Development of larval fish rearing techniques and nutrient requirements for the green mandarin, Synchiropus splendidus: a popular marine ornamental fish

The green mandarin fish, Synchiropus splendidus (Herre, 1927) is a small, brilliantly coloured benthic marine fish distributed in the tropical Pacific and Indian Oceans. It is within the group of fish known as dragonets (Family: Callionymidae) and is among the most desirable species to marine aquarium hobbyists worldwide. Unfortunately, the current supply of S. splendidus for the aquarium trade comes solely from the wild. Moreover, it has been observed that most specimens on sale are males, suggesting probably selective harvesting from the wild populations. As a result, the captive breeding of S. splendidus is urgently required as an alternative to ensure a sustainable supply of this popular species for the aquarium trade to reduce the pressure on its natural populations.\ud \ud S. splendidus is a pelagic-spawner and has been considered as a relatively hard-to-breed species due to its adults generally feeding on live small crustaceans, hence this species is difficult to maintain under captive condition. In addition, there is limited knowledge on its larval feeding habitat. Newly hatched larvae of S. splendidus is among the smallest larvae of marine fish, on average only 1.5 mm. The high vulnerability of pre-feeding larvae to handling stress and advent environmental conditions make their rearing even more difficult.\ud \ud A series of experiments were conducted to improve captive breeding techniques for S. splendidus, which could also serve as a model species for ornamental marine pelagic-spawners. The present thesis consists of 9 chapters: following the first general introduction chapter (Chapter 1), the second chapter describes general materials and methods used (Chapter 2). The subsequent 6 data chapters can be largely grouped into three main themes: 1) Chapter 3, 4 and 5 focused on the first feeding preys and culturing regimes for newly hatched and early larvae of S. splendidus, and investigated the underlying mechanisms for the superior performance of copepods as live prey from the perspectives of behavioral and lipid nutrition. 2) Chapter 6 then focused on the strategy and best timing for larval prey transition from copepods/rotifers to Artemia nauplii, and subsequently to larger enriched metanauplii during the critical metamorphosis period of S. splendidus. 3) Chapters 7 and 8 finally focused on optimizing the growth rate of post-settlement S. splendidus by their feeding with enriched Artemia metanauplii containing graded concentration of two key dietary LC-PUFAs, DHA and ARA. The final chapter (Chapter 9) summarizes the main results from all data chapters and posits implications for future research directions.\ud \ud The experiments from Chapter 3 confirmed that S. splendidus newly hatched larvae fed on rotifers (10 mL⁻¹) alone had very low survival (3%) while the addition of the calanoid copepod Parvocalanus crassirostris at 2 mL⁻¹ for co-feeding with the rotifers dramatically improved both survival (>50%) and growth. A further rearing experiment feeding larvae with different copepod densities showed that the copepods provided at 1 mL⁻¹ for co-feeding with rotifers could produce compatible results to those at the higher density of 2 mL⁻¹. A subsequent experiment demonstrated that co-feeding rotifers with the copepods at different tested densities did not produce any clear beneficial effects, suggesting co-feeding rotifers is not necessary. The subsequent larval ingestion rate experiment showed that the copepods were always positively selected over the rotifers by S. splendidus larvae at all ages tested (4, 6, 8 and 10 DPH). Based on these results, an optimal feeding regime for S. splendidus early larvae by feeding copepods solely at 1.0 mL⁻¹ was established.\ud \ud Results presented in Chapter 4 showed that compared to the continuous copepod feeding control, larval prey shifted from copepods to rotifers on 4 or 8 DPH, leading to significantly lower survival and growth, whereas if the prey shift occurred later on 12 DPH, no significant difference in both survival and growth were detected at the end of the experiment on 15 DPH. The larval feeding behavior experiment showed that larvae of all ages tested (6, 8, 10 and 12 DPH) attacked copepods more frequently than rotifers, showing a strong preference toward copepods. Moreover, copepods were never rejected after being captured by larvae of all ages, whereas rejection following capture was commonly observed for rotifers. Furthermore, the feeding intervals (i.e. the time between a prey being ingested and the time of the next active foraging for food by a larva) on copepods by larvae of all ages were also significantly shorter than for rotifers.\ud \ud The fatty acid analysis results (Chapter 5) demonstrated that the copepod P. crassirostris possessed a superior fatty acid profile compared to the rotifers, as P. crassirostris showed much better matched MUFA, PUFA and LC-PUFA profiles as well as DHA/EPA ratio to S. splendidus newly spawned eggs. It was also noted that of the most important essential fatty acids, DHA and ARA in the 2 DPH pre-feeding larvae were approximately only half that in the newly-released eggs (p<0.001), whereas EPA remained relatively stable (p>0.05), suggesting dietary supply of DHA and ARA are likely to be important for subsequent larval survival and development.\ud \ud Larval feeding experiments described in Chapter 6 showed that larvae had their prey switched from the copepods to Artemia nauplii on 18 and 21 DPH without a transitional rotifers feeding period, giving the highest larval survival compared to other feeding treatments. In particular, introducing a rotifer feeding period starting at 12 DPH led to significant lower larval survival when compared to the treatments in which larval prey switched from the copepods to Artemia nauplii directly on 15, 18 and 21 DPH, respectively (p<0.01); however the growth performance of larvae from different treatments were not significantly different. Nevertheless, for the larvae fed rotifers from 12 DPH, if rotifers were switched to Artemia nauplii on 18 or 21 DPH, larval survival were still reasonably high (>70%), which appears acceptable in the case of commercial production. Hence, in the case of limited copepod supply, which commonly occurs since copepod intensive culture techniques are still under development and generally very costly, rotifers might be used to substitute copepods for feeding larvae between 12 and 17 DPH.\ud \ud It was further demonstrated that among different treatments of prey switched on 22, 25 and 28 DPH in both live and dead forms, introducing live enriched Artemia metanauplii to replace the Artemia nauplii on 25 DPH obtained the best survival and growth of post-settlement S. splendidus on 42 DPH. On the other hand, those fed dead enriched Artemia from 22, 25 and 28 DPH had significantly lower survival and growth compared to those fed live enriched Artemia on corresponding days.\ud \ud Chapters 7 and 8 investigated the optimal dietary DHA and ARA for optimizing the growth of post-settlement S. splendidus from 25 DPH. The results showed that no single mortality was observed in any treatments feeding enriched Artemia containing graduated levels of DHA or ARA, respectively, at the end of the experiments on 45 DPH. However, the growth performance in terms of standard length (SL), body width (BW) and dry weight (DW) were significantly affected by dietary DHA level; and BW, DW but not SL by the dietary ARA level. The regressions of SGR data with dietary DHA level suggested that dietary DHA level for the highest SL, BW and DW was 6.91, 6.68 and 6.48 mg DHA/g DW, respectively. Meanwhile, for the dietary ARA, it was 7.29, 7.06 and 7.34 mg ARA/g DW, obtaining the highest SL, BW and DW, respectively. It is unexpected that the analysis of fatty acids profiles of the 45 DPH old S. splendidus from different treatments indicated that they might possess limited capacity of biosynthesizing LC-PUFAs from LOA or ALA, which is not common among marine fish larvae