Part 1: Artemia Cyst Production in Solar Earthen Ponds in Vinh Chau, Vietnam, and Its Associated Microbiome

Artemia franciscana farmed in Vinh Chau, Vietnam, have been known for their premium-quality cysts. This study provides an update on Artemia cyst production in solar earthen ponds in Vinh Chau and explores the microbiome of Artemia gut, pond water, and pond sediment in order to generate clues for research on cyst productivity improvement. We monitored cyst productivity of two pond groups as follows: a more productive one (HIGHER) and a less productive one (LOWER); and conducted two samplings, one mid-crop and the other at the end for microbiome analyses. Cyst productivity averaged at 112.0±22.3kg/ha/crop with large variation among the studied ponds. The HIGHER group produced 160.6±10.3kg/ha/crop or 2.5 folds significantly higher than that by the LOWER group (63.2±18.7kg/ha/crop) (p<0.05). Higher feeding rate and fertilization rate were strongly associated with higher cyst productivity (p<0.05). The microbiota of Artemia gut had the lowest diversity compared with pond water and pond sediment. Microalgae, such as diatoms and Dunaliella, halophilic bacteria (Halomonas spp), and Vibrio were the dominant taxa in Artemia gut. Their relative abundances changed over time as the result of increasing salinity and extreme water temperatures. Fungi were present in Artemia gut and pond sediment but did not appear to be important to Artemia. Bacterial communities were very different among Artemia gut, pond water, and pond sediment. The highest diversity was recorded for pond sediment, followed by pond water and Artemia gut. Ponds with higher cyst productivity tended to have higher bacterial diversity in pond sediment. More significant temporal variations were found with the bacterial communities of pond water compared with pond sediment and Artemia gut. Overall, the observed differences in microbiota of Artemia gut, pond water, and pond sediment between the two pond groups (or cyst productivity levels) emphasized the importance of microalgae as the major food source, interesting roles of Vibrio and pond management.

1. Introduction

Artemia or brine shrimp, has been known as an essential input for aquaculture hatcheries thanks to their good nutrition, uniform size, and attractiveness to many species at young stages [1–3]. Under extreme conditions, often a combination of high salinities, high temperature, and lack of food, Artemia produce diapause cysts with hard outer shells [4]. Dry dormant Artemia cysts can be stored for a long time and incubated in seawater to become nauplii for larval rearing of high-value species [5]. In commercial hatcheries Artemia nauplii can be fed to marine larvae either directly or after being enriched with commercial products or selected microalgae [1, 6]. Artemia nauplii naturally have high α-linolenic acid (ALA) and linoleic acid (LNA) concentrations, and relatively low arachidonic acid (ARA), eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) [7]. These essential fatty acids are important for larvae development since fish do not have the ability to synthesize them [8, 9]. For penaeid shrimps, Artemia nauplii has been the main live prey supplied to Mysis and postlarvae (PL) stages and clearly outperforms commercial microencapsulated diets [10, 11]. Furthermore, the amount of Artemia cysts used directly impacts PL quality. Higher Artemia inclusion levels generally improve PL survival, growth, and overall performance, especially during early stages. Artemia enrichment with HUFA can further improve PL robustness [12–14]. It is estimated that the current demand for Artemia cyst is 2000–3500mt/year with an average value of US$100/kg [15, 16]. In Australia, Artemia cyst production had been trialed in Western Australia in the early 1990s but failed to progress into commercialization [17]. A small number of cysts are still being collected as a by-product from salt pans in Western Australia and Queensland (Clive Keenan, pers. comm.).

Artemia cyst production has been heavily relied on wild collection, mostly in the United States and partly in other countries, such as Iran, Russia, and China [3, 18]. Each strain of cyst has a distinct quality, which is likely due to geographical isolation and peculiar habitat conditions [19, 20]. Such variations in quality are due to. Artemia cyst production is considered advanced in Vietnam and Thailand [21, 22], but production remains marginal due to low productivity and increasing competition of land use by shrimp farming, which is a lot more lucrative [23]. In 2023, Vietnam produced only 2.8% of the country’s demand for Artemia cysts for marine shrimp/fish hatchery production [24]. Interestingly, the cysts produced by Artemia franciscana in Vinh Chau, Vietnam have been considered the best globally, reflected by price and limited supply. It is widely known as Vinh-Chau Artemia [25]. Vinh Chau Artemia are uniquely smaller sizes with good nutrition and a high hatching rate [22,23,25]. Hatchery operators have a clear preference for Vinh-Chau Artemia over many other available brands in the market (Trinh Trung Phi, Nguyen Van Hoa pers. comm.).

This species originated from San Francisco Bay, USA, was first introduced to Vinh Chau in 1982 via an FAO project to improve the livelihood of artisanal farmers [26]. Since then, Artemia has been farmed during the dry season, often from late November to early June of the following year, or as soon as the wet season begins [25]. Cysts can be collected around 2 weeks from stocking. Salinity is maintained between 80 and 120ppt throughout the crop to facilitate cyst production. Chicken manure in combination with inorganic fertilizers is commonly used to bloom pond water, providing natural food for the farmed Artemia. Supplementary feeding by shrimp feed crumbles has been found beneficial to the farmed Artemia[25]. While the translocated A. franciscana has adapted very well to the hot environment of Vinh Chau[23], there is no explanation yet for the excellent quality of their cysts. More than 80% of ectotherm species like Artemia grow to a smaller body size when they are reared under warmer conditions [27] at an estimated magnitude of size reduction of 4.0%/°C [28]. While the highest yearly water temperature in San Francisco Bay, where A. franciscana is originated from only reach 20–24°C, water temperature in Artemia ponds in Vinh Chau can be 29.5–32.9°C in the morning and 34.5–40.6°C in the afternoon (Le &Hoa). [29] reported that cysts of Vinh Chau Artemia are smaller in size with a thicker outer shell compared with those produced by San Francisco Artemia. Therefore, the extreme heat in Vinh Chau is likely the most noticeable and influencing factor that has driven some unknown processes which eventually make Vinh Chau Artemia cysts highly preferred for aquaculture hatchery production regarding size. Lower dis solved oxygen (DO) levels in Artemia ponds because of higher Artemia densities can be another factor that limits Artemia growth and their body sizes or cyst size, as hypothesized by Deutsch et al. [28]. But apart from these two abiotic factors, what else in pond water or pond sediment in Vinh Chau helps create the unique values of A. franciscana farmed in this area?

Many studies have demonstrated the importance of micro algae, bacteria, and recently fungi to Artemia [25, 30, 31]. Artemia are nonselective feeders. They feed on small food particles such as microalgae, bacteria, detritus, and small organisms ranging in size from 1 to 50µm [32]. Among these, microalgae is considered as an indispensable feed source for Artemia [25, 31]. Feeding Artemia with carotenoid-rich microalgae, such as Dunaliella salina, Haematococcus pluvialis, Isochrysisgalbana, and Myrmeciaincisa can improve the growth and health conditions of Artemia, resulting in a higher survival rate, larger body size, and total antioxidant capacity [30]. As microalgae communities vary significantly among locations, their biodiversity and composition can strongly influence Artemia quality. Previous studies have shown a great level of diversity of microalgae in Artemia production ponds located in the Mekong River Delta of Vietnam, including Bacillariophyta, Dinophyta, Cyanobacteria, and Chlorophyta [22, 33]. Bacillariophyta is the dominant group with up to 174 different species. However, none was conducted to relate microalgae composition with cyst productivity or to understand the microbiome of Artemia guts, pondwater, and sediment.

Regarding industrial growth, Artemia cyst production in Vinh Chau has remained marginal despite its 30+ years of development. Generally, both farming area and production have been steadily declining over the last 10 years. In 2023, only 12 tons of Artemia cysts were harvested from Vinh Chau (Directorate of Fisheries 2024). Cyst productivity varies significantly from farm to farm ranging from mostly 30–40kg/ ha/crop to 300kg/ha/crop, indicating opportunities for significant improvements. Over the last few years, extreme weather conditions and discharges from nearby superintensive shrimp farms have been found associated with reduced cyst productivity of farmed Artemia in Vinh Chau (Mr Chym Suol, pers. comm.). Mass mortality of Artemia has occurred more often and it is getting harder to bloom pond water and maintain the bloom.

In this study we reported productivity of Artemia cyst pro duction in Vinh Chau and investigated the microbiome of Artemia gut, pond water, and pond sediment in two different Artemia pond groups. The new findings are expected to shed more lights on the current status of Artemia farming and direct future research effort, which eventually facilitate stronger development of Artemia farming for the benefit of the aquaculture industry.

2. Materials and Methods

2.1. Research Location

Eight commercial Artemia ponds in Lai Hoa Ward, Vinh Chau District, Soc Trang Province, Vietnam (9°15′42.6″N 105°49′58.1″E) were selected for this study. These ponds were constructed in the early 1990s and have been used solely to produce Artemia cysts over the last 15 years. Pond size was relatively similar, ranging from 0.3 to 0.5ha. Pond preparation was conducted a month before stocking by tilling the soil, liming, and removing unwanted organisms. Water was pumped into a series of shallow solar ponds to increase salinity from 10ppt to 70ppt over a period of 3 weeks. Artemia nauplii hatched from cysts for 24h were then stocked into these production ponds in the late afternoon between 17:00 and 19:00. Stocking rate was 850g of cysts/ha. During the crop salinity was maintained within a preferable range for cyst production, that is, 80–120ppt by compensating water loss due to evaporation with new low-salinity water daily. To bloom pond water and maintain the algal bloom, chicken manure was added in the morning at a rate of 100–150kg/ha/day. Shrimp feed crumbles were provided in the evening at 3.0–4.0kg/ha/ day as a supplementary feed. Data on fertilization and feeding rates were collected and recorded. Cysts appeared 14–15 days from stocking. Cyst harvesting started from the 18th day of culture (DOC18) and lasted for 75 days. The amount of fresh cysts daily harvested from each experimental pond was measured and recorded. Harvest data as recorded by the farmers in the first 53 days were used to group these ponds into two groups: HIGHER (1.7–2.6kg/day) and LOWER (0.3–1.4kg/ day) regarding cyst productivity (Table 1). The harvested cysts were soaked in 120ppt salt water to harden their outer shells for at least 1 month before being dried and canned.

Note: Data are means ±S.E.M. (n=53).

2.2. Sampling Design and Collection of Samples.

Two samplings were conducted as follows: the first one around midcrop ontheMarch,192024andthesecondoneattheterminationof the crop on the May, 4 2024. At the first sampling salinity (115–120ppt), pH (8.5–9.0) and alkalinity (170–180mg CaCO3/L) were not different among the experimental ponds when being measured between 06:00 and 07:00. Water temperature was not measured due to technical issues of the digital thermometer. The crop was terminated on DOC111 or about a month earlier than planned because of a prolonged heatwave in the research area. The heatwave started on DOC85 and lasted for 4 weeks with air temperature between 36.4° and 38.8°C. As a result, water temperature in the experimental ponds reached 30°C in the early morning and up to 37°C in the afternoon. Salinity increased to 158±23ppt in HIGHER pond group and 154±24ppt in the LOWER pond group on DOC111 when the second sampling was conducted. Higher water temperature (29.8±0.5°C) and lower pH (7.6–7.9) were recorded similarly across all the experimental ponds and were not different between the two pond groups. Alkalinity was 264± 74mg CaCO3/L for the HIGHER group and 322±51mg CaCO3/L for the LOWER group. Mass mortality of Artemia was observed from DOC89, followed by a series of algal crashes and re-blooms. Cyst collection ceased from DOC94. Artemia biomass was harvested from DOC99 to DOC111.

For each sampling, the microbiome samples (water, sediment, and Artemia) were collected between 06:00 and 07:00 from eight ponds. For each pond, water and sediment were collected from nine different locations along the surrounding ditch and over the shallow platform in the middle. These were then pooled in a sterilized container and mixed thoroughly. Then either 500mL of water or circa 300g of sediment were collected into preprepared sterilized containers and labeled. At the same time, up to 300 Artemia individuals were collected along the pond’s edge that opposed the wind direction. They were then stocked into new 5L polyethylene bags with 2L of pond water. The bag was enriched with pure oxygen, goose neck tightened, and labeled. All samples were transported by refrigerated vehicle to International University-Vietnam National University in Ho Chi Minh City for further processing. The transportation took ~6h. Once arrived, water and sediment samples were kept at 4°C till being processed, whereas the Artemia were immediately pretreated and stored at −80°C until processing. Water parameters including pH, DO, temperature, and salinity were measured using a hand-held YSI multiparameter water checker for each Artemia production pond at the time of sampling. Alkalinity was estimated by Sera KH test kit.

2.3. Sample Preparation.

To collect materials for DNA extraction 250mL of each pond water sample was filtered through a 47mm diameter sterile nitrocellulose filter membrane with a pore size of 0.22µm (Whatman Article No. 28413910) using a vacuum filtration apparatus (Sartorius, Germany). For pond sediment samples, the collected sediment was first filtered through 30 and 10mm sieves for the removal of large debris. Then, 1.0g of sieved sediment was mixed with distilled water and filtered through a 47mm diameter Whatman filter mem brane. The membranes were then washed with 20mL of dis tilled water to remove salt residue. They were then stored individually at −80°C. From the collection bags Artemia were dewatered and transferred into 5mL cryo-tubes prefilled with 3.0mL of RNA later (AM7021, Invitrogen). Each tube held 30–50 individuals. The tubes were kept at 4°C overnight before being transferred to −80°CuntilDNAextractiontime.Artemia guts were aseptically collected from 30 individuals of each sample under a stereomicroscope using pairs of sterile tweezers. They were then pooled for DNA extraction.

2.4. DNA Extraction.

Microbe DNA extraction from water and sediment samples was performed using commercial kits and the instructions provided by the manufacturers. DNeasy PowerWater Kit (Qiagen, Germany) was used for water samples. DNeasy PowerSoil Kit (Qiagen, Germany) was utilized for sediment samples, while CTAB DNA extraction methods with some modifications [34] was used for Artemia gut samples. For water samples, half of the prepared filter membrane was used for the extraction. For sediment samples, a quarter of the pre pared filter membrane (equivalent to 250mg of sediment) was used. All extracted DNA was assessed using a NanoDrop spec trophotometer (NanoDrop One, Thermo Scientific). DNA yieldvariedfrom10ng/µL up to more than 200ng/µL. A260/ A280 ratio ranged from 1.8 to 2.0, indicating the purity of the extracted DNA. DNA integrity was assessed, visualized, and documented by gel electrophoresis (Figure 1).

Regarding microeukaryotes, extraction of microeukaryotic DNA was also conducted using the same kits with additional heating at 60°C [34] and homogenizing phases. The presence of microeukaryotic DNA in the samples was confirmed using PCR products primers for ITS1 and ITS3 genes visualized in the gel electrophoresis (Figure 2). All qualified samples were stored at −80°C for sequencing.

2.5. Sequencing and Bioinformatic Analysis Protocol.

DNA samples were submitted to NovogeneAIT Genomics in Singa pore for Illumina sequencing using NovaSeq 6000 platform. First, the DNA samples were processed through multiple steps including samples quality control, library construction, and sequencing. Quality control was assessed at every step of the process to ensure data reliability. Two target amplified regions were chosen for this current study as follows: the V4 region (using 5′-GTGCCAGCMGCCGCGGTAA-3′ forward primer and 5′-GGACTACHVGGGTWTCTAAT-3′ reverse primer) of the 16S rRNA gene and the ITS1-5F region (using 5′ GGAAGTAAAAGTCGTAACAAGG-3′ forward primer and 5′-GCTGCGTTCTTCATCGATGC-3′ reverse primer) of the ITS rRNA gene. Once the DNA sample passed the quality control, it was amplified by primers depending on the interested regions and purified through gel electrophoresis. The amplified fragments underwent size selection, end repair, A tailing, and ligation with an Illumina adapter. NovogeneAIT used Agilent 5400 (AATI) instruments and qPCR to assess the integrity and size of the DNA fragments and the library’s effective concentration respectively. The quantified libraries were then pooled and sequenced on NovaSeq platform.

Dataprocessing was carried out with nf-core/ampliseq pipe line version 2.9.0 (https://github.com/nf-core/ampliseq) [35], part of the nf-core suite of workflows [36]. This workflow is based on the reproducible software environments from Bioconda [37] and Biocontainers [38] projects provided with some additional modifications (Figure 3). After assessing data quality, VSEARCH [39] clustered amplicon sequence variants (ASVs) into centroids with pairwise identity of 0.97. The ASVs are categorized by Barrnap into the origin domain. The V4 region of 16S rDNA was used to identify and characterize bacterial diversity [40]. Accordingly, DADA2 and QIIME2 [41] performed the taxonomy classification using the database “GTDB—Genome Taxonomy Database- R08-RS214.1” [42] and“Greengenes2 16S—Version2022.10”[43]. ASVsequences, abundance, and DADA2 taxonomic classifications were then loaded into QIIME2 for processing. Regarding to microeukaryotic communities, the sequenced ITS region on ITS rDNA gene was utilized as a key to precisely identify and depict the diversity of microeukaryotic communities [44]. Taxonomy was classified using the databases “UNITE QIIME release for eukaryotes—Version 9.0” [45] and “UNITE USEARCH/ UTAX release for eukaryotes—Version 9.0” [46] by QIIME2 and SINTAX, respectively. Similarly, ASV sequences, abundance, and SINTAX taxonomy designations were input into QIIME2 for analysis. The final microbial community data was displayed on this platform as a bar plot for the taxa composition, evaluated for relationships based on similarities in a principle component analysis (PCoA) plot, and examined for alpha (within-sample) in a box plot.

The quality of 16S and ITS amplicons was assessed to ensure sufficient sequencing output (approximate 100K tags per sample). For the water and sediment samples, total 64 samples, including 32 samples of 16S and 32 samples of ITS samples, were qualified for library construction. Regarding the Artemia gut samples, our examination showed only 10/16 samples and 4/16 samples qualified for library construction of 16S and ITS samples, respectively. Thus, data from 78 samples were used as input for raw sequence processing and downstream analysis. After trimming and filtering, 16S sequencing yielded 6,212,306 clean reads from 7,354,010 raw reads, while ITS sequencing produced 5,448,679 clean reads from 7,078,116 raw reads.

2.6. Statistical Analysis.

Descriptive statistics were summarized by groups of pond and sampling events. Where comparison is needed, one-way or two-way ANOVA was deployed at α=0.05 using SigmaPlot 15.1 (Inpixon). Important assumptions about normal distribution and homogeneity were all checked for the collected data before analysis [47].

…To be continued…

By Tung Hoang, Nguyen Thai Binh, Vo Thi Minh Thu, Tran Thi Hong Tham, Tran Minh Long

Read more:

• Evaluating the effect of protists in diets for Pacific white shrimp
• Grow-out protein requirements of Nile tilapia fed fishmeal-free diets
• Effects of pre-slaughter stress on the fillet quality of Nile tilapia