The development of a new candidate species for aquaculture requires control of the breeding cycle to ensure a consistent supply of fertile gametes. This initial investigation of captive reproduction in wild caught female YBF indicates that treatment with exogenous GnRHa successfully activates the reproductive axis in comparison to sham treated, control fish.
Only GnRHa treated fish were found with hydrated oocytes or observed to have ovulated. Injection with at least 50 µg/kg GnRHa stimulates FOM in wild caught female YBF. Interestingly, a higher dose of 100 µg/kg of GnRHa did not significantly stimulate FOM. Latency in the single fish which did ovulate in this higher dosage group, was considerably longer (28 days) compared to those in the 50 µg/kg GnRHa treatment (less than 6 days). While fertilisation rates were highly variable, both treatment groups yielded egg batches with ≥ 80% fertilisation, in contrast with low fertilization rates reported from hormonally induced spawning in some flatfish, e.g. southern flounder (Paralichthys lethostigma) (Watanabe et al. 2001; Wright-Moore et al, 2019). The reason for the increased latency of the higher GnRHa dose is unknown but could relate to temporary desensitisation of the GnRH receptors (GnRH-r) on the pituitary gonadotropic cells. Our recent work shows that pituitary luteinising hormone beta-subunit (lhb) expression is upregulated within 24 hours in YBF following injection with either 25 µg/kg or 50 µg/kg of GnRHa but not with 100 µg/kg GnRHa (unpublished data). Continuous exposure to GnRH has been demonstrated to desensitise pituitary LH release in both goldfish (Carassius auratus) and tilapia hybrids (Oreochromis niloticus x O. aureus). Levavi-Sivan et al. (2004) found that elevated doses of GnRHa dramatically decreased GnRH-r mRNA levels in comparison to lower doses in tilapia although plasma LH was not affected. These observations pose intriguing physiological questions with a view to the optimal dosage of GnRHa in YBF.
Hormonal therapy often needs fine tuning to develop an optimised protocol. While injection with 50 µg/kg GnRHa stimulated the progression of oocytes into advanced stages of FOM (GVBD and hydration), this did not necessarily result in ovulation. Studies of induced reproduction in other species of fish have also resulted in a portion of individuals ovulating following GnRHa treatment (Berlinsky et al. 1997; Poortenaar and Pankhurst 2000; Setiawan et al. 2016). The efficacy of exogenous hormone treatment is subject to a number of influencing factors, such as stage of ovarian development, presence of endogenous hormonal inhibition, as well as the delivery method of the therapeutant and dosage (Chang and Peter 1983; Cerdà et al. 1997; Forniés et al. 2001; Zohar and Mylonas 2001; Levavi-Sivan et al. 2004). The identification of an optimal dosage is critical to maximise ovulation in YBF and further refinement is required. The fact that some hormone-treated fish failed to enter FOM, may reflect insufficient ovarian development at the time of GnRHa administration (Berlinsky et al. 1997; Mugnier et al. 2000; Wright-Moore et al. 2019). The presence of both PVO and late stage atretic oocytes in many of these fish suggests that they were either post-spawned or had entered stress-related atresia, following capture. Breeding in wild YBF is reported to occur between early winter and early summer (June – December) (Colman 1973). It is yet to be reported as to whether individual fish spawn throughout this period or at specific times within this seasonal window. In the latter scenario, it is likely that any given group of wild caught fish will contain at least some individuals that are not in breeding condition. Based on the initial size of oocytes in the individuals that successfully responded to treatment; it is recommended that GnRHa is administered when oocytes are approximately 300 µm diameter. Oocytes of this size are likely to correspond to late-vitellogenic or early maturational stages of ovarian development. This is when FOM is typically induced via LH mediated stimulation of the MIS (Nagahama 1994; Nagahama and Yamashita 2008).
The failure of control fish to complete FOM to yield hydrated or ovulated oocytes suggests that wild caught YBF are, at least initially, prone to reproductive dysfunction in captivity. This is a common result of stress which has been observed in other wild caught broodstock fish (Corriero et al. 2021; Imanaga et al. 2014; Mehdi and Ehsan 2011). Failure to enter FOM is one of the most common forms of reproductive dysfunction in captive fish and typically responds to GnRHa treatment (Lim 2016; Prat et al. 2001; Rosenfeld et al. 2012). While hormonal interventions have been associated with reduced gamete quality in some fish (Mugnier et al. 2000; Forniés et al. 2001; Garber et al. 2009; Setiawan et al. 2016), high fertilization rates (> 80%) were achieved in this study indicating that treatment with 50–100 µg/kg GnRHa can yield good quality egg batches.
Milt production in males may prove to be a limiting factor for the development of artificial fertilisation protocols for YBF. Comparatively small volumes (< 0.5 ml) of milt were able to be collected from individuals in this study and production did not appear to be greatly enhanced with GnRHa treatment. It is important to note that this observation is anecdotal as the study primarily focused on induced reproduction in female YBF. Pankhurst and Poortenaar (2000) recorded similar volumes of milt in greenback flounder (Rhombosolea taparina) and concluded that human choriogenic hormone (hCG) and GnRHa both increased milt volume relative to saline controls. Low efficacy of GnRHa treatment on milt production was reported in male summer flounder (Paralichthys dentatus; Berlinsky et al. 1997) and Senegalese sole (Solea senegalensis; Guzmán et al. 2010). Evidence indicates that dopamine inhibition within the reproductive axis may suppress GnRHa efficacy in male Senegalese sole (Guzmán et al. 2010). The co-administration of a dopamine antagonist and GnRHa was effective at increasing spermiation and milt volume in Senegalese sole (Guzmán et al. 2011). Methodologies to increase expressible milt volumes in YBF may warrant further investigation to avoid future bottlenecks in hatchery production.
Reproductive behaviours appeared to be absent during the study and there was no evidence of spontaneous spawning occurring despite some of the GnRHa treated fish having ovulated. Similar results occur in captive Atlantic halibut (Hippoglossus hippoglossus) where hand-stripping is necessary (Skaalsvik et al. 2015). This may relate to the requirement of specific husbandry or environmental conditions. Pankhurst and Fitzgibbon (2006) found that captive greenback flounder undergo spontaneous spawning in the mid-water column when held in a low disturbance environment at reduced densities (~ 1kg/m3). Broodstock fish in the current study experienced regular handling disturbance and were held at greater stocking densities. While wild YBF are often caught on shallow mudflats in less than one metre of water, spawning typically occurs at depths of 12-30m (Colman 1973; Smith et al. 1999). The depth of the tanks used in the present study (0.6 m) may have been inadequate to support natural courtship and spawning. Future work should consider adapting the tank environment to encourage natural reproduction.
Few studies have focused on the reproductive biology of YBF although previous reports state that they seasonally spawn a single batch of eggs (Colman 1973; Mutoro 2001). The existence of distinct cohorts of oocytes at various stages of development (including vitellogenic and FOM) as observed in this study, suggests a different reproductive strategy. Moreover, the observation that some fish ovulated more than once, strongly indicates that YBF are multiple group synchronous batch spawners as suggested by Koverman (2018). A similar pattern of ovarian development also exists in greenback flounder (Poortenaar and Pankhurst, 2000) as well as many other temperate flatfish (Berlinsky et al. 1997; Earl 2014; Lim 2016; Poortenaar et al. 2001; Tingaud-Sequeira et al. 2009).