Fish Physiologyand Biochemistryvol. 6 no. 2 pp 113-120 (1989) Kugler Publications, Amsterdam/Berkeley
The growth and reproductive endocrinology of adult tripioid Pacific salmonids* Tillmann J. Benfey, Helen M. Dye, Igor I. Solar and Edward M. Donaldson Department o f Fisheries and Oceans, Biological Sciences Branch, West Vancouver Laboratory, 4160 Marine Drive, West Vancouver, British Columbia V7V 1N6, Canada Keywords: growth, endocrinology, reproduction, triploid, salmonid, sterilization, aquaculture
This paper describes the effect of triploidy on growth and reproductive endocrinology in the months leading up to and including spawning in rainbow trout, Salmo gairdneri, and pink salmon, Oncorhynchus gorbuscha. Growth rates were the same for diploid and triploid rainbow trout, but triploid female pink salmon were smaller than maturing diploid females and diploid and triploid males of the same age. Triploid males of both species developed typical secondary sexual characteristics and had normal endocrine profiles, although their cycle appeared to be delayed by about one month. Triploid females remained silvery in appearance and showed no endocrine signs of maturation, even at the level of the pituitary. Thus, although triploids of both sexes are genetically sterile, only the females do not undergo physiological maturation.
A major goal of controlled reproduction in salmonids for aquaculture is the prevention of sexual maturation prior to harvest (Donaldson 1986). This can be achieved practically by the production of all-female triploids (Chevassus et al. 1984; Okada 1985; Bye and Lincoln 1986; Donaldson and Benfey 1987). The methods used to produce and identify triploid fish have been thoroughly reviewed by Purdom (1983) and Thorgaard (1983). Triploids are genetically sterile because their cells cannot undergo normal meiotic division to produce euploid gametes. Only very small numbers of triploid cells complete meiosis, and these only reach full maturity in males (i.e. functional spermatozoa), and are aneuploid (Lincoln 1981a; Lin-
coin and Scott 1984; Okada 1985; Allen et al. 1986; Benfey et al. 1986). Post-meiotic oocytes probably do not reach full maturity in triploid females because there is no vitellogenin produced for oocyte growth (Benfey et al. 1988b). In spite of their sterility, triploid males still de: velop large testes because of the enormous number of pre-meiotic spermatogonia and primary spermatocytes that are produced. As a result, the gonads are generally larger in triploid males than females, and these males appear to mature. Triploid male rainbow trout, Salmo gairdneri, have normal sex steroid levels (Lincoln and Scott 1984) and exhibit all the physiological and morphological changes associated with maturation in diploid males (Thorgaard and Gall 1979; Chevassus et al. 1984; Lincoln and Scott 1984; Okada 1985; Benfey et al. 1986).
* Reported in part at the Third InternationalSymposiumon ReproductivePhysiologyof Fish, St. John's, Newfoundland,August2-7. 1987.
114 Thus, in an aquacultural sense, it is only the triploid females that can be considered to be sterile (Donaldson and Benfey 1987). Triploid female rainbow trout show no change in plasma sex steroid levels through the normal pre-spawning period when these hormone levels become highly elevated in diploid females (Lincoln and Scott 1984; Nakamura et al. 1987). Even one year prior to spawning, diploid females have higher plasma sex steroid, gonadotropin and vitellogenin levels than triploid females ever reach (Sumpter et al. 1984). Little is known about the reproductive endocrinology of triploids in other' species. Johnson et al. (1986) measured plasma 17/3-estradiol levels in coho salmon, Oncorhynchus kisutch, at the onset of maturity and found no difference between diploid and triploid females, but vitellogenin was only detectable in the diploids. Plasma sex steroid levels were found to be the same in spermiating diploid and tripioid male "plaice, Pleuronectes platessa (Lincoln 1981a) and in post-ovulatory diploid and triploid female plaice x flounder, Platichthys fiesus, hybrids (Lincoln 1981b). In amphibians, Cayr o l et al. (1985) showed that triploid newts, Pleurodeles waltl, of both sexes have seasonal cycles in sex steroid levels, but the hormone levels are never as high as in diploids. The major site for production of sex steroids in maturing fish is the steroidogenic cells within the gonads (Scott 1987). The available data from studies with rainbow trout indicate that these steroidogenic cells must be functioning in triploid males but not females. These steroids are known to have a profound effect upon gonadotropin synthesis and secretion by the pituitary in normal fish, which can be negative or positive depending on their stage of maturity (Peter 1983; Kah 1986). The aim of this study was to examine the reproductive endocrinology, both at the level of the gonad and the pituitary, of triploid rainbow trout and pink salmon, Oncorhynchus gorbuscha.
Materials and methods
Pink salmon eggs and sperm were obtained from the Quinsam River Salmonid Enhancement Pro-
gram Hatchery (Department of Fisheries and Oceans) and brought to the West Vancouver Laboratory for fertilization. Triploids were produced by hydrostatic pressure shock using a 40 ml standard French pressure cell and Carver laboratory press. Pressure shocks of 1,2, 3 or 4 minutes at 10,000 psi were used on separate batches of eggs, always 15 minutes after fertilization and incubation at 10.5~ All the pressure treated groups had to be combined prior to ploidy determination. The rainbow trout used were those previously described by Solar and Donaldson (1985) and Benfey et al. (1986). Female rainbow trout were sampled one year later than the males. Diploids and triploids of both species were identified by propidium iodide flow cytometry (Allen 1983), after which each fish was marked externally with a coded tag. Sampling of the fish began in July, as they entered their first reproductive phase. Spawning was expected about 6 months later, as 2-year-olds for pink salmon of both sexes and for male rainbow trout, and as 3-year-olds for female rainbow trout. Fish were anesthetized in 0.04~ 2-phenoxyethanol and had their weight and fork length measured. Blood was collected from the caudal sinus into heparinized vacutainer tubes or syringes, and then centrifuged to separate the plasma. Plasma samples were stored at -30~ until they were assayed. All the pink salmon developed bacterial kidney disease shortly after the first sampling, and further sampling was discontinued to minimize stress on these fish. All the surviving fish died in September due to a water failure. These fish again had their weight and fork length measured, and their gonads were removed and weighed for the calculation of gonadosomatic index (GSI). At the same time, pituitaries were removed and each was separately homogenized in 1 ml of radioimmunoassay buffer and stored at -30~ until assayed for gonadotropin content. All the female rainbow trout (12 diploids and II triploids) died in mid-December, again due to a water failure. At this date the diploids had just begun spawning (3 of the females had ovulated). The pituitaries from these fish were also homogenized and stored for later assay. Sampling of the male rainbow trout (5 diploids and 5 triploids) continued through the spawning period
115 until mid-March. No pituitaries were collected from these fish. Plasma steroid hormone levels in rainbow trout were measured using the radioimmunoassays described by Dye et al. (1986) for l l-ketotestosterone and Van Der Kraak et aL (1984) for testosterone, 17/%estradiol and 17~-hydroxy-20/3-dihydroprogesterone (17,20-P). Plasma testosterone and 17/3estradiol in pink salmon were measured using direct t25I kits (catalog nos. 1102-D and 1018, respectively, Radioassay Systems Laboratories Inc., Carson, California). Both these assays were validated by demonstrating that plasmas dilute parallel to the standards, and that unlabelled hormone added to plasma samples at known concentrations is recovered by the assays in the correct amount. Gonadotropin and vitellogenin were measured using the radioimmunoassays described by Pickering et al. (1987) and Benfey et al. (1988a), respectively, except that an immunologically less potent form of purified hormone was used for the gonadotropin assay.
The diploid and triploid pink salmon could be separated into two distinct groups, based both on their size and their endocrine status: those that were maturing (i.e. most of the diploid females and all of the diploid and triploid males) and those that were immature or sterile (i.e. 2 of the diploid females and all of the triploid females, respectively) (Table 1). Of the two diploid females judged to be immature, one had oocytes at endogenous vitellogenesis and the other had small, atretic oocytes; oocytes of the remaining diploid females were at late stages of exogenous vitellogenesis. The maturing fish were larger both in weight and length, and had larger gonads and GSIs. The maturing diploid females had the highest plasma testosterone and pituitary gonadotropin levels, followed closely by the maturing diploid and triploid males. These hormones were low or not detectable in the immature diploid and sterile triploid females. 17/~-Estradiol and vitellogenin were high only in the maturing diploid females. Plasma gonadotropin was not detectable in any of the fish.
There was no significant difference in growth rate between diploid and triploid male rainbow trout (Fig. 1), and both groups had similar endocrine profiles (Fig. 2). Testosterone and ll-ketotestosterone levels increased gradually from July to November, at which point testosterone levelled off and l l-ketotestosterone rose dramatically. Both steroids began to decline in January with the onset of spermiation, at which time 17,20-P levels began to rise. There was a slight lag in the increase in androgen levels observed in the triploids, with both testosterone and ll-ketotestosterone levels being significantly lower than in the diploids in September and again in December (p < 0.05), but higher than in the diploids in March (p < 0.05 for testosterone only). Plasma gonadotropin levels were low prior to spermiation, but began to rise in January. There was only a short peak of plasma gonadotropin in the diploid males, whereas levels continued to rise in the triploid males and were significantly higher by mid-March (p < 0.01). Growth rates were also the same for diploid and triploid females (Fig. 1). The females were larger than the males because they were one year older. Plasma steroid hormone and gonadotropin levels were low or not detectable in triploid females throughout the normal pre-spawning period (Fig. 3). Diploid females had constantly elevated 17/% estradiol and increasing testosterone levels until ovulation, at which time levels of these hormones began to drop. There was a sharp rise in 17,20-P levels in the three females that ovulated in December. Plasma gonadotropin levels began to rise in the diploids with the approach of spawning, and by D~cember were significantly higher than in the triploid females (p < 0.01 for the non-ovulated and p < 0.001 for the ovulated females). The ovulated diploids had significantly higher plasma 17,20-P and gonadotropin levels in December than those that had not yet ovulated (p < 0.001), but 17/~estradiol and testosterone levels were not significantly lower. Pituitary gonadotropin content was not significantly different between the nonovulated and ovulated diploids (0.26 _+ 0.14 and 0.55 _+ 0.14 mg, respectively), but was higher than in the triploid females (0.0017 _+ 0.0016 mg, p < 0.001 in both cases).
116 Table 1. Growth and endocrine status of diploid ( 2 n ) a n d triploid (3n) pink salmon.
Maturing 2n females
Sample size Weight (kg) Length (cm) Testosterone (ng/ml) 17B-estradiol (ng/ml) Vitellogenin (t~g/ml) Plasma gonadotropin (ng/ml)
0.37 31.6 4.5 1.9 6339 < 2.5
9 • + + + •
0.07 1.7 2.2 1.3 5794
Immature 2n females
Sterile 3n females
Maturing 2n males
2 0.26, 0.23 27.9, 27.6 < 0.72 0.07, 0.11 0.19, 10.3 < 2.5
11 0.26 _+ 0.04 28.9 • 1.1 < 0.72 0.05 + 0.01(6) 0.11 :t 0.02 < 2.5
0.37 31.3 2.7 0.13 0.10 < 2.5
10 • 0.13 • 2.9 _+ 1.1 + 0.02(5) +_ 0.04(9)
Maturing 3n males
0.33 31.4 2.4 0.07 0.11 < 2.5
8 +__ 0.07 • 2.1 • 0.6(7) + 0.03(4) + 0.05(6)
SEPTEMI3ER Sample size Weight (kg) Length (cm) Gonad weight (gin) GSI (%) Pituitary gonadotropin (ng)
2 0.69, 0.36 34.6, 31.1 42.6, 7.2 6.6, 2.1 7450, 1680
2 0.20 27.1 1.13 0.56 1.57
0.29, 28.4, 3.02, 1.04, 0.38,
0.31 28.9 0.09 0.03 1.6
7 +_ 0.11 __. 1.1 :t: 0.04 +_ 0.02 + 2.3
0.46 33.4 38.4 8.9 1538
8 _+ 0.18 + 4.0 + 16.9 • 1.7 _+ 1051
0.44 32.9 36.2 8.8 1473
4 + + • _+ •
0.14 2.8 15.5 1.4 794
Data are shown as mean "_+ standard deviation; numbers in brackets = reduced sample sizes because the remaining fish were below detectable levels.
Both the diploid and the triploid male pink salmon were yellowing and beginning to develop humps when they died in September, with no apparent differences between ploidies. The maturing females had not yet developed secondary sexual characteristics, and so were indistinguishable from the immature and sterile females. As previously discussed (Benfey et al. 1986), there was no difference in secondary sexual characteristics between diploid and triploid male rainbow trout. The diploid females developed normal secondary sexual characteristics (i.e., darkening of the skin, white edges on the ventral fins and a protruding vent), none of which appeared on the triploid females. The triploid females lost more scales during handling than any of the maturing fish.
Fig. 1. Growth rates of diploid and triploid rainbow trout (mean + standard deviations; o = diploid females, 9 = triploid females, a = diploid males, 9 = triploid males).
Triploid females showed no superiority in growth over maturing diploid females, and in fact were growing more slowly in the case of pink salmon. Triploid females of both species were always in the same tank as diploids of the same species, and were thus competing with the diploids throughout their
9 150 u J
z 0r 9 I00
~'~,,,16~ : 1
~- %- ~,- ~ - -