.5 ppb in seawater and from 213 ppb (dry weight) in their food was measured. It was found that, if equal concentrations of dieldrin were available to the larvae in their food and in seawater, the animals would accumulate the pesticide about 8000 times as fast from the water as from the food.     

Explaining long-distance dispersal: effects of dispersal distance on survival and growth in a stream salamander

Long-distance dispersal (LDD) may contribute disproportionately to range expansions, the creation of new evolutionary lineages, and species persistence in human-dominated landscapes. However, because data on the individual consequences of dispersal distance are extremely limited, we have little insight on how LDD is maintained in natural populations. I used six years of spatially explicit capture-mark-recapture (CMR) data to test the prediction that individual performance increases with dispersal distance in the stream salamander Gyrinophilus porphyriticus. Dispersal distance was total distance moved along the 1-km study stream, ranging from 0 to 565 m. To quantify individual performance, I used CMR estimates of survival and individual growth rates based on change in body length. Survival and growth rates increased significantly with dispersal distance. These relationships were not confounded by pre-dispersal body condition or by ecological gradients along the stream. Individual benefits of LDD were likely caused by an increase in the upper limit of settlement site quality with dispersal distance. My results do not support the view that the fitness consequences of LDD are unpredictable and instead suggest that consistent evolutionary mechanisms may explain the prevalence of LDD in nature. They also highlight the value of direct CMR data for understanding the individual consequences of variation in dispersal distance and how that variation is maintained in natural populations.     

Sequestration of phorbol esters by aposematic larvae of Hyles euphorbiae (Lepidoptera: Sphingidae)?

Summary. The larvae of the hawkmoth species Hyles euphorbiae have a conspicuous aposematic colouration and show gregarious behaviour. It has thus been suggested that they sequester phorbol esters from their food plants which include different species of the genus Euphorbia (Euphorbiaceae) for chemical protection against predators. To test this hypothesis in more detail, we fed larvae an artificial diet with three doses of 12-tetradecanoyl-phorbol-13-acetate (TPA), then examined the faeces and the larval tissues, such as integument, haemolymph and gut of the caterpillars for the presence of TPA. In order to determine the ability of the larvae to detoxify phorbol esters, other larvae were directly injected with a TPA solution and analysed in the same manner. Our study indicates that the larvae of Hyles euphorbiae do not sequester phorbol esters. Upon oral application TPA was not found in the larval integument or the haemolymph. Instead, it was mostly metabolised (about 70–90%). Nevertheless, about 10-30% were retained and recovered in the faeces. The larvae were also able to metabolise and thus detoxify the phorbol ester when TPA was injected directly into the body. These hawkmoth caterpillars are relatively large and have a gut full of plant material, which they regurgitate into the direction of the predator when attacked in nature. Since phorbol esters are very potent toxins and irritants, we postulate that the gut content (and especially the plant slurry disgorged as regurgitant from the anterior gut) alone could be aversive for a potential predator, even if some metabolism has taken place. Thus, although H. euphorbiae caterpillars do not actively sequester phorbol esters, their aposematic colouration appears to be based on chemical defence through phorbol esters retained in the gut.