 |
INTRODUCTION |
Reproductive impairment has been identified in salmonid populations inhabiting the Finger Lakes of New York State, the lower Great Lakes of Canada and the United States, and the Baltic Sea. In these regions, reproductive problems manifest as variably lethal syndromes in sac-fry larvae or first-feeding fry. In some of the Finger Lakes, the Cayuga syndrome afflicts sac-fry of landlocked Atlantic salmon (Salmo salar) (Fisher et al. 1995b
). In the lower Great Lakes, swim-up syndromes have been recorded in first-feeding lake trout (Salvelinus namaycush) (Fitzsimons et al. 1995, Mac et al. 1985), brown trout (Salmo trutta) (Marcquenski and Brown 1997), rainbow trout (Oncorhynchus mykiss) (Skea et al. 1985), and coho and chinook salmon (O. kisutch and O. tshawytscha) (Hornung et al. 1998). In Sweden and Finland, the M-74 syndrome kills sac-fry of anadromous Baltic salmon (Salmo salar) (Norrgren et al. 1993).
The severity of these early mortality syndromes (EMS)6 appears to be species, population (i.e., lake) and maternally specific. The Cayuga syndrome has been observed in progeny from every Cayuga Lake (CL) landlocked salmon female examined since 1974 (n = 3-10 females/y), with mortality usually reaching 100%; yet, lake trout, rainbow trout, and brown trout appear unaffected (Fisher et al. 1995b, 1998). Similarly, mortality of progeny from M-74-positive salmon is also generally 100%; however, the incidence among females varies substantially between years, and Baltic brown trout may also be affected (Norrgren et al. 1998). In contrast, not all Great Lakes salmonids afflicted with an EMS die from the disease. For example, maternally specific mortality of lake trout offspring with the swim-up syndrome ranged from 31-75%, yet signs of the condition were observed in offspring from 87.5% of the lake trout spawned (Fitzsimons et al. 1995). The incidence of EMS in specific Great Lakes populations also fluctuates between years, and certain populations appear more susceptible (Marcquenski and Brown 1997).
Results from studies in the Finger Lakes did not support a genetic basis for the Cayuga syndrome and implicated diet as a likely factor. Only salmon populations whose forage was dominated by the nonnative, thiaminase-rich alewife (Alosa pseudoharengus) exhibited syndrome-related mortality despite their common progenitor stock (Fisher and Spitsbergen 1990, Fisher et al. 1995b). Evidence from pathology (Fisher et al. 1995a), and from experiments where thiamin treatments effectively resolved syndrome-related mortality in lake trout (Fitzsimons 1995) and landlocked Atlantic salmon (Fisher et al. 1996b), further supported the role of thiamin deficiencies in the etiologies of these North American syndromes. Thiamin treatments initially developed for landlocked salmon sac-fry (Fisher et al. 1996b) also reduced M-74 mortality in sac-fry from anadromous Atlantic salmon (Bylund and Lerche 1995). These anadromous salmon feed heavily on baltic herring (Clupea harengus) and sprat (Sprattus sprattus L.) (Ikonen 1996), species that also exhibit high thiaminase activity (Anglesea and Jackson 1985, Nielands 1947). Given the pathological similarities between the Cayuga, M-74, and swim-up syndromes and their apparent dietary connections, it was proposed that forage high in thiaminase could result in thiamin deficient eggs through a mechanism that reduced the initial absorption of the vitamin from the maternal diet (Fisher et al. 1996).
View this table:
[in this window]
[in a new window]
|
Table 1.
Source, syndrome status, region, diet, and number (by sex) of landlocked Atlantic salmon (Salmo salar) sampled
for these studies
|
|
In this study, we explore the relationship between thiamin status of wild landlocked Atlantic salmon and the viability of their sac-fry progeny, and the connection between sac-fry mortality and a maternal diet of thiaminase-rich alewife. The initial research that identified thiamin deficiencies in Finger Lakes salmon quantified total thiamin levels only in sac-fry (Fisher et al. 1996b). In the present study, levels of free thiamin, thiamin pyrophosphate (TPP), thiamin monophosphate (TMP), and the total of these thiamin moieties (hereafter total thiamin) in blood, eggs, and larvae of wild landlocked Atlantic salmon from several New York watersheds were determined, and survival of their progeny was monitored. The objectives of this study were four-fold: 1) to determine whether salmon populations outside of the Finger Lakes catchment were also susceptible to the Cayuga syndrome if they consumed alewife and, if so; 2) to determine if their syndrome-related mortality was associated with depressed thiamin levels that could be resolved with treatment; 3) to compare the proportionate distribution of the thiamin moieties within the parental blood, eggs, and sac-fry of control, afflicted and thiamin-supplemented stocks and 4) to identify tissue threshold levels of thiamin in maternal blood, eggs or larvae that might prevent syndrome-related mortality. Results of these studies suggest that a natural diet containing high levels of thiaminase is most likely responsible for depressed thiamin levels in adult landlocked Atlantic salmon and their resultant eggs, without partiality for the watershed in which the salmon reside. Furthermore, the data suggest that threshold levels of egg thiamin must be met by the completion of oogenesis to prevent significant larval mortality prior to the initiation of exogenous feeding.
| |
MATERIALS AND METHODS |
Source of salmon.
To evaluate the role of thiaminase-rich alewife forage on reproductive viability and thiamin levels in landlocked Atlantic salmon, eggs, blood or both were sampled from two populations that lacked an alewife diet and four sources that feed heavily or exclusively on alewife. All salmon sampled were captured by using trap-nets or by nonlethal electroshocking; in previous studies with syndrome-positive CL salmon (Fisher et al. 1995b), capture by either method did not alter the incidence of Cayuga syndrome. Data on the watershed source, diet, and known syndrome status of salmon stocks evaluated for these studies are detailed in Table 1. Additional stock and watershed information relevant to salmon collected for these studies is provided in Fisher et al. (1995b, 1998) and Oglesby (1978).
Assessment of landlocked Atlantic salmon viability.
Survival of salmon progeny from each female was monitored from fertilization through the accumulation of 1400 centigrade dd, roughly 7.5 mo later. Degree-day units refer to the average daily incubation temperature multiplied by the number of days since fertilization and are commonly used to gauge embryonic and larval development of fish (Leitritz and Lewis 1980). If eggs could be manually expressed, females were spawned immediately after capture in the field. Salmon from CL were transported to aquaculture facilities at Cornell University, where they were held for 2 wk in a flow-through, 2.6-m-diameter tank until spawning on November 21. Salmon from CL and Green Pond (GP) were killed immediately before spawning by a blow to the head, and the entire egg mass was taken from each female. Salmon from Little Clear Pond (LC), Little Clear Hatchery (LCH) and Otsego Lake (OL) were briefly anesthetized in tricaine methane sulfonate (MS-222, Sigma Chemicals, St. Louis, MO) before spawning and having blood samples taken, and released alive immediately after recovery. Due to New York State restrictions, only 300-500 eggs were taken from each of the LC, LCH, and OL females.
Eggs from each female were fertilized dry (without added water) with the sperm of two males from the same lake source. Sperm were activated with dechlorinated and filtered CL water from the aquaculture facilities at Cornell University; conventional chemical parameters of this water were previously described (Fisher et al. 1995b). For collections at LC, LCH, OL and GP, the CL water was transported to the respective field sites. After 5 min, the sperm were rinsed away, and the eggs were water hardened for another hour before transfer into incubation facilities. (Water-hardening refers to the uptake of water through pores in the chorion; the 30-60 min process is initiated upon exposure of the eggs to water and increases egg volume by roughly 30%).
To assess the incidence of the Cayuga syndrome in landlocked salmon from outside the Finger Lakes region, 15 mL of water-hardened eggs from each female were placed in separate screened 7.6-cm-diameter polyvinyl chloride cups in 20-L glass aquariums. Because of size differences in eggs from individual females, the density of eggs ranged from 103 to 189 eggs per cup (mean = 130 ± 22 SD) on the initial distribution date. Four to six cups were placed in each aquarium, and the lake sources of eggs were separated. The eggs were suspended 6-8 cm below the water surface, and each aquarium received upwelling flow-through (1.5-2 L/min) dechlorinated and filtered CL water at ambient temperature. Water temperatures averaged 10.5 ± 1.1°C in November, 7.6 ± 0.7°C in December, 6.3 ± 0.4°C in January, 4.9 ± 0.4°C in February, 5.1 ± 0.6°C in March, 6.4 ± 0.5 °C in April and 9.2 ± 0.8°C from May 1 to May 28, 1995, when the study was concluded. The surplus eggs remaining after this initial distribution were either 1) pooled by lake source (i.e., all surplus LC, LCH, and CL eggs), or 2) kept separated by female in a second series of incubation cups (i.e., eggs from GP and OL females); these surplus eggs were used for subsequent thiamin treatment trials.
Separate records were maintained for embryo, larval (i.e., alevin or sac-fry) and fry mortality7. The temperature units of development accumulated at the time of mortality were recorded as dd. To reduce densities for growth, the number of sac-fry in each cup was normalized to a mean of 59 ± 7/cup on February 27, 1995. By this date, the number of centigrade dd accumulated for each stock was 650 (CL), 704 (OL) and 785 (LC, LCH, and GP).
Thiamin treatment of syndrome-positive stocks.
When early behavioral and pathological signs of the Cayuga syndrome (see Fisher et al. 1995b) were apparent in sac-fry progeny from OL and GP salmon, we proceeded to evaluate the efficacy of thiamin therapy at reducing or eliminating potential mortality that might result if these stocks were left untreated.
Thiamin bath treatments were performed using a protocol previously established (Fisher et al. 1996b), except that no carrier solvent was added to the treatment solutions, and the pH was raised to 6.5 instead of 5.5. Briefly, sac-fry were immersed for 1 h in aerated 10 g/L (active) thiamin hydrochloride (Sigma Chemicals, St. Louis, MO). Thus, these thiamin treatments were applied to provide a measure of confirmation or refutation for a possible thiamin deficiency in the OL and GP stocks. The treatments were not designed to evaluate, from a clinical-trial approach, the dose-response efficacy of thiamin treatments; those experiments were also previously conducted (Fisher et al. 1996a).
Surplus sac-fry from individual females were used for the treatment trials. The number of sac-fry treated per female ranged from 14 to 109, depending on remaining availability. Surplus sac-fry from the pooled syndrome-positive CL stock were also treated with thiamin, providing a positive-control to the trials. Previous treatments of syndrome-negative LC stock had shown that the thiamin treatments were not toxic to healthy sac-fry (Fisher et al. 1996b); hence, there was no syndrome-negative group incorporated into these trials. Untreated sac-fry from each female and lake stock were also already under observation, as described in the previous section. At the time of treatment, the sac-fry were approximately 5 (CL), 10 (OL) or 11 wk (GP) post hatching, and had accumulated 622 (CL), 762 (OL), or 839 centigrade dd (GP) since fertilization.
Assessment of egg, blood, and sac-fry thiamin status by stock.
Concentrations of free thiamin and its mono and diphosphate esters (i.e., thiamin monophosphate and thiamin pyrophosphate) were determined in the whole blood of all salmon captured. Thiamin levels were also determined in the eggs and sac-fry progeny of each female salmon. Despite its recognized role in nerve conduction (Gubler 1991), thiamin triphosphate (TTP) was not measured in these studies because it is not appreciably stored in adult animals (<3% of total in rat liver, 0.6% in brain; Kawosaki and Sanemora 1991). Furthermore, although identifiable by the method employed here, levels in our samples were below consistent quantification and no standard was available. Thus, total thiamin, as considered in these studies, was determined by combining the concentrations of TMP, TPP, and free thiamin. Ratios of each thiamin moiety relative to the total thiamin were compared among these tissues to assess potential differences among salmon populations in their metabolism of the vitamer.
Prior to blood sampling, we anesthetized the fish in 125 mg/L tricaine methane sulfonate (i.e., MS-222, Sigma Chemicals, St. Louis, MO), or killed with a blow to the head per euthanasia techniques approved by the New York Department of Animal Welfare. Two blood samples were drawn from the caudal vein of each fish using 21-gauge needles and 5 mL VacutainerTM collection tubes (Becton Dickinson and Co., Rutherford, NJ) without anti-coagulant. (Thiamin levels in blood from salmonids sampled both with and without lithium-heparin anticoagulant were not significantly different). All blood samples were briefly shaken by hand and immediately frozen in crushed dry ice and transferred to a -70°C freezer within 6 h. By this method, blood samples were frozen solid in no more than 30 sec. The samples were stored in the freezer for up to 10 mo before analysis. These collection and storage conditions have not appreciably altered thiamin levels (S. Brown, coauthor, Department of Fisheries and Oceans, unpublished data).
Egg samples were taken between 2 to 12 h after water hardening. Eggs (15-25 mL) from each female salmon were collected, drained of water by blotting, placed in 25-mL plastic screw-top centrifuge tubes and immediately frozen on crushed dry ice for up to 2 h before transfer to storage in a -70°C freezer. Egg samples were processed for thiamin analysis 2-3 mo after the initial sample date.
Sac-fry (a.k.a., larvae/alevins/eleutheroembryos) from the control LC and LCH stocks were taken on February 8, 1995, 695 dd after fertilization. Sac-fry samples from the CL, OL and GP stocks were taken on February 21, 622, 677 or 758 dd after fertilization, respectively. Five sac-fry from each female were euthanized by pipetting them onto fiberglass mesh, blotting them briefly on absorbent paper (atop the mesh), transferring them to Whirl-pak bags, and quick-freezing them live on crushed dry ice. Euthanasia by this method (Canadian Council on Animal Care 1993) was instantaneous and was approved by the NY Department of Animal Welfare. The frozen samples were stored in a 
70°C freezer within 2 h of sampling, and analyzed for thiamin within 2 mo of the sampling date.
Gradient reversed-phase HPLC was used for thiamin analyses of whole blood, egg and sac-fry samples by the method of Brown et al. (1998b). Briefly, samples consisted of approximately 200 mg of tissue homogenate from a pool of 3 eggs or sac-fry, or from whole blood. Extracts of each sample were run twice on the HPLC system, and the average of the two runs was recorded. Product recovery of samples spiked with free thiamin ranged between 85 and 90%; recovery in samples spiked with TPP ranged between 80 and 90%. TMP recoveries were not determined in these salmon samples, but were measured at 98.8 ± 1.9% in eggs from lake trout (Brown et al. 1998b). Assay detection limits were 0.002, 0.004 and 0.005 nmol/g for free thiamin, TPP and TMP, respectively. The coefficient of variation for total thiamin in an egg sample analyzed six times was 7.47%.
Statistical analyses.
Differences in offspring survival frequencies among females were evaluated for statistical significance with the chi-square goodness of fit test. Owing to heterogeneous variance, thiamin levels were log-transformed and stock differences were evaluated using ANOVA. Ratios of thiamin moieties in the eggs and sac-fry were computed as proportions of each thiamin moiety relative to total thiamin (i.e., 
free thiamin + TMP + TPP) or as the ratio of a single thiamin moiety in sac-fry relative to a moiety in the eggs from the same female. Differences in thiamin ratios were also compared among stocks using ANOVA; because ratio data are inherently normalized, log transformation was not performed. Findings of significance (P < 0.05) were followed with an analysis of standardized residuals (for chi-square evaluations) or with the Bonferroni multiple comparison test (for ANOVA). Finally, regression analyses were used to ascertain which of the thiamin moieties was the best predictor of survival and time-to-death (TTD). Statistical calculations were assisted by Data Desk version 4.1 (Data Description Inc., Ithaca, NY). Values cited in the text are means ± SD, unless otherwise noted.
| |
RESULTS |
Embryonic and larval survival by stock.
Hatching of eggs from all sources began around 400 dd post fertilization, peaked at approximately 450 dd, and was complete by 500 dd. Embryonic survival through hatching varied significantly between maternal sources (X2 = 1591, df 25, P < 0.0001; Table 1). Pre-hatch (embryonic) survival also differed significantly (P < 0.0001) between females within lake sources for all lakes where the Chi-square comparison could be made (i.e., LC, LCH and OL). Pooling all embryonic survival data by lake source revealed a significant difference among lakes (X2 = 963.0, df 4, P < 0.0001); analysis of standardized residuals indicated no difference in embryonic survival between the control LC stock and the syndrome-positive CL stock (X2 = 0.0133, df 1, P < 0.9081).
Maternally specific mortality of eggs from the control LC stock averaged 28.7 ± 29.0% and was most influenced by the poor survival of embryos from LC 5 and 8. Sac-fry mortality of the LC stock from hatching until density normalization (785 dd for the LC stock) averaged 5.1 ± 5.5%. There was no subsequent mortality recorded in LC sac-fry through yolk absorption (approximately 1,000 dd) or in LC fed-fry.
The maternally specific range in egg mortality from the syndrome-positive CL salmon could not be calculated because an interruption in the flow resulted in the loss of all CL progeny separated by maternal source. Thus, mortality for the CL stock was estimated based on data from the pooled surplus eggs. An examination of the dead eggs cleared with 10% acetic acid indicated that more than 95% of the dead CL eggs were infertile. This finding indicated the uncommonly high egg mortality of the pooled CL eggs (i.e., 30.5 ± 9.4%) was likely due to the infertility of one of the three females spawned. Between hatching and density normalization (622 dd for the CL stock), mortality of the pooled CL stock was less than 1%. After normalization, the CL sac-fry exhibited 100% mortality, with clinical signs that were consistent with previous observations of Cayuga syndrome preceding death (Fisher et al. 1995a and 1995b).
Egg mortality of the thiamin-fortified LCH stock averaged 86.9 ± 15.9%. This high egg mortality was greatly influenced by the 100% infertility of the eggs from LCH 2 and the > 80% infertility of the eggs from the other three LCH females. Sac-fry mortality before density normalization was also high relative to the control LC stock, averaging 46.6 ± 30.3%. No mortality was observed in the remaining LCH sac-fry or fed-fry after the normalization date, the period in which mortality occurs in the syndrome-positive CL stock.
Mortality of the OL embryos averaged 20.2 ± 11.7% through hatching and, like the LC control stock, was highly dependent on the maternal source of eggs. Prior to density normalization (704 dd), mortality of the OL sac-fry was low (mean 3.3 ± 3.9%). Syndrome-related mortality was observed in sac-fry from 7 of the 10 OL females spawned. The onset of mortality in the affected OL progeny was abrupt, occurring between 725 and 825 dd, depending on the family unit, with 100% mortality reached only a few days after clinical signs were recorded (90% mortality of all syndrome-afflicted stocks was reached by approximately 850 dd). Clinical signs exhibited by moribund OL sac-fry included yolk-sac opacities; vitelline hemorrhage and congestion; mild subcutaneous and retrobulbar edema; and neurological impairment such as ataxia, reduced photophobicity, convulsions, and eventually paralysis. The mortality pattern and clinical signs observed in these OL sac-fry were consistent with the current syndrome-positive CL sac-fry and with previous studies of CL sac-fry (Fisher et al. 1995a and 1995b). The few syndrome-affected OL sac-fry that survived to 900 dd, when food was initially presented, did not eat and died with substantial yolk reserves remaining.
In contrast to the progeny from the other OL females, survival of sac-fry from OL 5, 6, and 7 exceeded 90% through yolk absorption, which was consistent with the survival of the control LC progeny. Sporadic mortality was recorded in these OL offspring during the subsequent 6 wk of feeding. Nonetheless, food was accepted readily by these stocks, as evidenced by the observation of fecal matter in the incubation cups. Vigor of the progeny from OL 6 and 7 was normal; however, some of the feeding fry from OL 5 exhibited a loss of equilibrium similar to lake trout fry with the swim-up syndrome (Mac et al. 1985). Coincident with this observation, there was a slight increase in mortality in the OL 5 fed-fry after the accumulation of 1,150 dd.
Embryonic mortality of the GP stock was 52.5%, and sac-fry mortality prior to density normalization was 2.2%. After normalization, the mortality pattern and clinical signs in the GP progeny were consistent with those from the syndrome-positive CL salmon and affected OL progeny.
Effect of thiamin treatment on syndrome-afflicted salmon.
Relative to the complete mortality of the untreated syndrome-afflicted sac-fry, thiamin bath treatments substantially improved survival of the syndrome-afflicted salmon from CL, OL, and GP (Table 2). Given the near complete survival after treatment of these syndrome-afflicted stocks, statistical evaluations of thiamin effectiveness were not performed. Treatment also improved survival of the sac-fry progeny from OL 6 and 7, whose untreated sac-fry did not exhibit clinical signs similar to Cayuga syndrome (treated vs. untreated: X2 OL 6 = 9.120, df 1, P = 0.0025; X2 OL 7 = 6.632, df 1, P = 0.010).
View this table:
[in this window]
[in a new window]
|
Table 2.
Effect of a single 1-h bath treatment in 10 g thiamine-hydrochloride/L on survival of Atlantic salmon (Salmo salar) sac-fry from syndrome-positive Cayuga
Lake (CL), and syndrome-unknown Otsego Lake (OL) and Green Pond (GP)
|
|
Thiamin levels in landlocked salmon stocks.
Eggs. Egg levels (log-transformed) of each thiamin moiety differed significantly between stocks (Fig. 1; F Log TMP = 5.862, P = 0.0021, df 23, 4, F Log TPP = 15.298, P < 0.0001; F Log TMP+TPP = 11.309, P < 0.0001; F Log Free = 31.287, P < 0.0001; F Log Total = 33.191, P < 0.0001). The LCH stock fed the thiamin-fortified diet had significantly more of each thiamin moiety in its eggs than was detected in eggs from syndrome-positive stocks. Eggs from the LCH stock also had significantly more TMP, free thiamin and total thiamin than eggs from the LC progenitor stock.
View larger version (27K):
[in this window]
[in a new window]
| Fig 1.
Mean thiamine levels in eggs from thiamine-fortified Little Clear Hatchery (LCH) stock (n = 4 females); Little Clear Pond (LC) control stock (n = 10); and syndrome-positive Cayuga Lake (CL), Otsego Lake (OL), and Green Pond (GP) stocks (n = 3, 10 and 1, respectively). a, log-transformed value significantly different from LCH, P < 0.05. b, significantly different from LC, P < 0.05.
|
|
Of greater interest in this study was the comparison of thiamin levels between the syndrome-negative LC control salmon that feed on rainbow smelt and the syndrome-positive stocks (i.e., CL, OL and GP) that feed primarily on alewife. TPP levels in eggs from the control LC stock were 82%, 95% and 5.6-fold greater than that found in eggs from the CL, OL and GP stocks, respectively. Free thiamin levels in eggs from the LC stock were 3-, 7.3- and 30.4-fold greater than those detected in the CL, OL and GP stocks, respectively (Fig. 1). Total thiamin levels in eggs from the LC stock were 2.56-, 3.80- and 19.2-fold greater than those detected in eggs from the CL, OL and GP stocks, respectively. Eggs from the LC stock had significantly more TPP, free thiamin and total thiamin than eggs from the OL and GP stocks (Fig. 1). There were no significant differences in egg thiamin levels among any of the syndrome-positive stocks.
Sac-fry.
The overall pattern of significant statistical differences in thiamin moiety concentrations detected between stocks in eggs was also observed in their sac-fry (F Log TMP = 9.3012, P = 0.0002, df 20,4; F TPP = 18.304, P < 0.0001, F Log TMP+TPP = 8.642, P = 0.0003; F Log Free = 24.319, P < 0.0001; F Log Total = 20.450, P < 0.0001). Free thiamin, its phosphate esters and total thiamin were significantly more concentrated in sac-fry from the syndrome-negative LCH and LC stocks than in sac-fry from any of the syndrome-positive stocks (Fig. 2). Thiamin monophosphate levels in sac-fry from the control LC stock were 2.81-, 4.96- and 14.62-fold greater than sac-fry from CL, OL and GP, respectively. Levels of TPP in the LC sac-fry were 4.94-, 3.99- and 17.76-fold greater than that detected in the CL, OL and GP sac-fry, respectively. Free thiamin levels in sac-fry from LC were 3.37-, 4.10- and 10.1-fold greater, while total thiamin levels were 2.94-, 4.13- and 14.07-fold greater than those observed in sac-fry from CL, OL and GP, respectively. There were no significant differences in sac-fry thiamin levels between any of the syndrome-positive stocks.
View larger version (24K):
[in this window]
[in a new window]
| Fig 2.
Mean thiamine levels in landlocked Atlantic salmon sac-fry from thiamine-fortified Little Clear Hatchery (LCH) stock (n = 1 female); Little Clear Pond (LC) control stock (n = 10); and syndrome-positive Cayuga Lake (CL), Otsego Lake (OL) and Green Pond (GP) stocks (n = 3, 10 and 1, respectively). a, log-transformed value significantly different from LCH, P  0.05. b, significantly different from LC, P 0.05.
|
|
Thiamin ratios.
Thiamin moieties in the eggs and sac-fry were evaluated as one mean to ascertain whether proportionate differences in these analytes might help to explain the syndrome manifestations in the landlocked salmon. Unphosphorylated (i.e., free) thiamin was the principal form of the vitamin detected in eggs from all the stocks, averaging 73 ± 18% of the total (Fig. 2). Thiamin pyrophosphate constituted 22.1 ± 14.1%, and TMP amounted to 4.9 ± 6.41% of the total thiamin remaining in the eggs. The ratios of each thiamin moiety to total thiamin in the eggs differed significantly between stocks (F TMP/Total = 4.545, P = 0.0074, df 23, 4; F TPP/Total = 10.166, P < 0.0001; F Free/Total = 13.480; P < 0.0001), with a substantially greater proportion of free/total thiamin detected in eggs from the syndrome-negative LC and LCH stock relative to eggs from the syndrome-positive stocks (Table 3). Significant multiple comparisons are also depicted in Table 3.
View this table:
[in this window]
[in a new window]
|
Table 3.
Thiamine utilization and conversion between eggs and sac-fry in landlocked Atlantic salmon species from New York State waters1
|
|
In the sac-fry, TPP constituted over half of all thiamin detected when data for all stocks were combined (52.8 ± 7.7%). Free thiamin was proportionately reduced (33.2 ± 8.1%), while the ratio of TMP/total thiamin increased nearly threefold over that detected in the eggs (13.9 ± 5.1%). Ratios of the thiamin moieties to total thiamin in sac-fry differed significantly between stocks for TPP and unphosphorylated thiamin only (F TPP/Total = 9.611, P = 0.0002, df 20, 4; FFree/Total = 8.350, P = 0.0004). Degrees of freedom were reduced relative to the egg results because only one LCH female was represented in the ANOVA. Significant multiple comparisons are also depicted in Table 3.
To assess whether the syndrome affected thiamin utilization, ratios of sac-fry:egg-thiamin were compared between stocks (Table 3). These ratios differed significantly between stocks for each thiamin moiety except TMP (F TPP = 9.474, P = 0.0002, df 20,4; F TMP+TPP = 6.673, P = 0.0014; F Free = 4.195, P = 0.0014, F Total = 3.3260, P = 0.0305). Significant multiple comparisons are indicated in Table 3.
Although not a formal kinetic study of thiamin metabolism, ANOVA of the ratio of TMP + TPP in sac-fry:free thiamin in eggs provided a means to consider potential stock differences in the conversion of free thiamin to these physiologically active phosphorylated forms. This ratio differed significantly between stocks (F = 21.772, P < 0.0001, 20, 4), and was significantly greater in the syndrome-positive OL and GP stocks relative to the control stock (Table 3). Maternally specific data could not be obtained from the CL sac-fry; hence, this ratio was not computed.
Survival of salmon through first-feeding vs. egg and sac-fry thiamin levels.
Survival from hatch through first-feeding was directly related to both egg and sac-fry thiamin levels, revealing a steep sigmoidal dose-response curve (Fig. 3). The lack of data coupling intermediate survival and thiamin levels precluded probit transformations; thus, the interpolated best-fit curves of the arithmetic data are presented (Fig. 3). The interpolated best-fit curves of these relationships reflect progeny survival data from each of the OL and GP females and the three LC females with the lowest levels of thiamin. These data suggest that egg levels of free and total thiamin above 0.8 nmol free thiamin/g and 1.1 nmol total thiamin/g should yield survival rates through first-feeding of >80%.
View larger version (26K):
[in this window]
[in a new window]
| Fig 3.
Interpolated best-fit curves of survival of landlocked Atlantic salmon through first feeding versus egg and sac-fry concentrations of free and total thiamine.
|
|
Regressions using a justifiably truncated data set (i.e., data from the three syndrome-positive OL females with the highest amount of thiamin detected, the three OL females whose progeny exhibited relatively good survival and the three syndrome-negative LC control females with the lowest amount of thiamin) indicated that free and total thiamin levels in both eggs and sac-fry were highly significant predictors of survival through yolk absorption, and that the concentrations in sac-fry provided a slightly better indicator than those detected in eggs (Table 4).
View this table:
[in this window]
[in a new window]
|
Table 4.
Regression equations for survival of landlocked Atlantic salmon (Salmo salar) sac-fry from hatch through first-feeding versus egg and sac-fry thiamine levels
|
|
Significant positive linear relationships were also observed between the concentrations of total thiamin, and the summation of TMP and TPP in sac-fry and the 75% quartile of their TTD (Fig. 4; F Total = 7.89, df 6, 1; F TMP+TPP = 7.83).
View larger version (27K):
[in this window]
[in a new window]
| Fig 4.
Maternally specific concentrations of thiamine pyrophosphate (TPP), thiamine monophosphate plus pyrophosphate (TMP + TPP) and total thiamine in syndrome-positive sac-fry versus the 75% quartile for their time-to-death (n = 59 sac-fry/female, SD 7). *, P < 0.05, for Ho:  = 0, Ha:  0.
|
|
Thiamin in blood.
Thiamin in the blood of all salmon captured in the field was analyzed to ascertain whether differences in stock, sex and diet were reflected in the blood and/or egg profiles of each thiamin moiety.
Comparison by stock and sex.
Blood levels of thiamin differed significantly between salmon stocks (F Log TMP = 5.3057, P = 0.0006, df 34error, 6stock, 1sex, 2,interaction; F Log TPP = 6.3107, P = 0.0002; F Log TMP+TPP = 8.4251, P < 0.0001, F Log Free = 3.8534, P = 0.0049; F Log Total = 10.743, P = < 0.0001). The LCH stock fed the thiamin-fortified diet had significantly greater whole blood levels of TPP, free thiamin and total thiamin than the syndrome-positive CL and OL stocks (Table 5); blood from the LC control stock also had significantly higher TPP and total thiamin concentrations than the syndrome-positive CL and OL stocks. Thiamin levels in the blood of the two immature male salmon captured from SL did not differ significantly from the LC or LCH levels, nor were there significant differences detected within the syndrome-negative stocks (i.e., between LC and LCH) or the syndrome-positive stocks (i.e., between CL, OL and GP). Male salmon had 50.5, 33, 32 and 35% less TMP, TPP, free and total blood thiamin, respectively, than female salmon; however, this sex difference was significant for only TMP and TMP plus TPP (F Log TMP[sex] = 5.5391, P = 0.0245, F Log TMP+TPP[sex] = 6.913, P = 0.0128). Interaction terms between stock and sex were not significant for any of the thiamin moieties.
View this table:
[in this window]
[in a new window]
|
Table 5.
Blood concentrations of thiamine in landlocked Atlantic salmon (Salmo salar) from New York State lake stocks1
|
|
Blood and egg thiamin levels.
Maternal blood TPP levels averaged 0.263 ± 0.05 nmol/g for those females with syndrome-positive offspring (i.e., data from all CL, GP and OL females except OL 5, 6 and 7) versus 0.439 ± 0.12 nmol/g for those feral females with syndrome-negative offspring (i.e., all LC females and OL females 5, 6 and 7). Total blood thiamin averaged 0.329 ± 0.05 nmol/g in those females with syndrome-positive progeny and 0.616 ± 0.19 nmol/g in those with syndrome-negative offspring.
Blood levels of thiamin in the female salmon were subsequently regressed against the concentration detected in their eggs to ascertain whether thiamin deposition was affected in feral salmon whose progeny were syndrome-positive versus those with syndrome-negative progeny. Significant positive linear relationships were detected in salmon with syndrome-negative progeny (F Blood TPP vs. TPP Egg = 8.29, df 11, P = 0.0150; F Blood Total vs. Egg Total = 9.38, P = 0.0108, F Blood TPP vs. Egg Free = 11.6, P = 0.0059) (Fig. 5). However, there were no significant linear relationships observed between blood and egg thiamin levels in females with syndrome-positive progeny.
View larger version (27K):
[in this window]
[in a new window]
| Fig 5.
Regressions of total egg thiamine versus total maternal blood thiamine in landlocked Atlantic salmon stocks that produced syndrome-positive and syndrome-negative offspring.*, P < 0.05, for Ho: = 0, Ha: 0.
|
|
| |
DISCUSSION |
Stock assessment of Cayuga syndrome outside the Finger Lakes region.
The observation of a thiamin-responsive mortality syndrome in larval landlocked Atlantic salmon from OL and GP indicates that, within New York State, the Cayuga syndrome is not confined only to the Finger Lakes region. Because thiaminase-rich alewife are considered the major dietary component for salmon in these lakes, these current epizootiological data support the hypothesis that the thiamin deficiency responsible for the mortality of the salmons' progeny is attributable to a maternal diet high in thiaminase.
The role of thiamin in early life stage survival appears to be most critical during yolk absorption and the initiation of exogenous feeding. Thus, in this study there was no consistent relationship observed between embryonic survival and egg thiamin concentrations. Indeed, correlating the present data of total egg thiamin (X) to embryonic mortality (Y) might suggest the dubious conclusion that depressed thiamin levels actually increase embryonic survival (e.g., LC: r = (-0.514), OL: r = 0.146). The absence of a consistent effect of thiamin on hatchability in landlocked Atlantic salmon is consistent with results from recent studies on anadromous Atlantic salmon afflicted with M-74 (Amcoff et al. 1998) and on other salmonid species from the Great Lakes (Hornung et al. 1998).
To demonstrate a nutritional deficiency it is considered necessary to show not only that levels of the nutrient in question are depressed, but also that the clinical signs associated with the supposed deficiency can be reversed with treatment (Combs 1992). In the current study it was demonstrated that two additional salmon stocks within New York State (i.e., OL and GP) are suffering from a thiamin deficiency akin to that first observed in salmon from CL. Total thiamin levels in CL sac-fry analyzed in the current study were over twice those detected in the previous study (mean94 y class = 0.777 nmol/g, mean93 y class 
110 ng/g = 0.326 nmol/g, Fisher et al. 1996). In contrast, the control LC stock had thiamin levels nearly 70% greater than those detected in the 1993 year class (i.e., 3.06 nmol/g vs. 1.82 nmol/g
the equivalent of 615 ng/g). Given the small sample sizes of these respective studies, the differences in analytical methodology, and the slight variations in sampling times, the variation in sac-fry thiamin levels between years was not unexpected. In either case, it is clear that the thiamin levels in syndrome-afflicted sac-fry are significantly depressed relative to unaffected control stocks.
Thiamin deposition and utilization.
To identify mechanism(s) involved in yielding thiamin deficient sac-fry, we investigated how thiamin moiety ratios changed during development. Proportionately more free thiamin was detected in the control LC and LCH stocks than in the syndrome-afflicted stocks. As egg levels of thiamin were generally reflective of those found in whole blood, regardless of stock, the current results suggest that the transovarian deposition process was not impaired. Thus, either the diet consumed by the adult salmon producing syndrome-positive offspring is deficient in thiamin, or the vitamin is inactivated prior to absorption.
Muscle tissue of alewife reportedly contains more thiamin (Fitzsimons et al. 1998) than is needed to prevent the gross deficiency signs documented in salmonids fed thiamin-deficient diets (Halver 1957); yet, thiamin is rapidly degraded by thiaminase in homogenates of raw alewife (Gnaedinger 1964, Ji and Adelman 1998). Atlantic salmon fry-fed diets of raw thiaminase-rich clupeids, or raw clupeids mixed with thiamin-rich beef liver (50/50), eventually succumbed to thiamin deficiency, with mortality ranging between 45 and 73% (Saunders and Henderson 1974). Therefore, we maintain that egg levels of thiamin are reduced in syndrome-positive stocks owing to inadequate maternal absorption of the vitamin (Fisher et al. 1995a).
The present results also suggest that phosphorylation of thiamin was not impaired by the Cayuga syndrome, as there was a consistent shift in all stocks from the unphosphorylated vitamer principally present in the egg to the physiologically active mono and pyrophosphate forms most abundant in the sac-fry. If phosphorylation was impaired, then the treatments performed with unphosphorylated thiamin-hydrochloride would have been ineffective. What appears most likely, therefore, is that a critical level of the free form of the molecule is required in the eggs to facilitate survival through first-feeding.
Thiamin threshold requirements for survival in Atlantic salmon.
It is premature to emphasize absolute tissue-thiamin requirements for the survival of Atlantic salmon from hatch through first-feeding. Nonetheless, it is clear from the regressions of survival versus tissue thiamin that distinct steep thresholds exist for total thiamin at around 1.1 nmol/g egg and 0.8 nmol/g sac-fry (Fig. 4). Females with whole blood TPP and total blood thiamin levels above ~0.31 and 0.44 nmol/g, respectively, should also produce sac-fry with high survival. That there was no linear relationship observed between blood and egg thiamin in females with syndrome-positive progeny suggests that a limiting systemic concentration of thiamin is required by the female to satisfy physiological requirements before transovarian deposition will occur (Fig. 5). The noise in the relationships of Atlantic salmon survival to blood, egg and sac-fry thiamin at or near the threshold requirement is consistent with recent findings in lake trout (Fitzsimons and Brown 1998), and Pacific salmon from the Great Lakes (Hornung et al. 1998). The egg thiamin requirement for the larval survival of each of these species is also near 1 nmol thiamin/g; however, more study is needed to identify specific requirements. Brown et al. (1998a) cite a 63% incidence of EMS in female lake trout with red cell TPP levels below 0.33 nmol TPP/g. While the current whole blood results cannot be directly compared against Brown et al.'s findings with red cells, it is noteworthy that again there appears to be little difference in the blood levels required to prevent EMS-type syndromes in salmonids, regardless of species.
Results from the current study support the role of thiaminase as the primary cause of thiamin deficiency and Cayuga syndrome in sac-fry of landlocked Atlantic salmon from New York State. Although thiamin deficiency does not appear to directly disrupt oogenesis, spermatogenesis, fertilization success or embryonic development of salmonids, the thiamin-responsive Cayuga syndrome effectively prevents the transitional development of endogenous-feeding yolk-sac larvae to exogenous-feeding fry. As such, thiamin could be considered indirectly critical for the successful reproduction of salmonid fishes. The Cayuga syndrome manifest in Cayuga Lake Atlantic salmon represents the only case, to our knowledge, where a vitamin deficiency has been correlated to the complete reproductive failure of a feral animal population (i.e., 100% mortality of offspring of all family units). The current results suggest specific minimal thiamin requirements that must be met to prevent deficiency-related mortality in the salmon larvae. Studies to concurrently measure thiamin and thiaminase levels in salmon gut contents and prey are needed to substantiate the epizootiological link of the Cayuga syndrome to a thiaminase-rich diet.
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
, 5,
Abstract
Introduction
Abstract
- tocopherol acetate/g, 110.2 IU ergo calciferol/g, 1,653 USP retinyl palmitate or acetate/g); 12) trace mineral mixture, 0.5 g/kg (mixture contains 74.96 mg zinc sulfate/g, 20.06 mg manganous sulfate/g, 1.5 mg copper sulfate/g, and 10.01 mg potassium iodate /g).
