Multiple spawning in several commercial fish species and its consequences for fisheries management, cultivation and experimentation

The strategy of multiple spawning is examined in several commercially important species—in particular turbot, Scophthalmus maximus (L.); halibut, Hippoglossus hippoglossus (L.); cod, Gadus morhua L.; and northern anchovy, Engraulis mordax Girard. Aspects of special relevance to fishery science and commercial and experimental cultivation are discussed. These include changes in batch fecundity and egg size as the spawning season progresses; the timing and methodology of fecundity estimations; ovulatory/spawning rhythms; the effect of temperature on the ageing rate of retained ovulated oocytes and the interovulatory period; ovulation prediction in hand-stripped broodstocks to facilitate the collection of good quality, freshly ovulated eggs, and the effect of maternal stress on the ovulatory/spawning rhythm. The implications of these phenomena are discussed in terms of recommended practices for the culturist, experimentalist and fisheries scientist.     

Ovulation in the merino ewe in the breeding and anoestrous seasons

The pattern of ovulation of Merino ewes was studied by repeated laparoscopy each 14 days in the anoestrous (n = 97) and breeding (n = 87) seasons. In the anoestrous season the proportion of ewes ovulating did not decrease below 11%, 42% of ewes never ovulated and the remainder fluctuated between the two states. On 20 occasions a clear anovulatory period was interrupted by an isolated spontaneous ovulation. In the breeding season the overall mean proportion of ewes with corpora lutea or albicantia at laparoscopy was 87%, 54% of ewes ovulated regularly throughout while in another 31% absence of corpora lutea or albicantia coincided with the follicular phase of an oestrous cycle as evidenced by an appropriately aged corpora lutea at the next laparoscopy. Of the remaining 15% of the flock 3% had anovulatory periods greater than 14 days while the remainder experienced irregular ovulatory cycles--the majority due to short periods of anovulation but some ewes retained corpora lutea for longer than 14 days while others ovulated twice between successive laparoscopies.     

The influence of postovulatory age on the rate of cleavage in in-vitro fertilized rat oocytes

The purpose of this study was to analyze the effect of postovulatory ‘aging’ in the oviduct on the rate of zygotic development. Two ovulatory ages were tested: oocytes collected from the oviducal ampullae 1) soon after ovulation (denoted freshly ovulated) or 2) 7-hour postovulation. All the oocytes were from superovulated immature rats. By manipulation of the timing of the ovulatory hormone treatment, it was possible to place both types of oocytes into sperm suspension from the same pool and at the same time. The oocytes and spermatozoa were coincubated overnight. Cleavage was established by interference contrast microscopy. The time of the first cleavage of ova from the 7-hour postovulation group was clearly advanced. Because the cleavage time curves were not parallel, no reliable estimate of the time difference could be made, but it was clearly in the range of 2 hr. This shift could not be related to any difference in the time of sperm penetration. Both groups of oocytes underwent penetration by spermatozoa at the same time. The time interval between maximal sperm penetration (94% of oocytes in both groups) and maximal cleavage (50% in both groups) was 23 hr in the freshly ovulated and 21 hr in the 7-hour postovulatory eggs. Nor was the difference related to polyspermy, which was approximately 14% in both groups. These results support the hypothesis that developmental processes are under way in the oocyte before fertilization, but at a much slower rate than after fertilization.     

Fertility, ovulation and maturation of eggs in mares injected with HCG

Pony mares were observed from January to August for incidence of oestrus, duration of oestrus, length of the oestrous cycle and for ovulation and fertility after injection of HCG. From January to 15 May most mares showed oestrus but the duration of oestrus was quite variable and few mares ovulated in response to HCG. From 15 May to 17 August oestrous cycles were more regular and ovulation was induced within 40-50 h by an intramuscular injection of 1500-5000 i.u. HCG. Pregnancy was established by one mating at a fixed time after HCG in 20 of 69 mares. Degenerate eggs were recovered from the oviducts of anoestrous recently ovulated, mated, unmated and pregnant mares. The first polar body was formed before ovulation in 2 eggs and had not formed in 2 recently ovulated eggs flushed from the oviduct. The second polar body formed after sperm penetration 10-12 h after ovulation. After formation of pronuclei, the first cleavage division occurred at 20 h and the second at 32 h after ovulation. Oestrus was inhibited by progesterone administered by vaginal devices but occurred within 1-3 days in 12 of the 20 mares after withdrawal of the devices.     

Induced spawning by LHRHa and pimozide in the Asian catfish Clarias macrocephalus (Gunther)

Experiments were conducted to determine the optimum dose of luteinizing hormone-releasing hormone analogue (LHRHa) and pimozide (PIM) injected simultaneously to yield a high ovulation rate and produce sufficient eggs in the Asian catfish Clarias macrocephalus . In June 1990, injection of 0.05 or 0.10 μg LHRHa/g body weight (BW) + 1 μg PIM/g resulted in 100% ovulation, while only 80% of gravid catfish injected 0.025 μg LHRHa + 1 μg PIM/g ovulated. Most of the eggs stripped from 6 out of 8 control fish were not mature. Fertilization and hatching rates of LHRHa + PIM-induced fish (75–90% and 39–51%, respectively) were higher than those of control fish (36–39% and 0–1% respectively). In August and September 1990, at gravid catfish ovulated after injection of 0.05–0.10 μg LHRHa + 1 μg PIM/g BW. However, only 20% of the fish given 0.025 μg LHRHa/g + 1 μg PIM/g BW in August ovulated. No eggs could be striped from any of the control fish in August and September 1990. Techniques developed in this study, showed a simple and effective way of spawning captive catfish, C. macrocephalus . A simultaneous intramuscular injection of 0.05 μg LHRHa + 1 μg PIM/g and stripping of eggs at 16–20 h post-injection have been tested to yield high ovulation, fertilization and hatching rates.     

Natural and artificially induced ovulatory models related to lactation in the rat: role of prolactin

The presence and the importance of a preovulatory prolactin (PRL) peak was determined in four, natural or artificially induced, ovulatory models related to lactation in the rat. Gonadotrophin peaks were determined in the afternoon preceding ovulation in four models: postpartum ovulation (PPO), ovulation after the lactational period (AL) (natural models), ovulation after litter removal at midlactation (ML), and ovulation in lactating rats (LR) (artificially induced models). In PPO, AL, and ML rats a preovulatory PRL surge was detected, showing that its presence is a common characteristic of ovulation in the rat. Bromocriptine inhibition of PRL levels in PPO and AL rats did not modify the percentage of rats which ovulated. In contrast, this treatment was able to significantly increase ovulation percentage in ML rats. Moreover, in LR rats strong dopaminergic inhibition of PRL levels, induced by pergolide, was necessary for ovulation to take place, but if pergolide-treated rats were injected with ovine PRL ovulation was completely inhibited. These data suggest that while a PRL surge seems to be always present in natural ovulatory models, it is not essential for ovulation to take place. On the other hand, in artificially induced ovulatory models, suppression of prolactinemia is able to induce ovulation or to increase the percentage of rats which ovulated. This effect of PRL on ovulation may be direct or indirect.     

Synchronization rate, size of the ovulatory follicle, and pregnancy rate after synchronization of ovulation beginning on different days of the estrous cycle in lactating dairy cows

Recently a protocol was developed that precisely synchronizes the time of ovulation in lactating dairy cows (Ovsynch; GnRH-7d-PGF2 alpha-2d-GnRH). We evaluated whether initiation of Ovsynch on different days of the estrous cycle altered the effectiveness of this protocol. The percentage of cows (n = 156) ovulating to the first GnRH was 64% and varied (P < 0.01) by stage of estrous cycle. Treatment with PGF2 alpha was effective, with 93% of cows having low progesterone at second GnRH. The overall percentage of cows that ovulated after second GnRH (synchronization rate) was 87% and varied by response to first GnRH (92% if ovulation to first GnRH vs 79% if no ovulation; P < 0.05). There were 6% of cows that ovulated before the second injection of GnRH and 7% with no detectable ovulation by 48 h after second GnRH. Maximal diameter of the ovulatory follicle varied by stage of estrous cycle, with cows in which Ovsynch was initiated at midcycle having the smallest follicles. In addition, milk production and serum progesterone concentration on the day of PGF2 alpha affected (P < 0.05) size of the ovulatory follicle. Using these results we analyzed pregnancy rate at Days 28 and 98 after AI for cows (n = 404) in which Ovsynch was initiated on known days of the estrous cycle. Pregnancy rate was lower for cows expected to ovulate larger follicles than those expected to ovulate smaller follicles (P < 0.05; 32 vs 42%). Thus, although overall synchronization rate with Ovsynch was above 85%, there were clear differences in response according to day of protocol initiation. Cows in which Ovsynch was initiated near midcycle had smaller ovulatory follicles and greater pregnancy rates.     

Euploid chicken embryos from eggs containing one, two or several yolks

A high incidence of haploid/diploid chimerism in chick embryos from strains of chickens selected for large size was postulated to be caused by the propensity of such hens to ovulate erratically. To test the hypothesis karyological analysis was made of embryos in eggs containing 1 or greater than 1 yolk. The eggs were from a line selected for multiple ovulation for 20 generations. Double and multiple-yolk eggs are a manifestation of an irregular ovulatory pattern. Ova in multiple yolk eggs were significantly less fertile and significantly fewer embryos survived to 18 h of incubation than single ovulated ova. In the sample of 342 embryos analysed, only 2 forms of heteroploidy occurred in frequencies of greater than 1.2%; 2n/4n mosaicism (5.8%) and 3n (5.0%). Only triploidy occurred significantly more frequently in eggs containing greater than 1 yolk (7.0%) than in single yolk eggs (none). The overwhelming majority of 3n embryos had a digynic origin (i.e from ova with 2 maternal pronuclei), as inferred from the sex chromosome complement. Erratic ovulation therefore resulted in suppression of second polar body extrusion leading to digynic triploidy. Multiple yolks had no effect on dispermy, the primary cause of 1n/2n chimaeric embryos, in single-yolked chicken eggs.     

Ovulation rate and serum LH levels in rats treated with indomethacin or prostaglandin E2

Serum LH levels were determined by radioimmunoassay at the normal time of the proestrous LH peak (17.30 – 18.00) and ovulatory performance was examined on the morning of estrus in rats treated with indomethacin, an inhibitor of prostaglandin synthesis. When the drug was administered at 14.30 on the day of proestrus, only 21% of the rats ovulated and the total number of ova shed was reduced to 4% of that found in the untreated control group, but there was no significant change in peak serum LH level (1122 ± 184 vs. 975 ± 240 ng/ml ± S.E., treated vs. control). Prostaglandin E₂ (PGE₂) given late on the day of proestrus (25 to 750 μ g/rat at 24.00) was effective in overcoming this antiovulatory action of indomethacin: 71–90% of the rats ovulated, though the number of eggs shed was low (24–55% of control value). Indomethacin was still effective in blocking ovulation when given at 20.00, that is after completion of the proestrous LH surge, but not at 24.00. Administration of PGE₂ (2 × 750 μ g/rat) in the early afternoon of proestrus elicited a rise in serum LH levels in rats in which the cyclic LH surge had been blocked with Nembutal (470 ± 87 vs. 106 ± 17 ng/ml ± S.E.) and induced ovulation in two-thirds of these animals.The results confirm, by direct measurement, that indomethacin does not block LH release but interferes with a late phase of the ovulatory process. PGE₂ reverses this action of indomethacin on the ovary. In addition, PGE₂ has a central effect causing LH release.     

The pattern of reproduction in female Columbian black-tailed deer, Odocoileus hemionus columbianus.

Ovarian cycles and the pattern of reproduction in female black-tailed deer in British Columbia were ascertained largely through examination of the ovaries from 444 females. Cyclic development and degeneration of single follicles of ovulatory size occurred several weeks before first ovulation. As the breeding season approached, a second or third large follicle developed in each cycle but in 48% of adult females the follicles were at different stages of maturation. Those failing to rupture at first ovulation luteinized 1 to 2 days thereafter. The first ovulation of the season, in November, never resulted in a lasting pregnancy even though some ova were penetrated by spermatozoa and began to cleave. First ovulation was apparently 'silent' in five of seven females for their ova lacked spermatozoa. Of sixty-one pregnant females, fifty-nine conceived at second ovulation; the other two conceived at subsequent ovulations more widely spaced than the 8- to 9-day interval between first and second ovulations. The synchrony of ovulatory cycles among adult females was such that half of them ovulated for the second time in a span of 7 or 8 days. Primary CL that formed after first ovulation grew to an average maximum volume of only about 45 mm3, whereas those originating at second ovulation grew to twice that size within 5 to 8 days. First generation CL shrank from 35 mm3 to 10 mm3 within 2 days. They disappeared within 18 months but corpora albicantia persisted for the life of the female. The possible ecological significance of the reproductive pattern is discussed.     

Familial influences on ovulatory function in common marmosets (Callithrix jacchus)

Previous studies have indicated that many, but not all, female common marmosets (Callithrix jacchus) housed with their natal families undergo social suppression of ovulation. In this study, we further characterized ovulatory activity in common marmoset daughters to determine the prevalence of social suppression of ovulation and to elucidate familial influences on daughters' ovarian activity. Blood samples were collected twice weekly from each of 46 daughters for 5–12 months, usually beginning when the daughters were 12 months of age. Plasma progesterone concentrations indicated that 46.3% of daughters in intact natal families ovulated at least once, with the age at first ovulation averaging 17.2 months; however, none of these daughters became pregnant. Daughters' ovulatory cycles showed several significant differences from those of older females housed with unrelated adults, including longer periods between successive luteal phases, shorter luteal phases, and lower peak and mean luteal-phase progesterone levels. Daughters were significantly more likely to ovulate in families in which the mother was experimentally prevented from sustaining pregnancies, and in families in which the father had been replaced by an unrelated adult male and when the daughter was approximately 10–11 months of age. Daughters in families containing an older sister never ovulated; in contrast, those with a female littermate were not less likely to ovulate than were other daughters, but had more sporadic ovarian cycles and significantly lower mean luteal-phase progesterone levels. These results confirm and extend previous findings that up to half of female common marmosets may ovulate while housed with the natal family but that virtually none sustain pregnancies, suggesting that suppression of ovulation is only one of several components of reproductive failure. Furthermore, these results demonstrate that daughters' likelihood of ovulating, as well as the endocrine profiles of their ovulatory cycles, can be modulated by numerous social influences within the family. Am. J. Primatol. 41:159–177, 1997. © 1997 Wiley-Liss, Inc.     

Effects of forced egg-retention in Aedes albopictus on adult survival and reproduction following application of DEET as an oviposition deterrent.

The insect repellent DEET (0.1% concentration), used as a mosquito oviposition deterrent in the laboratory, influenced the retention and maintenance of mature eggs by caged gravid female Aedes albopictus Skuse. This egg-retention mechanism could benefit survival because the gravid females were ultimately able to lay maintained eggs upon availability of water, but the length of forced egg-retention time reduced the number of eggs laid per female. Gravid females with retained eggs also laid a higher percentage of eggs that failed to tan, and this percentage increased with time duration of egg-retention. Percent egg hatch was not significantly affected by DEET when used as an oviposition deterrent; however, percent hatch was affected by time duration of egg-retention in both treated (exposed to DEET) and untreated (control) gravid females. The rate of egg hatch was considerably reduced after three weeks of retention; this reduction declined to zero for treated and control females at six and four weeks post-treatment, respectively. The fecundity and fertility of gravid female Ae. albopictus were affected by the time duration of forced egg-retention.     

Post ovulatory changes in the concentration of prostaglandins in rabbit Graafian follicles

In previous studies we have demonstrated that prior to hCG induced ovulation the levels of PGF and PGE in rabbit Graafian follicles increase markedly as ovulation approaches. We have now extended the study to include follicles obtained from animals at ovulation time and up to 48 hours after hCG injection. We have found that PGF reaches a maximum in ovulated follicles at the time of ovulation and then quickly decreases, whereas PGE continues to rise for several hours and then declines. The increase in both prostaglandins is limited to the follicles that actually ovulate. These data further document the proposed role for prostaglandins in the ovulatory process.     

Hysterectomy in the rat: surgical disruption of the vascular channels reduces number of ova shed.

Cycling rats were hysterectomized and/or unilaterally ovariectomized (ULO) on day 2 (metestrus). Collateral blood supply to the remaining ovary via the uterine artery was left intact or disrupted. Animals were killed in metestrus after one complete estrous cycle. Control rats were also killed at this time. Counts of tubal ova revealed that intact rats ovulated an average of 4.4 +/- 0.4 eggs per ovary (N = 8). Following ULO, rats (N = 8) ovulated 9.6 +/- 0.2 EGGS. Ligation of the uterine artery decreased the number of eggs ovulated in ULO rats (N = 8) to 5.4 +/- 1.1. Hysterectomized rats (N = 8) ovulated 4.8 +/- 0.5 eggs per ovary. If the blood supply was disrupted, a reduction to 2.7 +/- 0.2 eggs per ovary occurred (N = 8). Hysterectomized and ULO rats (N = 8) ovulated 10.3 +/- 0.4 eggs from the remaining ovary but only 5.0 +/- 1.0 eggs if the collateral blood supply of the uterine artery was not intact (N = 10). The results demonstrate that disruption of the vascular channels during the surgical procedures of hysterectomy and/or ULO results in a reduction of the expected ovulation number.     

Is slow follicular growth the cause of silent estrus in water buffaloes?

The present experiment was conducted to study the growth profile of the ovulatory follicle in relation to the expression of estrus following administration of PGF(2alpha) to subestrus buffaloes. After detection of a mature corpus luteum by examination per rectum, confirmed by ultrasound scanning, subestrus buffaloes (n=20) were treated (Day 0) with single dose of Dinoprost tromethamin (25 mg, i.m.). Blood samples were collected at 0, 24 and 48 h after treatment for estimation of plasma progesterone concentration. Growth profile of the ovulatory follicle was monitored daily through ultrasound scanning starting from Day 0 until ovulation and the regression profile of CL was monitored at 0, 24 and 48 h of treatment. Estrus was detected by exposure to a fertile buffalo bull three times a day until expression of overt estrus or ovulation. Behavioral estrus was recorded in 14 animals and 6 animals ovulated silently. Sixteen animals including six animals with silent estrus ovulated from the dominant follicle present at treatment (Group A) and remaining four animals ovulated from the dominant follicle of succeeding follicular wave (Group B). The intervals from treatment to estrus (6.5+/-0.25 versus 3.2+/-0.27 days, P<0.001) and treatment to ovulation (7.5+/-0.25 versus 5.4+/-0.46 days, P<0.005) were significantly longer in animals of Group B compared with animals of Group A. Significant differences were observed in growth profile of the ovulatory follicle between animals of Groups A and B with respect to size of the follicle on Day 0 (9.8+/-0.7 versus 5.3+/-0.45 mm, P<0.001), daily growth rate (0.97+/-0.07 versus 1.6+/-0.2 mm/day, P<0.01) and increase in diameter (4.1+/-0.6 versus 7.8+/-0.7 mm, P<0.01). The animals with silent estrus (subgroup A-2) had significantly smaller diameter of the ovulatory follicle on Day 0 (7.7+/-0.4 versus 11.0+/-0.7 mm, P<0.005), its daily growth rate was significantly slower (0.7+/-0.02 versus 1.1+/-0.1 mm/day, P<0.01) and they recorded significantly longer interval from treatment to ovulation (7.3+/-0.56 versus 4.2+/-0.27 days, P<0.001) compared with the animals that showed overt estrus (subgroup A-1). The corpus luteum area (CL area) and plasma progesterone (P(4)) concentration declined continuously from 0 to 48 h after PGF(2alpha) treatment in the animals of both the Groups A and B. Non-significant differences were observed in mean CL area and plasma P(4) concentration at 0, 24 and 48 h post-treatment between animals of Groups A and B and also between animals of subgroups A-1 and A-2. The small size and the slow growth rate of the ovulatory follicle were identified as the possible cause of silent estrus in subestrus buffaloes after PGF(2alpha) treatment.     

Absorbable deslorelin implants (Ovuplant) prolong postpartum anestrus in early ovulating dairy cows

Two experiments were conducted to investigate the use of a bioabsorbable implant of the GnRH agonist deslorelin to temporarily delay the resumption of postpartum ovulatory cycles in Holstein cows. In Experiment 1, recently calved cows were paired and received either a single implant (Ovuplant); Peptech Animal Health, Sydney, NSW, Australia) within 48 h of parturition (OVP; n=17), or remained as untreated controls (CON; n=17). Blood samples were collected for plasma progesterone assay three times weekly for 6 weeks to profile the pattern of resumption of ovulatory cycles. In Experiment 2, there were 15 CON and 15 OVP cows initially treated as for Experiment 1 as well as 15 OVP+SYNCH cows. Each cow in the CON and OVP+SYNCH groups received a progesterone vaginal insert (CIDR); Genetics Australia, Bacchus Marsh, Vic., Australia) for 7 days at 23 days postpartum (23 dpp) to synchronise estrus in cycling animals or to induce an ovulation with estrus in anestrus animals. Blood samples were collected weekly until removal of the CIDR insert, and then twice weekly until 56 dpp to monitor plasma P4 for retrospective determination of ovulation. Milk yield was monitored by twice daily electronic volume measurements and milk composition with once weekly milk composition analysis.In Experiment 1, CON cows began ovulating from 9 dpp; 15 of 17 had ovulated by the end of blood sampling at 42 dpp. None of the OVP cows ovulated until at least 24 dpp, and only 6 of 17 had ovulated by 42 dpp. The average day of first ovulation was extended from 22.4+/-2.7 dpp to 39.3+/-2.7 dpp (P<0.05). In Experiment 2, ovulation had occurred in 8 of 15 CON cows at the time of CIDR insertion (23 dpp), 0 of 15 OVP cows and 1 of 15 OVP+SYNCH cows. By 40 dpp (or 10 days following removal of the CIDR insert) every CON cow (15/15) had ovulated, but only 2 of 15 OVP+SYNCH cows and 1 of 15 OVP cows. None of these effects of treatment was associated with any changes in milk yield or composition in either experiment.In conclusion, inserting a bioabsorbable implant of deslorelin within 48 postpartum extended the interval to first ovulation to at least 24 dpp in 46 of 47 cows. Recovery periods were highly variable. This variability was not reduced by using a form of intravaginal progesterone supplementation that did produce a synchronised estrus with ovulation in anestrus animals that had not been treated with deslorelin.     

The tempo of reproduction in Hyphessobrycon pulchripinnis(Characidae), with a discussion on the biology of ‘multiple spawning’ in fishes

Synopsis We describe the short-term patterns of egg production and release in the lemon tetra, Hyphessobrycon pulchripinnis (Characidae) as observed over a six month aquarium study and then use our results and those of others to both describe general patterns and derive comparative predictions. Female lemon tetras ovulated about once every four days; differences among individuals were small and inconsistent. As in other species, the probability of ovulation depended strongly on time since last ovulation, indicative of an ovarian cycle; furthermore we found no obvious long-term patterns of ovulation within females and batch fecundity was independent of the length of the two previous interovulation intervals and of the one following. Each batch of ovulated eggs was released over an average of 23.1 spawning acts, beginning as soon as the lights went on in the morning. The number of spawning acts increased with ovulation fecundity but decreased with the number of other females spawning in the tank on that day. The mass of eggs produced every four days was greater than that of the remaining ovary, and the mass of eggs that could be produced in a season was greater than that of the female. This latter observation highlights the most important consequence of repeated reproduction within a season - increased reproductive output - and leads to predictions associating it with less seasonal environments (e.g. low latitudes), as well as with small ovaries and small body size. The significance of releasing a batch of eggs over many spawning acts remains unknown.     

The effect of platelet activating factor on ovulation.

The mechanism of ovulation has been compared to an inflammatory reaction. Platelet activating factor (PAF) is an important mediator of inflammation as it may induce the production of prostaglandins and lysosomal enzyme. We evaluated the potential role of PAF in PMSG-HCG induced ovulation using CV3988, a specific PAF receptor antagonist in a superovulated ICR mice (9-12 weeks old). CV3988 blocked the ovulation in a dose dependent manner, and the significant reduced ovulatory efficiency was observed at more than 500 micrograms dose (p less than 0.001). The ovulatory efficiency reduced by CV3988 was reversed by PAF in a dose dependent manner. In vitro fertilization (IVF) rate of follicular oocytes with treatment of CV3988 was not different from that of ovulated ova without treatment. These results suggest that PAF may be involved in the ovulation process but the presence of PAF may not be essential for the fertilization of the ova as IVF.     

Ovarian follicular development in Holstein cows following synchronisation of oestrus with oestradiol benzoate and an intravaginal progesterone releasing insert for 5-9 days and duration of the oestrous cycle and concentrations of progesterone following ovulation

The aim of this study was to determine if the duration of treatment with an intravaginal progesterone releasing insert (IVP(4)) after treatment with oestradiol benzoate (ODB) at the time of insertion and 24 h after removal would affect selected variables including: size of ovarian follicles at the time of removal of inserts, diameter of ovulatory follicles, plasma concentrations of progesterone following ovulation, and duration of the following oestrous cycle. Characteristics of oestrus at a synchronised and spontaneous oestrus were also monitored. Non-lactating Holstein cows were synchronised with an IVP(4) for 5 (n = 10), 7 (n = 10), 8 (n = 9) or 9 (n = 9) days together with injections of ODB at device insertion (2 mg) and 24 h after removal (1 mg). Ultrasonography showed no significant effect of treatment on the day of emergence of preovulatory follicles relative to the day of removal of inserts (overall mean = -4.22 +/- 0.58; P = 0.15) for cows that ovulated within 120 h insert removal (n = 36). Treatment with ODB and an IVP(4) for 5 days reduced the diameter of preovulatory follicles at the time of removal of inserts and for the following 2 days compared to cows treated for 7-9 days (mean difference 2.56 +/- 1.15 mm; P = 0.033) but did not reduce the diameter of the ovulatory follicle (P = 0.21). Day of emergence relative to removal of inserts was associated with the diameter of the ovulatory follicle (R2 = 0.69; P < 0.001). Concentrations of progesterone and the diameter of the corpus luteum following ovulation were not affected by treatment (P > 0.20), but were affected by the diameter of the ovulatory follicle (P < 0.01). Diameter of the ovulatory follicle did not affect interoestrous and interovulatory intervals (P > 0.40). We conclude that treatment with an IVP(4) for 5 compared to 7-9 days with ODB administered at device insertion, and 24 h after removal reduced the diameter of preovulatory follicles at the time of removal of the insert but did not reduce the diameter of the ovulatory follicle or concentrations of progesterone in plasma. Emergence of preovulatory follicles closer to the time of removal of inserts reduced the diameter of the ovulatory follicle when oestrus was induced with ODB. Ovulation of smaller follicles reduced concentrations of progesterone in plasma following ovulation but did not affect oestrous cycle duration.