Microsatellite–centromere mapping in bighead carp ( Aristichthys nobilis) using gynogenetic diploid families

Gene–centromere mapping using half‐tetrad analysis is a powerful tool for understanding chromosomal behaviour and determining the position of centromeres in relation to genes or markers in fish. In this study, eight gynogenetic diploid families induced by inhibiting of the second meiotic division were genotyped at 66 microsatellite loci for microsatellite–centromere (M‐C) mapping in bighead carp ( Aristichthys nobilis). The absence of paternal alleles verified the success of gynogenetic development in all gynogenetic families. All loci were consistent with Mendelian segregation in control families. The estimated recombination frequency (y) ranged from 0.057 to 0.875 with an average of 0.477 ± 0.222. Seventeen loci (25.76%) showed high M‐C recombination frequencies of over 0.667. M‐C distances ranged from 2.85 to 43.75 cM under the assumption of complete interference. Thus, these loci are distributed from the centromeres to the telomeres of their respective chromosomes, while mainly in the intermediate region. Information on centromere mapping could serve as a starting point to consolidate the genetic linkage groups in bighead carp.     

Microsatellite–centromere mapping in walking catfish Clarias macrocephalus (Günther, 1864) using gynogenetic diploids

Thirty new microsatellite loci were isolated from a microsatellite‐enriched genomic library of walking catfish Clarias macrocephalus. The CA motif was the most abundant, although several other motifs were also isolated. Two gynogenetic diploid families, each consisting of a female and 50 offspring, were genotyped at 56 microsatellite loci, including 30 loci developed in the present study and 26 loci reported in the previous study. Overall, 33 anonymous microsatellites derived from genomic library and 11 EST‐linked microsatellites were mapped to their centromeres. High levels of chiasma interference were apparent in the walking catfish genome as indicated by an average frequency of second‐division segregation (y) of 0.643 ± 0.248. Twenty‐six loci (59%) showed a high microsatellite–centromere recombination, with a frequency >0.67, and three loci had recombination frequencies >0.9. This study demonstrated that gene–centromere mapping provided a rapid method for the expansion of the initial linkage map of walking catfish.     

Test of Mendelian segregation among 10 microsatellite loci in the fourth generation of bester (Huso huso L. ×Acipenser ruthenus L.)

Ten microsatellite loci were tested on Mendelian segregation in the bester – hybrid of beluga, Huso huso L. and sterlet, Acipenser ruthenus L. in the fourth generation. All studied loci showed disomic inheritance and Mendelian segregation of their alleles. The molecular approach by microsatellite DNA analysis represents a reliable tool for interspecific characterization of sturgeons and their hybrids (different alleles of several loci are present in sturgeon fishes, and interspecific hybrids posesses both maternal and paternal specific alleles). This is the first report of the Mendelian segregation testing in interspecific hybrid of acipenserid fishes using microsatellite DNA markers.     

Genetic linkage map of bay scallop, Argopecten irradians irradians (Lamarck 1819)

The bay scallop (Argopecten irradians irradians Lamarck 1819) has become one of the most important aquaculture species in China. Genetic improvement of cultured bay scallop can benefit greatly from a better understanding of its genome. In this study, we developed amplified fragment length polymorphisms (AFLPs) and simple sequence repeat markers from expressed sequence tags (EST‐SSRs) for linkage analysis in bay scallop. Segregation of 390 AFLP and eight SSR markers was analysed in a mapping population of 97 progeny. Of the AFLP markers analysed, 326 segregated in the expected 1:1 Mendelian ratio, while the remaining 74 (or 19.0%) showed significant deviation, with 33 (44.6%) being deficient in heterozygotes (A/a). Among the eight polymorphic EST‐SSR loci, one marker (12.5%) was found skewing from its expected Mendelian ratios. Eighteen per cent of the markers segregating from female parent were distorted compared with 21% of the markers segregating from male parent. The female map included 147 markers in 17 linkage groups (LGs) and covered 1892.4 cM of the genome. In the male map, totally 146 AFLP and SSR markers were grouped in 18 LGs spanning 1937.1 cM. The average inter‐marker spacing in female and male map was 12.9 and 13.3 cM respectively. The AFLP and SSR markers were distributed evenly throughout the genome except for a few large gaps over 20 cM. Although preliminary, the genetic maps presented here provide a starting point for the mapping of the bay scallop genome.     

镜鲤多家系体长和体质量QTL定位分析

为了克服单个家系数量性状位点(QTL)检测效率低、假阳性高等缺点,实验利用250对微卫星(SSR)标记对镜鲤8个全同胞家系的522尾子代进行基因组扫描,采用半同胞家系的分析策略对镜鲤体长(SL)和体质量(BW)性状进行QTL分析。结果显示,基于父系的QTL分析,共检测到4个QTL区间,其中,3个体长的QTL中,1个为95%基因组水平(genome-wide)显著性,位于LG24,可解释表型变异率为20.3%;其余2个均为95%染色体水平(chromosome-wide)显著性,分别位于LG6和LG30,可解释表型变异率分别为11.9%和11.6%。1个体质量的QTL达到99%基因组水平,位于LG24,可解释表型变异率达到38.3%,且与体长QTL区间重叠。基于母系的QTL分析,共检测到8个QTL区间,其中,5个体长的QTL中,1个为99%染色体水平,位于LG8,可解释表型变异率为16.6%;其余4个均为95%染色体水平,分别位于LG24、LG30、LG31和LG45,可解释表型变异率为9.6%~14.2%,且位于LG24和LG30上的QTL为父母本共有;3个体质量的QTL均与体长QTL区间重叠,1个为95%染色体水平,位于LG24,其余2个均为99%染色体水平,位于LG30和LG45,可解释表型变异率分别为14.1%和13.6%。进一步分析发现,位于LG24上的体长和体质量QTL区间重叠且均为父母本共有,体质量的3个QTL均与体长QTL存在重叠区域且呈现成簇分布的特点。本研究结果不仅可以为鲤分子育种提供更可靠的标记,而且为家系和品种间QTL变异规律的探索提供基础数据。     

A preliminary genetic map of Zhikong scallop (Chlamys farreri Jones et Preston 1904)

Zhikong scallop (Chlamys farreri Jones et Preston 1904) is one of the most important aquaculture species in China. The development of a genetic linkage map would provide a powerful tool for the genetic improvement of this species. Amplified fragment length polymorphism (AFLP) is a PCR‐based technique that has proven to be powerful in genome fingerprinting and mapping, and population analysis. Genetic maps of C. farreri were constructed using AFLP markers and a full‐sib family with 60 progeny. A total of 503 segregating AFLP markers were obtained, with 472 following the Mendelian segregation ratio of 1:1 and 31 markers showing significant (P<0.05) segregation distortion. The male map contained 166 informative AFLP markers in 23 linkage groups covering 2468 cM. The average distance between markers was 14.9 cM. The female genetic map consisted of 198 markers in 25 linkage groups spanning 3130 cM with an average inter‐marker spacing of 15.8 cM. DNA polymorphisms that segregated in a 3:1 ratio as well as the AFLP markers that were heterozygous in both parents were included to construct combined linkage genetic map. Five shared linkage groups, ranging from 61.1 to 162.5 cM, were identified between the male and female maps, covering 431 cM. Amplified fragment length polymorphism markers appeared to be evenly distributed within the linkage groups. Although preliminary, these maps provide a starting point for the mapping of the functional genes and quantitative trait loci in C. farreri.     

Isolation of microsatellite markers by in silico screening implicated for genetic linkage mapping in Japanese pufferfish Takifugu rubripes

ABSTRACT:   The search for dinucleotide repeat microsatellites within scaffolds 1–25 of genome database JGI Fugu v3.0 for the pufferfish Takifugu rubripes revealed that 80% of microsatellite loci consisted of five to 13-fold repeats with locus-specific differences in density. Eleven out of 15 microsatellite loci isolated from the database with which genotyping using wild pufferfish was successfully performed showed polymorphism; that is, the means of the number of alleles and expected and observed heterozygosities at these 11 loci were 21.8, 0.915 and 0.829, respectively. It was confirmed that eight out of the 11 polymorphic loci were inherited through the Mendelian law and one pair of microsatellite loci derived from the same scaffold was linked. These results demonstrated that these loci are useful for constructing a linkage map in the pufferfish as DNA markers.     

Quantitative trait loci (QTL) associated with growth rate trait in common carp (Cyprinus carpio)

Growth rate is one of the important economic traits in aquaculture species that is influenced by quantitative trait loci (QTLs). Recently published complete genetic map of common carp (Cyprinus carpio) with 307 markers (including 109 SSR, 31 EST-SSR and 167 SNP markers) was used to determine the number and position of QTLs effecting body weight (BW) and growth rate. Ten-, 11- and 12-month-old individuals of common carp were measured, and seven QTLs were detected in three linkage groups (LG1, LG6 and LG20); three were significant QTLs (qBW2-6, qBW3-6 and qBW3-20, P < 0.001), and four were suggestive QTLs (qBW1-1, qBW1-6, qBW2-20 and qBW3-1, P < 0.01). The three significant QTLs related to BW2 and BW3 were in LG6 and LG20 and covered 18 and 41 cM, respectively. The explained variances of qBW2-6, qBW3-6 and qBW3-20 were 34.4, 32.1 and 33.4 %, respectively. The four suggestive QTLs affecting BW1, BW2 and BW3 were in LG1, LG6 and LG20 and had variances that ranged from 17.9 to 26.7 %. The similarity of the QTL intervals for BW1, BW2 and BW3 in LG6 suggested that these genes may affect growth continuously during the early stage of growth and may be important for growth. The identification of QTLs for growth rate and BW in C. carpio will lead to a better understanding of the factors that influence these traits at a genetic level and will promote the growth to higher production through marker-assisted breeding.     

Application of amplified fragment length polymorphism markers to assess molecular polymorphisms in gynogenetic haploid embryos of turbot (Scophthalmus maximus)

Among the variety of cultured marine species, the turbot Scophthalmus maximus is a fish of growing importance in European aquaculture. In this paper, an advanced application of AFLPs to estimate the genetic diversity of haploid gynogenetic families with the aim of obtaining a preliminary genetic map is presented. Ten EcoRI/TaqI primer combinations were tested in four families comprising diploid mothers and their haploid progenies. The amplified fragment length polymorphism (AFLP) analysis revealed an average of 6.8 polymorphic bands per primer combination and a total number of 88 polymorphisms out of 579 fragments. Among various primer pairs, seven combinations were selected in relation to the quality of profiles and number of polymorphic fragments, to be used in the determination of genetic linkage relationship between AFLP markers within the largest haploid family. Co‐migration of non‐homologous fragments was also investigated in one primer combination adding a fourth selective nucleotide to the three used in the classic TaqI AFLP protocol. Surprisingly, a rate of 38.7% of non‐homologous fragments co‐migrating with monomorphic bands was identified, due to the combined effect of homoplasy and the protocol used. Additional polymorphic markers discovered by this protocol were included in the linkage map. The turbot AFLP linkage map comprises 52 AFLP markers distributed in 12 linkage groups. On the basis of this map, turbot expected total genome length sums up to 1225.6 cM. The results confirm the usefulness of AFLPs in revealing genome segregation in haploid turbot progeny.     

Quantitative trait locus mapping of growth‐related traits in inter‐specific F1 hybrid grouper (Epinephelus fuscoguttatus × E. lanceolatus) in a tropical climate

Growth‐related traits are the main target of genetic breeding programmes in grouper aquaculture. We constructed genetic linkage maps for tiger grouper (Epinephelus fuscoguttatus) and giant grouper (E. lanceolatus) using 399 simple sequence repeat markers and performed a quantitative trait locus (QTL) analysis to identify the genomic regions responsible for growth‐related traits in F₁ hybrid grouper (E. fuscoguttatus × E. lanceolatus). The tiger grouper (female) linkage map contained 330 markers assigned to 24 linkage groups (LGs) and spanned 1,202.0 cM. The giant grouper (male) linkage map contained 231 markers distributed in 24 LGs and spanned 953.7 cM. Six QTLs affecting growth‐related traits with 5% genome‐wide significance were detected on different LGs. Four QTLs were identified for total length and body weight on Efu_LG8, 10, 13 and 19 on the tiger grouper map, which explained 6.6%–12.0% of the phenotypic variance. An epistatic QTL with a reciprocal association was observed between Efu_LG8 and 10. Two QTLs were identified for body weight on Ela_LG3 and 10 on the giant grouper map, which explained 6.9% of the phenotypic variance. Two‐way analysis of variance indicated that the QTL on Efu_LG13 interacts with the QTLs on Ela_LG3 and 10 with large effects on body weight. Furthermore, these six QTLs showed different features among the winter, summer and rainy seasons, suggesting that environmental factors and fish age affected these QTLs. These findings will be useful to understand the genetic structure of growth and conduct genetic breeding in grouper species.     

Inheritance mode of microsatellite DNA markers and their use for kinship estimation in kuruma prawn Penaeus japonicus

The allelic inheritance mode of microsatellite DNA markers was examined using seven copulated wild females and their offspring. Five microsatellite loci, CSPJ002 *, CSPJ010 *, CSPJ012 *, CSPJ014 *, and CSPJ015 *, were used in the study. At almost all family/locus combinations, one sire was determined and distributions of genotypes in offspring were consistent with the Mendelian segregation ratio. Distributions of genotypes were consistent with the ratio after assuming a null allele at some loci. Consequently, the alleles of CSPJ002 * and CSPJ012 * were inherited following the Mendelian inheritance mode in every family; however, the null allele was expected in CSPJ010 *, CSPJ014 *, and CSPJ015 * in some families. Thus, these loci should be used carefully in population genetic analysis, but siblings could be detected in the dendrograms based on unweighted pair-group method using arithmetic averages (UPMGA).     

A microsatellite marker tool for parentage analysis in Senegal sole (Solea senegalensis): Genotyping errors, null alleles and conformance to theoretical assumptions

A set of 50 potential microsatellite loci previously described in several flatfish species was evaluated for designing a suitable tool for parentage assessment in Senegal sole (Solea senegalensis). Low cross-amplification within Pleuronectiformes, even between congeneric species, was achieved. This suggests a low conservation of microsatellite flanking regions in this order regarding other fish orders of similar evolutionary age. Twelve loci showed appropriate amplification and polymorphism to acquire enough statistical power for parentage analysis. All forty-one individuals from one broodstock and 320 offspring were analyzed using this set of markers. The high potential of microsatellites for parentage exclusion (Excl1 = 0.9999; Excl2 = 1) permitted us to trace back all offspring to single pairs of parents. The families obtained were used to analyze the performances of the loci applied and to identify possible sources of error for parentage analysis in Senegal sole. The 2.5% of single mismatches in the families studied were attributed mostly to genotyping errors and to a lesser extent to null alleles. The average mutation rate estimated over 7680 gametes was in the range reported within fish (3.9 × 10^− 4) and did not compromise parentage analysis. All loci conformed to theoretical assumptions (Mendelian segregation, independent segregation), excluding the F13-II8/4/7 and Sol 9A loci, which showed strong genotypic association in the families analyzed. A set of 4 microsatellites (Sol CA13, F13-II8/4/7, Smax-02 and Sse GATA38) was finally selected for parentage evaluation in S. senegalensis combining both enough exclusion potential (> 0.970) and the lowest economic cost.     

Growth and age at first maturity in turbot and halibut reared under different photoperiods

The effect of extended photoperiods on growth and age at first maturity was investigated in 166 (79 females and 87 males) individually tagged Atlantic halibut Hippoglossus hippoglossus and in 114 (50 females and 64 males) individually tagged turbot Scophthalmus maximus. The halibut were reared at 11 °C on four different light regimes from 10 February to 6 July 1996: simulated natural photoperiod, (LDN), continuous light (LD24:0), constant 8 h light and 16 h darkness (LD8:16) and LD8:16 switched to continuous light 4 May 1996 (LD8:16–24:0). From 6 July 1996 to 9 February 1998 the LD24:0 and LD8:16–24:0 were reared together under continuous light and the LDN and LD8:16 together under natural photoperiod. The turbot were reared at 16 °C on three different light regimes: constant light (LD24:0), 16 h light:8 h darkness (LD16:8), or simulated natural photoperiod (LDN). After 6 months on the different photoperiods, the turbot was reared together on LDN for approximately 12 months until first maturation. Juveniles subjected to continuous light (halibut) or extended photoperiods (halibut and turbot) exhibited faster growth than those experiencing a natural photoperiod or a constant short day. Moreover, when the photoperiod increased naturally with day-length or when fish were abruptly switched from being reared on short-day conditions to continuous light, a subsequent increase in growth rate was observed. This growth enhancing effect of extended photoperiods was more apparent on a short time scale in Atlantic halibut than in turbot, but both species show significant long-term effects of extended photoperiods in the form of enhanced growth. In both species lower maturation of males was seen in groups exposed to extended or continuous light compared to LDN and this could be used to reduce precocious maturation in males leading to overall increase in somatic growth. This revised version was published online in August 2006 with corrections to the Cover Date.     

中华绒螯蟹37个多态性微卫星标记在亲子遗传中的分离方式分析

为了评估基因组扫描方法获得的中华绒螯蟹微卫星标记的遗传方式,随机选择60个候选三核苷酸微卫星标记,首先利用中华绒螯蟹F1家系双亲及其6个F1子代共8个样品进行PCR扩增验证和多态性检测,随后对多态性标记位点在80子代个体中的亲子遗传分离类型及连锁关系进行了分析。结果显示,42个（70.00%）位点得到清晰扩增产物,其中有5个单态微卫星位点和37个多态微卫星位点。多态位点中23个位点子代基因基因型为1：1：1：1分离类型,11个位点属1：1分离类型,余下3个位点为1：2：1分离类型。37个多态位点中35个位点（94.59%）的分离符合孟德尔分离比（P>0.05）,scaffold430598_213690和scaffold21303_16865位点显著偏离1：1：1：1分离比。35个分离比符合孟德尔分离比的位点中,scaffold240262_150253、scaffold216209_138892、scaffold293154_172768三个标记发生连锁关系,scaffold285640_169721和scaffold427534_212914发生连锁关系,scaffold507500_231891和scaffold92860_68250标记发生连锁关系。以上结果表明开发的候选微卫星标记适用于中华绒螯蟹亲子鉴定和遗传图谱构建。     

A first genetic linkage map of bluegill sunfish (<Emphasis Type="Italic">Lepomis macrochirus</Emphasis>) using AFLP markers

Genetic linkage maps were constructed for bluegill sunfish, Lepomis macrochirus, using AFLP in a F₁ inter-population hybrid family based on a double-pseudo testcross strategy. Sixty-four primer combinations produced 4,010 loci, of which 222 maternal loci and 216 paternal loci segregated at a 1:1 Mendelian ratio, respectively. The female and male framework maps consisted of 176 and 177 markers ordered into 31 and 33 genetic linkage groups, spanning 1628.2 and 1525.3 cM, with an average marker spacing of 10.71 and 10.59 cM, respectively. Genome coverage was estimated to be 69.5 and 69.3% for the female and male framework maps, respectively. On the maternal genetic linkage map, the maximum length and marker number of the linkage groups were 122.9 cM and 14, respectively. For the paternal map, the maximum length and marker number of the linkage groups were 345.3 cM and 19, respectively, which were much greater than those on the maternal genetic linkage map. The other genetic linkage map parameters of the paternal genetic linkage map were similar to those in the maternal genetic linkage map. For both the female and male maps, the number of linkage groups was greater than the haploid chromosome number of bluegill (2n = 48), indicating some linkage groups may distribute on the same chromosome. This genetic linkage mapping is the first step toward to the QTL mapping of traits important to cultured breeding in bluegill.     

A set of highly polymorphic microsatellites useful for kinship and population analysis in turbot (Scophthalmus maximus L.)

The turbot (Scophthalmus maximus) is a flatfish species of great commercial value for aquaculture. In this study we describe the isolation and characterization of 30 novel highly polymorphic microsatellite markers in this species obtained from a genomic library enriched for seven short tandem repeated motifs. Much higher polymorphism (mean number of alleles: 13.37; mean expected heterozygosity: 0.869) and potential for parentage assignment than previously reported for microsatellites in turbot were found after the analysis of 24 wild individuals. Most loci conformed to Hardy–Weinberg expectations, excluding Sma‐USC20 and Sma‐USC28, which showed a high heterozygote deficit probably due to the presence of null alleles. No significant genotypic disequilibrium was observed between any pair of loci, suggesting no close linkage between them. These loci are potentially useful for kinship and population analysis in turbot.     

三角帆蚌HcTyr基因内壳色性状相关SNP筛选及图谱定位

为研究三角帆蚌HcTyr基因与内壳色性状的相关性,本实验根据已分离的三角帆蚌HcTyr基因进行引物设计和片段扩增,并用144只三角帆蚌对其多态性进行筛选和验证,卡方检验分析了SNP位点和三角帆蚌内壳色相关性,并对HcTyr基因进行了图谱定位。结果显示,在HcTyr基因上筛选出8个SNP位点,其中有7个SNP位点与三角帆蚌内壳色L、a及d E呈显著相关性,用这7个SNP位点做单倍型构建和分析,发现Ⅱ、Ⅲ及Ⅳ这3种单倍型在白色群体中出现的频率极显著高于在紫色群体中出现的频率,而Ⅴ和Ⅶ这2两种单倍型在紫色群体中出现的频率极显著高于白色群体。进一步在商业养殖群体中验证发现,Ⅳ和Ⅶ这2种单倍型可分别作为白色和紫色群体的优势单倍型。研究表明,三角帆蚌HcTyr基因可作为内壳色相关的候选基因,其部分SNP位点可用于三角帆蚌分子辅助育种。另外本实验还将HcTyr基因定位在三角帆蚌LG16连锁群上,这为进一步解析该基因调控珍珠颜色形成的分子机制奠定了基础。     

Genotyping‐by‐sequencing for construction of a new genetic linkage map and QTL analysis of growth‐related traits in Pacific bluefin tuna

Pacific bluefin tuna (Thunnus orientalis) has high market value, but its wild populations have decreased in recent years. The broodstock of Pacific bluefin tuna that were hatched artificially and reared under aquaculture conditions is beginning to be used for production. The creation of broodstock with commercially valuable traits, such as rapid growth, is therefore of great interest. Genetic linkage map‐based identification of markers associated with quantitative trait loci (QTLs) facilitates marker‐assisted selection (MAS) breeding and allows efficient genetic improvement of broodstock. Single nucleotide polymorphism (SNP)‐based genetic linkage map construction using the genotyping‐by‐sequencing method can expand the number of mapped markers and help identify growth‐related QTLs. In this study, we constructed sex‐specific maps for 24 linkage groups consisting of 677 SNP and 651 microsatellite markers. The total lengths of 93 progenies in the mapping population followed normal distribution, with an average length of 9.4 mm. We performed composite interval mapping in the mapping population. QTL analysis revealed one significant QTL in LG10 on the female linkage map. The genetic linkage map—the second such map generated for Pacific bluefin tuna—and the growth‐related QTLs detected in this study will be useful for tuna aquaculture MAS programs.     

Low optimal temperatures for food conversion and growth in the big-headed turtle, Platysternon megacephalum

We held juvenile big-headed turtles, Platysternon megacephalum, from eastern China, at temperatures from 20 to 29.4 °C to determine effects on feeding, growth and food conversion. Food intake increased significantly from 20 to 22.4 °C, remained high until 27.1 °C, and then decreased dramatically at 29.4 °C. Digestive efficiency for energy decreased as temperature increased, whereas the digestive efficiency of protein increased from 20 to 25 °C, and decreased at higher temperatures. The relationships between specific growth rate (SGR), food conversion coefficient (C_c) and temperature (T) were curvilinear, and could be described by quadratic equations: SGR = −0.01 T² + 0.47 T − 5.24 and C_c = −0.37 T² + 17.20 T − 181.85. Maximum growth was estimated to occur at 23.9 °C, with 90% of the maximum being achieved within the range of 21.9–25.8 °C; maximal food conversion occurred at 23.2 °C, with a 90% range from 21.0 to 25.4 °C. The temperature range (22–25 °C) found to promote best growth and food conversion in juvenile P. megacephalum is lower than for many other freshwater turtles. Temperatures of 22–25 °C are recommended for use in culture of this species to maximize growth and food conversion.     

Effects of manganese on juvenile mulloway (Argyrosomus japonicus) cultured in water with varying salinity—Implications for inland mariculture

The effects of 5 mg/L of dissolved manganese on juvenile mulloway at salinities of 5, 15 and 45 ppt were determined by comparing their survival, growth and blood plasma and organ chemistry with those of fish grown at the same salinities without manganese addition. Survival of mulloway at 45 ppt in the presence of 5 mg/L of manganese (73 ± 13%) was significantly lower than all other treatments, which achieved 100% survival. Those fish grown in water without manganese exhibited rapid growth, which was not affected by salinity (SGR = 4.05 ± 0.29%/day). Those fish grown at 5 ppt and 45 ppt in the presence of manganese lost weight over the 2-week trial (SGR − 0.17 ± 0.42 and − 0.44 ± 0.83%/day, respectively), whilst those at 15 ppt gained some weight (SGR 1.70 ± 0.20%/day). Manganese accumulated in the gills, liver and muscle of the fish and significant differences in blood plasma chemistry were observed. Blood plasma sodium and chloride of fish exposed to manganese were significantly elevated in hyperosmotic salinity (45 ppt) and depressed at hyposmotic salinity (5 ppt) compared with unexposed fish at the same salinity; consistent with manganese causing apoptosis or necrosis to chloride cells. We did not, however, observe any gill epithelial damage under light microscopy. Blood plasma potassium was significantly elevated at all salinities in the presence of manganese and liver potassium and glycogen reduced. These findings are consistent with manganese interfering with carbohydrate metabolism.