Introduction

Fuelling larvae with essential compounds to maximise reproductive success is a determinant step for the recruitment of marine organisms. Fatty acids (FA) and carotenoids both have metabolic roles in embryonic and larval development. Yet, their changes in sea cucumber gonads during gametogenesis are still unknown. FA are major components of cell membranes, they play various roles in cell signalling and immunity (Tocher and Glencross 2015), they serve as energy reserve and they may even act as semiochemicals (Watson et al. 2009). Studies in the wild have generally recognised that changes in FA composition of sea urchin gonads (the closest relatives to sea cucumbers) primarily reflect dietary inputs, although reproductive status could also alter relative FA abundances (Hughes et al. 2005; Martinez-Pita et al. 2010; Zárate et al. 2016; Díaz de Vivar et al. 2019; Rocha et al. 2019). Yet, feeding trials have evidenced that the FA balance of marine animal gonads, including sea urchins, influence egg performance and larval survival (Gago et al. 2009; Nhan et al. 2020; Steinberg 2022).

Carotenoids have excellent antioxidative activities for quenching singlet oxygen and inhibiting lipid peroxidation, they may contribute to eggs photoprotection and serve as energy reserves (Shimidzu et al. 1996; Tsushima 2007; Maoka 2011; Galasso et al. 2017). To date, most studies dealing with carotenoids in echinoderms focused on identifying new compounds (Matsuno and Tsuchima 1995; Mariutti et al. 2012), while seasonal variations were only described for sea urchins (Symonds et al. 2007, 2009; Hagen et al. 2008; Rocha et al. 2019). Comparisons between sea cucumber species revealed a strong heterogeneity of compounds among taxonomic groups, with one compound among astaxanthin, canthaxanthin, echinenone or cucumariaxanthin generally dominating the pool of carotenoids (Matsuno et al. 1969; Matsuno and Ito 1971; Matsuno and Tsuchima 1995; Bandaranayake and Des Rocher 1999; Hudson et al. 2003). In addition, contrary to FA, carotenoids have been shown to reflect gametogenesis and/or embryogenesis in marine organisms, both in sea urchins (Hagen et al. 2008; Rocha et al. 2019) and in other species such as crabs (Zadorozhny et al. 2008), suggesting that they may fulfil a specific function for reproduction (Galasso et al. 2017).

The black sea cucumber Holothuria (Panningothuria) forskali Delle Chiaje, 1823 is a temperate species of deposit-feeding Holothuroidae (Echinodermata) commonly found in the East Atlantic from the British Isles to the Canary Islands and along the Mediterranean coast (Koehler 1921; MNHN & OFB 2002). Its reproductive cycle is synchronous with a brief spring spawning (March–April), when the temperatures are increasing; a brief resting phase (May–June; not always observed in recent studies focusing on the same species; Santos et al. 2016; Ballesteros et al. 2021); a long reabsorption/maturation period (June–October) during the warm season and a long mature phase (November–March) during the cold season (Tuwo and Conand 1992). The species follows the “tubule recruitment model” (Smiley 1988; Sewell et al. 1997) with five cohorts of tubules of different development stages (four maturing cohorts + one cohort of resorbed tubules). Basically, germ cells at the basis of the gonad develop in distinct primary, secondary and fecund cohorts, with each cohort of tubules having synchronous development of oocytes. The entire process of ovary development and oocyte maturation can be viewed as a slowly moving conveyor belt carrying containers of maturing oocytes to the posterior end of the gonad basis. After spawning, the relict fecund gonad tubules are resorbed but the primary and secondary tubules persist on the gonad basis and migrate posteriorly as new primary tubules are formed. Consequently, the gonad must contain tubules at various stages of maturation whatever the period of the year (Tuwo and Conand 1992).

Temperate holothurians production is gaining interest across Europe and North Africa to meet Asian demand and reduce the pressure on natural populations (Santos et al. 2016; Domínguez-Godino et al. 2018, 2019; Rakaj et al. 2018, 2019). As various other related species, H. forskali has recently been reproduced in controlled conditions (Laguerre et al. 2020) and gaining knowledge on its reproductive cycle is necessary to improve its reproductive success in captivity. To date, oocytes maturation in temperate sea cucumbers has essentially been studied through histological examination and gonad index determination. In the present study, we aimed at describing gametogenesis of both male and female H. forskali from a biochemical point a view, especially focusing on FA and carotenoids. In addition, we intended to separate the effect of season (i.e. diet availability) from that of the reproductive status (i.e. maturing, spawning, reabsorbing) on the biochemical composition of H. forskali gonads, being aware that both are necessary interrelated.

Materials and methods

Sample collection and processing

We sampled 6–11 sea cucumbers of the species Holothuria (Panningothuria) forskali Delle Chiaje, 1823 approximately every 2 months from December 2019 to July 2021 east of the Glenan Islands (Brittany – France; 47.710°N, 3.948°W) at a depth of 8–12 m (Table 1). The samples were collected as part of the HoloFarm research project (2018–2021), which aimed at developing rearing methods for European sea cucumbers. Given that years 2020 and 2021 were marked by the COVID-19 outbreak the sampling schedule was sometimes irregular, leading to gaps in the dataset. The sampling area was characterised by a rocky substrate where Laminaria sp. (Phaeophyceae—Laminariales) forests develop along with various species of red algae, and flat sandy bottom nearby. Patches of sediment dominated by sand and shell fragments accumulated in the depressions. Sea cucumbers were brought to the Concarneau marine station and maintained in aquaria with a continuous flow of sand-filtered seawater and no food supplying until dissection, which usually took place within 1–5 days after collection of the specimens.

Table 1 Number of samples and organ weights of female and male Holothuria forskali sampled in the Glénan Islands, France, from December 2019 to July 2021

Individuals were tranquilised for 30 min in a freezer before being dissected. Sea cucumbers were incised dorso-longitudinally and all internal organs were removed, leaving empty body walls referred as gutted (i.e. eviscerated) individuals. Gonads were isolated from internal organs, weighed and sex was determined by microscopic examination. We calculated gonad index as the percentage of gonad wet weight to body wall wet weight (i.e. eviscerated weight). All gonad material was then kept at – 25 °C and freeze-dried at – 80 °C and 0.8 mBar during 24 h before analysis of FA and carotenoids. Gonads were crushed to fine powder and all tubule classes were mixed, assuming that fecund tubules (T3 and/or T4) would largely dominate the gonad pool (Tuwo and Conand 1992 and supplementary file S1). However, this procedure may have confounded the effects of gametes development stages (by mixing earlier stages of development of the eggs/sperm with the later stages, each with different relative proportions at different times of the year) with those of season (possibly induced by diet changes and/or physiological responses to water parameters). Carotenoids were analysed on all samples (14 sampling events; 116 individuals; see Table 1), while FA were determined on broader time intervals (11 sampling events; 94 individuals). Sea urchins, employed for reference material for carotenoid identification (Fig. 1), were taken from the living collections of the Concarneau marine station (Psammechinus miliaris (P.L.S. Müller, 1771)) or gathered manually in the nearby tidal flats (Paracentrotus lividus (Lamarck, 1916)). Individuals from the sea cucumber Holothuria (Halodeima) atra Jaeger, 1833 were collected near Moorea (French Polynesia) and kindly provided by Guillaume Caulier (University of Mons – Belgium).

Fig. 1
figure 1

High-performance liquid chromatography (HPLC) carotenoid profile of common echinoderm gonads and Holothuria atra viscera (absorbance at 470 nm). Only peaks 1, 3 and 6 were identified with certainty and match with commercial standards. Peak 1 + 2 astaxanthin; 3 canthaxanthin; 4 canthaxanthin isomer; 5 echinenone isomer; 6 all-trans-echinenone

Fatty acid profiling

We analysed FA following a slightly modified method of Bligh and Dyer (1959) as developed by Meziane et al. (2007) and updated as in David et al. (2020). Lipids were extracted from 20 to 35 mg of freeze-dried gonad material using a methanol:water:chloroform mixture (2:1:1 v:v:v) after adding 40 μg of tricosanoic acid (23:0) provided by Sigma-Aldrich as internal standard. Chloroform, containing lipid extracts, was isolated and evaporated under nitrogen. Dried lipid extracts were then saponified with a solution of methanol and sodium hydroxide (NaOH, 2N) (2:1, v:v) at 90 °C for 1h30 min. Subsequently, methylation was performed using methanolic boron trifluoride 14% (BF3–MeOH).

Fatty acid methyl esters (FAME) were analysed by gas chromatography (Agilent, 7890B GC) equipped with an Agilent VF-WAXms capillary column (30 m length × 0.25 mm inner diameter × 0.25 μm film thickness) and quantified using a flame ionisation detector (FID). The oven temperature was set at 60 °C and held for 1 min, raised at 40 °C min−1 to 150 °C and held for 3 min, and then increased to 240 °C at 3 °C min−1 and held for 8 min. We identified FA using coupled gas chromatography with simple quadrupole mass spectrometry (Agilent, 5977B MSD) using the NIST mass spectral library and comparison of GC retention times with commercial standards (Supelco® 37 component FAME mix). The internal standard (23:0) was used to determine the concentration of each FA in µg per mg of freeze-dried gonad material.

Carotenoid determination

Carotenoids were analysed by high-performance liquid chromatography (HPLC) according to Brotas and Plante-Cuny (2003). We disrupted 10–20 mg of female and 30–50 mg of male freeze-dried gonad material (due to higher carotenoid concentrations in females compared to males) in 1.5 mL of 95% methanol (buffered with 2% w:v ammonium acetate) with one stainless steel bead in a 2 mL plastic tube using a MM 400 Retsch bead mill at 30 Hz for 10 min, thus allowing a complete transfer of pigments from tissues to the extraction solvent. Extracts were then filtered with 0.2 μm PTFE syringe filters and analysed within 16 h using an Agilent 1260 Infinity HPLC composed of a quaternary pump (VL 400 bar), a UV–VIS photodiode array detector (DAD 1260 VL, 250–900 nm) and a 100 μl automatic sample injector refrigerated at 4 °C in the dark. Chromatographic separation was carried out using a C18 column for reverse phase chromatography (Supelcosil, 25 cm long, 4.6 mm inner diameter). The solvents used were A: 0.5 M ammonium acetate in methanol and water (85:15, v:v), B: acetonitrile and water (90:10, v:v), and C: 100% ethyl acetate. The solvent gradient was set according to Brotas and Plante-Cuny (2003), with a 0.5 mL min−1 flow rate. Identification and calibration of the HPLC peaks were performed with astaxanthin, canthaxanthin and all-trans-echinenone standards provided by DHI Laboratory Products (Fig. 1). Compounds with no available standard were identified by comparison of H. forskali with previously studied echinoderms (Bandaranayake and Des Rocher 1999 for Holothuria atra; Symonds et al. 2007 for Paracentrotus lividus; Symonds et al. 2009 for Psammechinus miliaris). Peak profiles (300–600 nm in 2 nm increments) recorded by the HPLC were provided in Fig. 1 and used to confirm carotenoids identification. Quantification was performed at 470 nm using standard calibration curves built with repeated injections of standards over a range of dilutions. The relative abundance of each pigment (%) was calculated from its respective concentration in the sample (μg g −1).

Statistical analysis

We performed univariate male/female comparisons using Student's t test with Welch's modification of the degrees of freedom for non-homogeneous sample variances. Comparisons involving both sex and sampling events were done with the non-parametric Van der Waerden test (R package “agricolae”) due to small sample sizes (n = 1–6). Potential differences in FA or carotenoid proportions of sea cucumber gonads were assessed using permutational multivariate analysis of variance (PERMANOVA) on Bray–Curtis dissimilarity matrices. Prior verification of multivariate group dispersion was performed using betadisper function and date was taken as a categorical factor, meaning that the statistical comparison of each sampling event was not influenced by the results of the previous one. We visualised the underlying structure of gonads FA profiles using Principal Correspondence Analysis (PCoA) on Bray–Curtis dissimilarity matrices. Samples were agglomerated using the Ward method and clustered with the cutree function into three groups. Statistical analysis and graphical representations were performed using R 4.1.2 (R Core Team 2021) and type I error (α) was set to 5%.

Results

Reproductive cycle

A total of 116 sea cucumbers were dissected and weighed, including 61 females and 55 males. The weight of gutted individuals ranged from 71.1 to 223.1 g and the weight of gonads from 1.2 to 64.2 g. Highest gonad indices were measured in females sampled in March 2020 (33.0 ± 10.6%) and lowest values were recorded in females collected in April 2021 (2.2 ± 1.2%; Fig. 2), with significant temporal differences in both females (Van der Waerden test; F(13) = 42.6; p < 0.001) and males (Van der Waerden test; F(13) = 38.4; p < 0.001). Briefly, the gonad index increased from April–May to February–March of the following year, and then dropped abruptly, due to spawning. Unexpectedly, low values were, however, measured in January of both years (Fig. 2).

Fig. 2
figure 2

Temporal changes in the gonad index of female () and male () Holothuria forskali sampled in the Glénan Islands, France, from December 2019 to July 2021. n = 2–6 females and 1–6 males for each sampling event. Error bars: ± SD

Fatty acids

A total of 35 FA were identified using the GC-FID/MSD system in all samples. Highest total FA concentrations were measured in female gonads in May 2021 (90.5 ± 8.4 mg g−1 dw) and lowest values were recorded in male gonads in April 2021 (23.4 ± 0.4 mg g−1 dw; Fig. 3). Females exhibited significantly higher concentrations of FA in gonads than males (Welch test; t(74) = 14.1; p < 0.001). Significant temporal differences in total FA concentrations were revealed in females (Van der Waerden test; F(10) = 27.7; p < 0.01) but no difference could be highlighted in males (Van der Waerden test; F(10) = 15.4; p = 0.12).

Fig. 3
figure 3

Temporal changes in the concentration of fatty acids in the gonads of female () and male () Holothuria forskali sampled in the Glénan Islands, France, from December 2019 to July 2021. n = 3–5 females and 1–5 males for each sampling event. Error bars: ± SD

The permutational multivariate analysis of variance on relative FA compositions revealed significant differences in the FA assemblages between females and males and for at least one sampling event (PERMANOVA; Fsex = 388.2; p < 0.001; Fdate = 9.6; p < 0.001; Finteraction = 4.0; p < 0.001). As the interaction between factors was significant, both sexes were compared separately in one-way PERMANOVA. Both males (F(10) = 10.1; p < 0.001) and females (F(10) = 4.5; p < 0.001) showed significant temporal variations in their FA assemblages. The PCoA on Bray–Curtis dissimilarity matrix distributed the samples on a two-dimensional representation following a typical arch-shaped configuration with female gonads sampled from July to March at one extremity of the gradient (cluster 1; Fig. 4) and male gonads sampled from September to March at the other extremity (cluster 3). Females of cluster 1 (43 individuals) were characterised by especially high levels of branched FA (BrFA) and 16:1ω7 (Table 2), while males of cluster 3 (26 individuals) were marked by especially high levels of long-monounsaturated FA (LC-MUFA; MUFA ≥ 20 carbons atoms) and polyunsaturated FA (PUFA), particularly 20:4ω6 (ARA) and 20:5ω3 (EPA). Individuals of cluster 2 (7 females and 18 males) correspond to individuals in the reabsorption phase, soon after spawning (mostly samples from April in females and from May to July in males). They exhibited intermediate values for most FA and comparatively high levels of 22:6ω3 (DHA). The three FA or group of FA provided in Fig. 5 exhibited significant temporal differences in both sexes. Proportions of FA 16:1ω7 significantly varied for at least one sampling event in both females (Van der Waerden test; F(10) = 27.3; p < 0.01) and males (Van der Waerden test; F(10) = 35.2; p < 0.001). Proportions of branched FA significantly varied for at least one sampling event in both females (Van der Waerden test; F(10) = 24.3; p < 0.01) and males (Van der Waerden test; F(10) = 31.9; p < 0.001). Proportions of 22:6ω3 significantly varied for at least one sampling event in both females (Van der Waerden test; F(10) = 29.7; p < 0.001) and males (Van der Waerden test; F(10) = 33.2; p < 0.001).

Fig. 4
figure 4

Two-dimensional representation of the principal correspondence analysis of FA relative abundances in the gonads of female () and male () Holothuria forskali sampled in the Glénan Islands, France, from December 2019 to July 2021. Clusters were agglomerated according to the Ward method on a Bray–Curtis dissimilarity matrix. n = 3–5 females and 1–5 males for each sampling event

Table 2 Relative abundance of FA in the gonads of female and male Holothuria forskali sampled in the Glénan Islands, France, from December 2019 to July 2021
Fig. 5
figure 5

Temporal changes in the relative abundance of specific fatty acids in the gonads of female () and male () Holothuria forskali sampled in the Glénan Islands, France, from December 2019 to July 2021. n = 3–5 females and 1–5 males for each sampling event. Error bars: ± SD

Carotenoids

We separated six pigments using the HPLC system in all samples. Females exhibited significantly higher concentrations of carotenoids in gonads than males (Welch test; t(97.7) = 13.2; p < 0.001; Fig. 6), but no significant temporal variations in total carotenoids could be highlighted whether in females (Van der Waerden test; F(13) = 18.4; p = 0.14) or in males (Van der Waerden test; F(13) = 17.2; p = 0.19).

Fig. 6
figure 6

Temporal changes in the concentration of carotenoids in the gonads of female () and male () Holothuria forskali sampled in the Glénan Islands, France, from December 2019 to July 2021. n = 2–6 females and 1–6 males for each sampling event. Error bars: ± SD

The carotenoid assemblage was dominated by astaxanthin that was most probably present under two isomers (peak 1 and 2; Fig. 1). Yet, our system was not optimal for the quantification of this pigment, as shown by the long-tailed peak of the standard. As both peaks of astaxanthin could not be clearly distinguished we reported the sum of their concentrations. These two forms were also detected in the viscera of Holothuria atra where astaxanthin has previously been reported as the dominant carotenoid (Bandaranayake and Des Rocher 1999; Fig. 1). Traces of zeaxanthin and lutein were also detected but could not be quantified as both pigments eluted within the long tail of astaxanthin. Other clearly identified compounds were canthaxanthin (peak 3) and all-trans-echinenone (peak 6), while peak 4 and peak 5 were most probably isomers of the former and the latter. Peaks 5 and 6 were also detected in both Paracentrotus lividus and Psammechinus miliaris gonads where echinenone has been reported to dominate over all other carotenoids (Symonds et al. 2007, 2009).

The permutational analysis of variance on relative carotenoid compositions revealed significant differences in the carotenoid assemblages between females and males and for at least one sampling event (PERMANOVA; Fsex = 87.1; p < 0.001; Fdate = 3.6; p < 0.01; Finteraction = 2.8; p < 0.01). As the interaction between factors was significant, both sexes were compared separately in one-way PERMANOVA. Both males (F(13) = 3.1; p < 0.01) and females (F(13) = 2.2; p < 0.05) showed significant temporal variations in their carotenoid assemblages. Females were characterised by higher proportions of astaxanthin and echinenone (Table 3), while males were marked by higher relative abundances of the two forms of canthaxanthin. The range of variations in astaxanthin proportions was higher in males (47.2–86.5%) compared to females (70.0–86.6%) and increased proportions of astaxanthin were measured in males sampled from May to August (spawned and growing stages) compared to those collected the rest of the year (Welch test; t(53.0) = – 2.7; p < 0.01; Fig. 7). No significant differences were revealed in females between the same periods (Welch test; t(55.1) = – 1.4; p = 0.15).

Table 3 Sexual differences in the relative abundance of carotenoids in the gonads of female and male Holothuria forskali sampled in the Glénan Islands, France, from December 2019 to July 2021. Mean ± SD
Fig. 7
figure 7

Temporal changes in the relative abundance of specific carotenoids in the gonads of female () and male () Holothuria forskali sampled in the Glénan Islands, France, from December 2019 to July 2021. n = 2–6 females and 1–6 males for each sampling event. Error bars: ± SD

Discussion

Our results show that the fatty acid composition of sea cucumber gonads follows a seasonal cycle with major changes occurring in late spring/early summer, just after spawning and synchronously with the occurrence of the spring bloom in this area. However, we were unable to evidence a clear seasonal pattern in the carotenoid composition of H. forskali gonads, possibly because of the large inter-individual variability and the mixing of tubule classes.

Early partial spawns

The temporal variations in gonad indices are consistent with the annual reproductive cycle described by Tuwo and Conand (1992), a brief spring spawning (March–April), when the temperatures are increasing; a brief resting phase (May–June; not always observed in recent studies focusing on the same species; Santos et al. 2016; Ballesteros et al. 2021); a long reabsorption/maturation period (June–October) during the warm season and a long mature phase (November–March) during the cold season. Gutted weights were also very similar to those reported 30 years ago in the same location (Tuwo and Conand 1992). Yet, the small drops in gonad indices that were observed in January were not measured in the Glénan Islands 30 years ago. Santos et al. (2016) measured a similar decrease in December–January in Peniche-Portugal and Ballesteros et al. (2021) reported it in November in Galicia-Spain. In Peniche, this drop was attributed to evisceration due to especially stressful climatic conditions in this period (samples were picked up by hand at low tide), while in Galicia, it was attributed to early spawning events. As a small drop in gonad index was noticed for two consecutive years in January (at least for females), the hypothesis of stress was unlikely (storms do not necessarily occur at the same period in temperate regions and their effect at a depth of 8–12 m is strongly reduced) and we thus attributed declines to partial spawns followed by immediate gonad regeneration (Ballesteros et al. 2021), which is new for this location. Echinoderms are known to adapt their reproductive activity to local environmental conditions (Guettaf et al. 2000; Marquet et al. 2017) and we suggest that these early partial spawns were triggered by unusual and randomly occurring climatic conditions at this period (e.g. temperature increase, etc.). Yet, a 2-year survey as in our study is not enough to evidence an annual pattern and a specific trigger for such early partial spawns.

Sex-related biochemical variations

The sexual difference in both FA and carotenoid composition of sea cucumber gonads is consistent with existing knowledge. It has previously been evidenced in H. forskali and in Apostichopus japonicus regarding FA (David et al. 2020; Zhang et al. 2023) and in A. japonicus and Holothuria leucospilota regarding carotenoids (Matsuno et al. 1969; Matsuno and Ito 1971). As previously suggested by David et al. (2020), the twice as high total concentration of FA in female compared to male gonads (on average 72.4 ± 15.7 mg g−1 in female vs. 36.9 ± 7.9 mg g−1 in male gonads over all our samples) is likely due to lipid storage for future use as an energy source by eggs, while spermatozoids require energy only until fertilisation. Resource allocation may differ in both sexes as females maximise the storage of energy in oocytes, while males maximise the performance of spermatozoids. This would explain why males exhibited higher relative proportions of highly unsaturated FA (HUFA; ≥ 20 carbon atoms and three double bonds) that are FA of high metabolic significance in gonads compared to females (46.5 ± 2.0 in males of cluster 3 vs. 26.9 ± 2.5% in females of cluster 1). Zhang et al. (2023) recently evidenced that the fatty acid synthase gene was up-regulated in the intestine of males A. japonicus compared to females and speculated that higher levels of PUFA in male gonads might be related to intestinal metabolism. Sexual differences in gonad FA proportions were also shown for the sea urchins Arbacia lixula, A. dufresnii, Paracentrotus lividus, Strongylocentrotus intermedius and Sphaerechinus granularis (Martinez-Pita et al. 2010; Zárate et al. 2016; Díaz de Vivar et al. 2019; Rocha et al. 2019; Wang et al. 2019; Lourenço et al. 2022) and sex-related differences in the relative abundance of carotenoids were revealed for P. lividus, Psammechinus miliaris and Strongylocentrotus droebachiensis (Symonds et al. 2007, 2009; Hagen et al. 2008; Rocha et al. 2019). In line with our hypothesis of different allocation of FA in males vs. females gonads, Zárate et al. (2016) measured higher proportions of structural lipids (that we expect to be essentially allocated to metabolic functions and thus present higher levels of HUFA) in testes and higher levels of energy lipids (with more generalist FA compositions and lower levels of HUFA) in ovaries of the sea urchin Arbacia dufresnii. However, although the proportions of some FA vary according to sex in sea urchins (Zárate et al. 2016), differences in HUFA proportions were never as marked as for H. forskali in the present study. Finally, higher levels of carotenoids in female gonads suggest that such compounds play a role in embryogenic and/or larval development rather than in fertilisation (e.g. photoprotection, antioxidant, energy reserve).

Seasonality in FA profiles

Our results show that the FA composition of sea cucumber gonads follows a seasonal cycle and that the magnitude of sex-related differences seems to changes on a seasonal basis (i.e. gonads strongly differ according to sex in autumn/winter while both sexes are more similar in late spring/early summer). In addition, early partial spawns in January of both 2020 and 2021, necessarily modifying the balance between maturing, resorbing and fecund tubules had negligible influence on the biochemical composition of gonads, suggesting that diet during the reabsorption phase is rather the main determinant of gonads FA composition in late spring/early summer. Data interpretation is, however, to be done with care as only two years were analysed and the dataset presents various gaps. It would have been ideal to sample food sources along with sea cucumbers to evidence that seasonal changes are related to diet. Unfortunately, H. forskali lives on hard surfaces and although sand in found in its gut, it is known to select its sediment particles, leaving its diet rather unclear when bulk sediment is studied (David et al. 2020). Individuals sampled in spring, soon after spawning, did not show a clear sexual separation according to the FA profile of their gonads, as shown by their proximity on the two-dimensional representation of PCoA (Fig. 4; cluster 2). Contrastingly, individuals sampled in autumn and winter exhibited a clear sex separation (Fig. 4; cluster 1 vs. cluster 3). We suggest that sea cucumbers with nearly empty gonads take advantage of the phytoplankton spring bloom to rapidly accumulate nutrients in the form of lipids in their gonads on an opportunistically basis, thus explaining the higher variability of FA profiles among individuals within a same sex-sampling event. The effect of such fresh inputs is especially notable in males where total FA content is lower. Increased levels of FA 16:1ω7, branched FA and 22:6ω3 in May and July of both years are actually, respectively, markers of a phytoplankton bloom (especially diatoms), the bacterial degradation of freshly produced organic matter and carnivory on filtering invertebrates (Dalsgaard et al. 2003; Hughes et al. 2005; Arafa et al. 2012). Indeed, the French phytoplankton monitoring network (REPHY) that surveys water parameters near Concarneau since 2004 (station 047-P-016; 13 km away from our sampling site) reports a peak in chlorophyll-a concentrations from 0.2–1.8 μg L−1 to 5–9 μg L−1 generally occurring during a few weeks between March and May (https://cutt.ly/uLrJpoG; REPHY 2021). This pulse to the ecosystem may not only favour larval development, as generally observed in marine invertebrates (Starr et al. 1990; Venâncio et al. 2022), but also the replenishment of adults energetic reserves through the opportunistic consumption of sinking organic matter (e.g. phytodetritus, microbial aggregates, dead larvae) and/or recently fixed invertebrates larvae (especially for this species living on hard substrates). Sex differences are reduced to their minimum in cluster 2 (gonads in the reabsorption phase) and the FA compositions of gonads nearly reflect that of H. forskali’s expected food resources, which do not differ between sexes (David et al. 2020). The close proximity of HUFA ratios between the digestive tract and the gonads that we previously highlighted on samples collected in May 2019 also suggests that FA are poorly processed before being stored in the gonads (David et al. 2020). Echinoderms are known for their ability to desaturate and elongate saturated FA to HUFA (Kabeya et al. 2017; Liu et al. 2017). Once rapidly stored in the gonads during the period of large resources availability, we suggest that FA are slowly elongated, desaturated and probably rearranged within lipids classes according to the specific requirements of both sexes, leading to increasingly marked sex differentiation in the gonads FA composition and reduction in compositional variability. We believe that a selective turnover of FA in favour of HUFA may also act synergistically with elongation/desaturation to allow the gonads to progressively reach the certain FA profile that is necessary for reproduction, allowing diet manipulation in aquaculture conditions.

Constancy in carotenoid profiles

We expected the carotenoid composition of sea cucumber gonads to vary on a seasonal basis. Instead, we could not establish a clear link between the carotenoid composition of female gonads and their reproductive status and/or the period of the year. This absence of pattern may have been caused by the averaged values obtained from the mixing of tubule classes; especially in recently spawned gonads (i.e. May 2020 and April 2021) where nearly transparent primary and secondary tubules (T1 and T2) were mixed with strongly pigmented spent tubules (T5; see pictures of gonads before and after spawning in supplementary file S1). Further studies on sea cucumbers fitting to the tubule recruitment model would have to separate tubule classes to obtain a more detailed description of the carotenoid dynamics. Yet, males H. forskali showed temporal trends in the relative abundance of carotenoids that could be related to the gametogenic cycle, with a slight increase in the relative proportion of astaxanthin during late spring/early summer, when gonads were in their growing stage. The generally admitted route for astaxanthin synthesis, consistent with carotenoids we identified in the gonads, is β-carotene—> echinenone—> canthaxanthin—> adonirubin—> astaxanthin (Maoka 2011; Prado-Cabrero et al. 2020). Astaxanthin being a terminal product of this route, it is highly probable that its relative increase in nearly empty gonads is due to preferential retention during spawning rather than opportunistic dietary inputs. Carotenoid retention in males was previously evidenced in the sea urchin Strongylocentrotus droebachiensis (Hagen et al. 2008). Animals do not synthesise carotenoids de novo. They must accumulate them directly from food or partly modify them through metabolic reactions (Maoka 2011). Yet, none of the intermediate products of astaxanthin biosynthesis were found in substantial proportions in the gut contents of H. forskali (David et al. 2020), suggesting that their acquisition through dietary inputs occurs synchronously with gonads filling and/or through the reabsorption of spent tubules (T5). Furthermore, the predominance of one carotenoid over the others in sea cucumber gonads (cantaxanthin in Parastichopus japonicus, astaxanthin in Holothuria leucospilota, H. atra and H. forskali, and cucumariaxanthin in Cucumaria japonica and C. echinata; Matsuno and Tsuchima 1995; Bandaranayake and Des Rocher 1999) raises questions about the specific function of these carotenoids in reproduction. Although the absorption spectra and molecular structures of these compounds are almost similar, suggesting similar photoprotective and energy reserve roles, the oxidative property of astaxanthin is tenfold greater than that of other carotenoids such as canthaxanthin (Galasso et al. 2017). Whether one or another carotenoid brings a particular advantage for the reproduction of a given species and thus could induce a selective pressure in the evolutionary lineage of sea cucumbers remains to be elucidated.

Potential implications for aquaculture

It appears that H. forskali stores nutrients (i.e. lipids) in its gonads primarily during the warm spring/summer months. Then, follows a long maturation period before the onset of spawning in early spring. In September, gonads are already nearly at their maximum size and exhibit a relatively stable FA and carotenoid composition. At this period, gametogenesis has already been launched and all results (gonad index, FA and carotenoid assemblages) suggest that the species is waiting for warmer waters of the following spring to initiate the final stage of maturation and release gametes. Partial spawns in November in Galicia-Spain actually support the hypothesis that sea cucumbers are ready to spawn already in autumn (Ballesteros et al. 2021). Broodstock could possibly be captured at this moment (September–October) and stored in controlled conditions until the production of larvae is required. This approach could help to extend the breeding season that is now restricted to early spring (Laguerre et al. 2020). Whether controlling dietary intake (i.e. lipids) during a few weeks/months before spawning induction can improve the reproductive success of H. forskali, as it is the case for other groups of marine animals such as crustaceans (e.g. Palacios et al. 2001; Yoshida et al. 2011) and fishes (e.g. Zakeri et al. 2011; Liang et al. 2014) will depend on the turnover of nutrients within the gonads during this maturing period. Indeed, feeding Cucumaria frondosa with fish eggs during the 3 months preceding expected spawning period increased oocyte maturity index and fuelled gonads with higher levels of essential FA compared to natural populations, but did not trigger spawning (Gianasi et al. 2017). In contrast, feeding them with diatoms, despite not exhibiting the best growth, organ indices or lipid profiles effectively provided the adequate stimulus to trigger natural spawning (Gianasi et al. 2017). An experiment conducted with Parastichopus japonicus concluded that females fed with the green algae Tetraselmis sp. had higher fecundity, but showed reduced larval survival and lower SFA and HUFA concentrations in eggs relative to females fed with the diatom Thalassiosira sp. (Whitefield et al. 2018). Modulation of carotenoids in broodstock diet was not challenged yet in sea cucumbers but larvae of the sea urchin Lytechinus variegatus from parents fed xanthophylls (i.e. zeaxanthin and lutein) were larger throughout development, developed faster, had higher survival rates and attained metamorphic competence faster than those having received vitamins or vitamins and β-carotene (George et al. 2001).

The ideal situation for hatcheries would be to maintain broodstock in controlled conditions for consecutive years. Our results evidencing a short period of nutrient loadings in the gonads during the spring bloom followed by a long maturation period suggests that the few months of warm temperature (i.e. late spring/summer) may be crucial in terms of nutrition to launch gametogenesis and switch on the conveyor belt of tubule recruitment. It has been hypothesised in Parastichopus californicus that germ cells of year N migrate on the gonad basis to become the primary tubules of year N + 1 and consecutively until spawning after four years of maturation (Smiley 1988; Sewell et al. 1997). This hypothesis has strong implications for aquaculture as the reproductive success of H. forskali, that fits to this tubule recruitment model (Tuwo and Conand 1992), may be affected by life history of broodstock during the 4 years preceding reproduction. Further studies will have to elucidate the dynamics of tubule recruitment and search for potential triggers of gametogenesis to handle the complete cycle of reproduction in H. forskali.