Morphometric parameters
We did not observe any changes in morphometric parameters during the experiment with the exception of male GSI (Figure
1b) and female K (Figure
2a). We observed the GSI in males was lower at the end of the experiment in males that had been stripped of milt during the experiment. This is expected and may be explained by an effect of handling and stripping on the males resulting in higher demands on the gonad for milt (and thus a reduced gonad weight) when compared to unstripped fish. The difference detected in female K was likely due to slightly, but not significant (p = 0.0564), longer mean female fork length in the exposed stripped tanks. Changes in morphometric parameters were not observed during the shortened 5-day fathead minnow spawning assay in the study by Kovacs (2007). In the 5-day study of 7 mechanical mill effluents tested in the 5-day FHM spawning assay, an increased gonad weight from two mill effluent exposures was the only morphometric parameter observed to change [
16], while in a study of 7 Kraft mill effluents, female body weight was lower in only one treatment [
17]. The authors of these studies state that these differences do not indicate effluent-related effects since measurements were taken at the end of the experiment after the fish had been removed from the effluents and kept in well water for five days [
16,
17]. In the longer 21-day assay, no changes in morphometric parameters were seen in males but in females mean fork-length was lower in one of the treatments (a 2% Kraft mill effluent) while mean weight and condition factor was higher in another treatment (a different Kraft mill effluent at 40% concentration) [
18].
Egg production in female fathead minnows
Fewer eggs were laid in the female groups exposed to a TMP effluent. In order to assess if the females were not ovipositing despite ovulating, we stripped ovulated eggs from the exposed females. Reasons for females ovulating but not spawning might include males not detecting the release of pre- and post-ovulatory pheromones due to pheromone binding/adsorption to effluent constituents or blockage of male olfactory receptors resulting in male failure to initiate spawning behaviour or perhaps ovulated females are able to detect non-ideal conditions for their eggs and decide to forgo spawning until conditions improve.
Our results (Figure
4) clearly demonstrate that ovulation is impaired in effluent-exposed fish compared to controls because we were unable to strip many ovulated eggs from the exposed females. If the exposed females were indeed ovulating, we would expect the total eggs per tank (oviposited + stripped) to be similar to the total eggs in the control tanks, only with a higher proportion of the total eggs being stripped compared to controls.
In teleost fish, ovulation is triggered by a surge in luteinizing hormone (LH) from the pituitary, which is under the direct stimulation of gonadotropin-releasing hormone [
19]. Dopamine and GABA are very important neurotransmitters in the reproductive axis because they respectively inhibit and stimulate LH release [
20‐
22]. This inhibitory input by dopamine is potent, such that co-injection of a dopamine antagonist with a GnRH agonist is required to induce the LH surge and spawning in teleosts [
23].
In vitro experiments on effluent extracts by Basu et al. [
9] identified that extracts of a Canadian TMP effluent contained ligands for the dopamine type-2 receptor among other neuroendocrine targets important to reproductive control. In follow-up experiments, several effluents [
13] and hardwood [
10] and conifer [
12] feedstocks have also identified potential neuroactivities.
It is not possible to measure LH in FHM because the LH radioimmunoassay has not been developed for this species. Nevertheless, we hypothesize that LH release may be reduced such that the female FHM exposed to the PPME were unable to ovulate. Prior to our work, a study by [
24] on a white sucker (
Catostomus commersoni) population exposed to a bleached Kraft mill effluent from a mill in Terrace Bay, ON, Canada, clearly demonstrated that LH and sex steroids in exposed wild white suckers were depressed compared to those from control sites. Upon injecting females with a GnRH agonist, the size of the resulting LH surge observed in control fish was not seen in those populations exposed to effluents, nor did ovulation occur (while it did in controls). This study indicated that either the pituitary had lost GnRH sensitivity (due to lesions or perhaps other mechanisms) or that other inhibitory signals, such as dopamine, might be suppressing LH release. Due to the inhibitory dopaminergic tone, for example, it is necessary to co-inject dopamine receptor antagonists with GnRH to stimulate spawning in many fish species.
It is also possible that while an LH surge is indeed occurring in the exposed females, the ovaries of these individuals may not be responding appropriately, but this is speculative because we did not directly measure steroid product.
In vitro studies (Gibbons et al. [
25]; McMaster et al. [
26]) have demonstrated an impaired induction of steroidogenesis in the gonadal tissue of fish exposed to pulp mill effluents: both testosterone and 17β-estradiol production was inhibited in ovarian follicles stimulated by human chorionic gonadotropin in tissues collected from wild white sucker (
Catostomus commersonii) downstream of a sulphite pulp mill [
26], and forskolin-stimulated 17β-estradiol production was inhibited in ovarian follicles collected from trout-perch (
Percopsis omiscomaycus) downstream of a thermomechanical/de-inked pulp mill [
25]. However, the significance of inhibited steroid production is unclear since previous studies have demonstrated that reduced spawning rates in fish exposed to effluents do not appear to be associated with several steroidogenic endpoints [
27].
Additional mechanisms of action of pulp mill effluents that may help explain the results seen in our study. Pharmacological inhibition of steroidogenic brain aromatase or 3β-hydrosteroid dehydrogenase/Δ
5-Δ
4 isomerase (3β-HSD) was also able to elicit rapid inhibition of spawning in fathead minnows. The aromatase inhibitor fadrozole results in decreased brain aromatase activity in fathead minnows and in these fish, rapid spawning inhibition and impairment of oocyte maturation was observed [
28]. Plasma estradiol and vitellogenin concentrations were also decreased in these females exposed to fadrozole. Inhibition of 3β-HSD by trilostane also caused a rapid inhibition of spawning in fathead minnows as well as decreased levels of vitellogenin [
29], so the observed effects in our study could be explained by effects at the level of the gonad or liver in addition to effects on brain. However, many effluents are strongly anti-reproductive, but inconsistently affect steroidal pathways [
30,
31]. It is possible that effluents are inducing a stress response and that increased cortisol levels play a part in the inhibition of ovulation, but there is little evidence for stress induction as measured by cortisol in fish [
2,
32]. Regardless of the hormonal response, pulp effluents are acting somewhere in the hypothalamus-pituitary-gonad axis and it is more than likely that this includes effects at all three levels.
Milt production in male fathead minnows
We observed no effects from exposure to a TMP effluent that inhibits spawning on male milt production (Figure
5), indicating that the reproductive inhibition is a female rather than a male effect. Males are likely detecting sex pheromones that are released by females to signal to males that they have ovulated, since the males are producing milt.
Typically, in female cyprinid fish the LH surge is followed shortly thereafter by the release of a pre-ovulatory hormone (17α,20β-P or an analog) from the somatic follicle cells in the ovary, which immediately precedes ovulation. This pre-ovulatory hormone is released to the environment as a sex pheromone and is detected by males. Upon detection, the males also experience an increase in LH and within hours milt volume begins to increase (for review, see [
33]). While we conclude that this TMP effluent inhibits ovulation in female fathead minnows, it is evident that a few eggs are still ovulated and oviposited (Figures
3 and
4). On day 3 of the experiment, we were able to strip eggs from 8 of the 16 females in the effluent treatment and 0 of 16 females in the control tanks. On day 5, we were again able to strip eggs from 8 of the 16 exposed females, and only from 4 of the control females. Presumably, any ovulated eggs in the control tanks had already been oviposited, explaining the difference in egg stripping success rate between the control and exposed tanks. Since some females were still able to ovulate a few eggs in the effluent-treated tanks, despite a dramatic decrease in oviposition (Figure
3), it seems that there may be sufficient pre-ovulatory hormone released to the water to stimulate the appropriate milt production in males.