Pairing ultrasonography with endocrinology to elucidate underlying mechanisms of successful pregnancy in the northern fur seal (Callorhinus ursinus)

General and Comparative Endocrinology 255 (2018) 78–89

Contents lists available at ScienceDirect

General and Comparative Endocrinology journal homepage: www.elsevier.com/locate/ygcen

Research paper

Pairing ultrasonography with endocrinology to elucidate underlying mechanisms of successful pregnancy in the northern fur seal (Callorhinus ursinus) Michelle R. Shero a, Don R. Bergfelt b, J. Ward Testa c, Gregg P. Adams d,⇑ a

Department of Biological Sciences, University of Alaska Anchorage, 3101 Science Circle, Anchorage, AK 99508-4614, USA Department of Biomedical Sciences, Ross University School of Veterinary Medicine, Basseterre 00265, Saint Kitts and Nevis National Marine Mammal Laboratory, Alaska Fisheries Science Center, 7600 Sand Point Way NE, Seattle, WA 98115-6349, USA d Western College of Veterinary Medicine, University of Saskatchewan, 52 Campus Drive, Saskatoon, SK S7N 5B4, Canada b c

a r t i c l e

i n f o

Article history: Received 26 July 2017 Revised 11 October 2017 Accepted 16 October 2017 Available online 16 October 2017 Keywords: Diapause Embryo Pinniped Pregnancy Reproduction Ultrasonography

a b s t r a c t Reproductive success is one of the central tenets of conservation management programs, yet the inability to study underlying physiological processes in a minimally-invasive manner and the unpredictable nature of wild animal populations leaves large gaps in our knowledge of factors critical to successful reproduction in wild species. This study integrated ultrasonography of the reproductive tract and analysis of reproductive hormones in 172 northern fur seals (Callorhinus ursinus) to identify intrinsic factors associated with reinitiating embryonic growth at the end of diapause. Within the first 3–4 weeks of active gestation, pregnant fur seals (n = 126) had a larger corpus luteum and fewer antral follicles than nonpregnant fur seals, or those still in diapause (n = 46). This suggests that the conceptus drives changes in ovarian status to convey its presence to the female. Morphological changes in the reproductive tract associated with pregnancy were not reflected in differences in endocrine profiles (estradiol, estrone, progesterone, and relaxin) between pregnant and non-pregnant individuals. Hormone concentrations correlated more strongly with calendar date than with the presence or size of the conceptus, demonstrating that none of these reproductive hormones were reliable markers for early pregnancy diagnosis. Instead, the northern fur seal’s long diestrus may serve to reduce the probability of a temporal mismatch between corpus luteum regression and embryo implantation. Indeed, conception rates were high and confirmed rates of pregnancy loss were relatively low (11%). In this study, minimally-invasive ultrasonography was used in wild pinnipeds to detect very early pregnancy (embryonic vesicles >2 mm) in combination with ovarian and endocrine dynamics at the time of embryo implantation, shedding light on mechanisms for maternal recognition of pregnancy. This study is also the first to track whether these same animals carried the embryo to term, by observing fur seals during the birthing season the following year. Data do not support the notion that decreased pregnancy rates or higher pregnancy loss rates are major contributing factors to the northern fur seal’s population decline. Ó 2017 Elsevier Inc. All rights reserved.

1. Introduction Maintenance of a stable population is contingent on the ability of animals to survive and reproduce. Mechanisms governing reproduction vary widely across and within taxa, but logistical, technical, and ethical challenges often prohibit the study of intrinsic Abbreviations: CA, corpus albicans; CL, corpus luteum; EV, embryonic vesicle.

⇑ Corresponding author.

E-mail addresses: [email protected] (M.R. Shero), [email protected] (D.R. Bergfelt), [email protected] (J.W. Testa), [email protected] (G.P. Adams). https://doi.org/10.1016/j.ygcen.2017.10.007 0016-6480/Ó 2017 Elsevier Inc. All rights reserved.

physiologic factors that influence reproductive patterns in nonmodel or wild animal species (Comizzoli et al., 2009; Lopes et al., 2004; Wildt et al., 2010). In assessing reproductive rates, common practices once included sacrificing individuals and their pregnancies for dissection and collection of the reproductive tracts (Craig, 1964; Daniel, 1971). Endocrine profiles have also been used as markers for pregnancy (Boyd, 1991; Browne et al., 2006; Daniel, 1975; Gales et al., 1997; Reijnders, 1990); however, interpretation of hormone changes during the post-breeding period is confounded by the lack of knowledge about normal temporal changes in non-pregnant females (e.g., not bred), and those that are in early

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89

stages of gestation (Browne et al., 2006; Daniel, 1974, 1975). The development of reliable, minimally-invasive technologies for assessing reproductive phenology are particularly useful when sacrificing individuals is not a viable option, whether it be due to logistical constraints or to conservation measures set to protect species in declining populations (Testa et al., 2010). Following conception, a pivotal step for reproductive success is maternal recognition of pregnancy (Roberts et al., 1996) where the embryo signals its presence in the uterus by either inhibiting luteolysis or by having a luteotrophic effect (Bazer, 2013; Bazer et al., 2010; Clemente et al., 2009; Geisert and Bazer, 2015). Without such a signal, the corpus luteum (CL) will begin to regress, marking the point at which pregnant and non-pregnant females become physiologically distinct (Roberts et al., 1996). Inadequate or mistimed maternal recognition of pregnancy results in the loss of the conceptus prior to placentation, and is a major contributor to female infertility in humans and other large mammals (Cross et al., 1994; Macklon et al., 2002; Teklenburg et al., 2010). While the mechanisms critical to maternal recognition of pregnancy are species-specific, well-studied laboratory mammals and agricultural livestock have revealed common endocrine changes associated with initiation of pregnancy. For example, during estrus there is a pronounced rise in estrogen production by preovulatory follicles, and this induces proliferative changes in the uterine endometrium via increased transcription of cell cycle genes, mitotic rates, and edema (Geisert and Bazer, 2015). Progesterone increases dramatically soon after ovulation and is responsible for differentiation of the endometrium to a secretory phase, and breaks down the transmembrane glycoproteins that act as a physical barrier to embryo attachment in the uterus (Geisert and Bazer, 2015). As gestation progresses and placentation occurs, relaxin concentrations increase and play a role in maintaining the differentiated endometrium, increasing angiogenesis and blood flow, and reducing rates of pregnancy loss (Anand-Ivell and Ivell, 2014). The occurrence of embryonic diapause adds complexity to the otherwise well-characterized sequence of endometrial proliferation/secretion, embryo attachment (i.e., implantation), and maternal recognition of pregnancy in other species (Renfree and Shaw, 2000, 2014). In species with diapause, the fertilized zygote develops only until reaching the blastocyst stage, before entering a period of metabolic quiescence and dramatically slowed mitotic growth (Daniel, 1971; Fenelon et al., 2014). In pinnipeds, the CL becomes vacuolated and progesterone secretion wanes during embryonic diapause (Atkinson, 1997; Boshier, 1981; Craig, 1964; Daniel, 1975; Harrison, 1948; Reijnders, 1990), but this process is reversible and luteinization occurs around the time of embryo attachment and the start of active gestation (Boshier, 1981; Craig, 1964). The mechanisms signaling the end of the blastocyst’s period of suspended animation remain largely unknown. Pinnipeds offer a unique opportunity to study the physiologic shifts that occur at the time of embryo attachment in wild animals due to the predictability with which animals congregate at on-shore rookeries. Because of commercial harvests for pelts prior to the Fur Seal Act (years: 1911, 1944, 1966, 1983) and the North Pacific Fur Seal Convention (1957), northern fur seals (Callorhinus ursinus) have been studied extensively to monitor their behavior, movements, and reproductive record. Northern fur seals give birth in July–August each year (mean date: July 9; York and Scheffer, 1997; Trites and York, 1993), with
70% of the world’s population residing on the Pribilof Islands off the coast of Alaska. In a highly polygynous breeding system, the males develop a hierarchical breeding territory to build harems as females return to terrestrial rookeries for birthing (Bartholomew and Hoel, 1953). Within one week of parturition, increased ovarian follicular development leads to ovulation, and

79

females are bred (Bartholomew and Hoel, 1953; Craig, 1964). After fertilization, the embryo enters an obligate (seasonal) period of diapause thought to be controlled by photoperiodic cues (Fenelon et al., 2014; York and Scheffer, 1997). Simultaneously, female fur seals nurse their pups for the next four months, utilizing an income-breeding strategy by alternating between foraging for self-maintenance and returning to the rookery to provision the pup (Bartholomew and Hoel, 1953; Boltnev and York, 2001). Embryonic diapause in this species is thought to keep the demands of lactation separate from the demands of active gestation; that is, re-initiation of embryonic growth and attachment does not occur until mid-November, after the pup is weaned (Gentry, 1998; York and Scheffer, 1997). In humans and domestic species in which it has been critically examined, the incidence of pregnancy loss is greatest during the embryonic phase, declining to a small fraction after the start of the fetal stage (i.e., after implantation and organogenesis). Further, age and nutritional plane may exert control over the probability of conception (Boyd, 1984; Testa and Adams, 1998). Therefore, the end of diapause in northern fur seals coincides with the time of the year that females are in relatively poor condition, and may serve as a critical period that distinguishes which females will continue with gestation and which will undergo embryonic loss. Northern fur seals are currently listed as a vulnerable species under the International Union for Conservation of Nature (IUCN; Gelatt et al., 2015) and non-lethal methods are needed to assess reproductive status in an effort to determine the cause of population stasis. In this regard, the use of minimally-invasive ultrasonography has been developed as a means of early pregnancy detection in pinnipeds (Adams et al., 2007; Shero et al., 2015; Testa et al., 2010). To date, the temporal dynamics between hormone concentrations and early embryo development in marine mammals have not been critically investigated, nor have events during the peri-implantation period been paired with direct observations of pregnancy status or gestational outcomes (i.e., embryo loss vs. pup birth). Direct and immediate knowledge of pregnancy status by ultrasonography provides the opportunity to relate endocrine dynamics with morphologic characteristics of the ovaries and uterus, and establish physiologic linkages with reproductive events such as non-pregnancy, embryonic growth and viability, implantation, and gestational loss in the same individuals. The goals of this study were to characterize early pregnancy in northern fur seals, and assess rates of reproductive success (i.e., giving birth) and gestational loss. In an effort to elucidate the mechanisms involved at the end of diapause in northern fur seals, we examined the morphologic and hormonal dynamics of the reproductive system near the expected time of embryo attachment. Results are expected to provide a basis for the establishment of indices that may be used to identify pregnancy during the earliest stages of active gestation for use in captive breeding programs and wildlife conservation management.

2. Methods 2.1. Animal handling Female northern fur seals were captured and physically restrained (Gentry and Holt, 1982) on St. Paul Island, Alaska on 13–14 November 2005 (n = 10), 11–16 November 2007 (n = 96), and 18–24 November 2008 (n = 66) between 10:00 and 18:00 h. All were captured and weighed at the Polovina Cliffs rookery (5 7°100 4600 N, 170°90 4200 W) except for 26 captured at Tolstoi rookery (57°080 1100 N, 170°200 5000 W) in 2008. Twenty-six seals in 2008 were anesthetized with isoflurane for tooth extraction. All others were examined without sedation while restrained in sternal

80

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89

recumbency using a head stock and board, with the pectoral limbs held tightly against the body with a neoprene vest, and the pelvic limbs held caudally. Blood was collected from 133 animals, from the hind-flipper plexus, kept chilled until transported to the laboratory (within 6 h), and serum was collected and stored at 20 °C until hormonal analysis. A lower and upper limit of each animal’s age was estimated based on whisker and pelage coloration (Baba et al., 1991). 2.2. Detection of pregnancy & ovarian structures The reproductive tract was examined by transrectal ultrasonography using a portable ultrasound unit (Aloka Co Ltd., Aloka SSD 500; Tokyo, Japan) equipped with a 7.5 MHz linear-array, sidefire transducer designed for transrectal prostate examination in humans (Adams et al., 2007), and a 12 V/15A external battery with an AC-DC power inverter. The tail was elevated, and the ultrasound probe was inserted into the rectum with the aid of standard methylcellulose gel lubrication. The probe was advanced with the lens facing ventrally and the entire reproductive tract was systematically scanned beginning with the cervix and uterine body in sagittal section immediately dorsal to the urinary bladder, up one uterine horn to the ipsilateral ovary, down the same uterine horn and back to the uterine body, and up the opposite uterine horn and ovary (Adams et al., 2007; Shero et al., 2015). Pregnancy was diagnosed by detection of a clearly circumscribed spherical, fluidfilled chamber (embryonic vesicle; EV), within the lumen of the uterus. An embryo proper, with or without a detectable heartbeat, was visible in some of the larger EVs; in combination these will be referred to as the embryo throughout this study. The diameter of the EV was measured in maximum cross section, with the average of two perpendicular measures used if the structure was not circular. Ovarian follicles and CL were identified based on similarity to other mammalian species (Adams et al., 2007). Follicles and CL were enumerated and measured at their maximum diameter. Images were taped using a digital video camera (Sony Network Handycam DCR-TRV950) for later review as necessary. 2.3. Hormone assays Serum hormone concentrations of estradiol, estrone, progesterone, and relaxin, were measured by radio-immunoassay, and validated for use in northern fur seals with parallelism and standard addition tests. Serum estradiol and estrone were measured from dried acetone extracts of 0.2 mL serum using commercial kits (Diagnostic Systems Laboratories, Webster, TX, USA; Estradiol – DSL-4400; assay sensitivity: 1.0 pg mL1; intra-assay CV: 5%; inter-assay CV: 10%; Estrone – DSL-8700; assay sensitivity: 1.2 pg mL1; intra-assay CV: 6%; inter-assay CV: 12%). Progesterone was measured after ether-extraction (DSL-3400; assay sensitivity: 0.05 ng mL1; intra-assay CV: 5%; inter-assay CV: 8%). Relaxin concentrations were determined with a canine assay previously adapted and validated for use in northern fur seals, using rabbit polyclonal canine relaxin antiserum as the primary antibody (Bergfelt et al., 2010; Steinetz et al., 1996). Assay sensitivity was 0.6 ng mL1 with intra- and inter-assay CVs of 5% and 15%, respectively. All samples were analyzed in duplicate. 2.4. Reproductive outcomes As part of the National Oceanic and Atmospheric Administration National Marine Mammal Laboratory tagging program, an Allflex (Dallas, TX, USA) and/or a VHF tag was applied to the caudal edge of the fore-flipper at the time of handling. This facilitated re-sighting animals the following year to determine reproductive outcomes (i.e., birth vs. no birth). Re-sighting efforts were done

only at Polovina Cliffs from July 3 to August 25 in all study years. As northern fur seals give birth within 1–3 days of arriving at the rookeries (Gentry, 1998), the observed arrival date was considered the pup’s date of birth. 2.5. Statistical analyses An embryo growth curve constructed using Bayesian models from Testa et al. (2010) was used to extrapolate backwards to the date when the embryo was 1.5 mm in diameter, the approximate size of blastocysts when growth resumes after diapause (Craig, 1964). This is referred to hereafter as the estimated date of embryo reactivation in analyses. The interval between the estimated date of embryo reactivation and parturition date the following year was used to calculate the duration of active gestation. The detection of pregnancy and anatomical features of the embryo were examined in relation to female body mass and age using binomial generalized linear models (GLM), and date was always included as a covariate. Age, body mass, and embryo size at the time of examination were related with birthing dates the following year (t + 1) using linear regressions. To assess differences in the number and size of follicles between ovaries (CL-bearing ovary, versus ovary without a CL), zero-inflated generalized linear mixed models for count data (Poisson distribution) from the ‘‘glmmADMB” package in R were used. Date, pregnancy status, and their interactive effects were included as covariates, with animal ID as a random effect in models (Skaug et al., 2012). Due to collinearity between calendar date and EV size, separate GLM analyses were used to assess hormone concentrations relative to: (1) calendar date, pregnancy status (EV detectable/not detectable), and interaction terms, and (2) size of the conceptus. When two CL were detected, only the largest and presumably functional CL was correlated with date, embryo size, and hormones. Female age and the time of day that blood was collected were added to the models to assess whether they improved fit. Best-fit models were selected using Akaike Information Criterion corrected for small sample sizes (additional variable was retained when DAICc > 2), and were only accepted with the absence of over-dispersion and heteroscedasticity of residuals (Zuur et al., 2013). Outliers were identified using Cook’s distance plots generated from model fit, and removed. The residuals from these models were then used to assess whether fur seals with higher hormone concentrations for the respective date of sampling, had larger embryos or were more likely to be seen with a pup the following year, using linear regressions and student’s t-tests (or Kruskal-Wallis H tests for nonnormal data), respectively. Results are presented as the mean ± S E, and all analyses were conducted in R (v. 3.2.1; R Core Team, 2016) with the significance level set as a = 0.05. 3. Results 3.1. Pregnancy detection & characterization The reproductive outcomes of fur seals in which an embryo was detected versus those with no detectable pregnancy are summarized in Table 1. An EV was detected in 73.3% of the fur seals examined in this study (126 of 172). Of those that were not detected pregnant at the time of transrectal ultrasound examination and observed the subsequent year, 87.5% returned with a pup, demonstrating that the embryo of these seals was still in diapause or so recently after diapause that it was too small to be detected by ultrasonography (<2–3 mm). Of fur seals that were detected pregnant at the time of ultrasound examination and were also resighted the following year, the estimated active gestation length

81

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89

Table 1 Reproductive outcome of northern fur seals which had a detectable versus no apparent embryonic vesicle at the time of ultrasound examination in November 2005–2008. Since many of the seals were re-sighted the following birthing season (t + 1), their pupping status was known. Mean ± SE (sample sizes in parentheses). Embryonic vesicle detected

n Mass (kg) Age (Lower Estimate; Years) Age (Upper Estimate; Years)

No detectable embryonic vesicle

Year t + 1 Pup

No pup

Not re-sighted

Pup

No pup

Not re-sighted

65 33.2 ± 0.6 (64) 7.5 ± 0.3 (63) 10.3 ± 0.7 (63)

8 31.4 ± 1.6 (8) 7.3 ± 1.2 (6) 9.5 ± 2.2 (6)

53 34.2 ± 0.8 (53) 7.7 ± 0.4 (45) 11.6 ± 0.9 (45)

28 32.7 ± 1.0 (28) 7.4 ± 0.5 (28) 8.8 ± 0.7 (28)

4 28.6 ± 1.7 (4) 5.0 ± 1.4 (4) 6.0 ± 1.6 (4)

14 28.2 ± 1.9 (14) 5.1 ± 0.7 (13) 6.5 ± 1.2 (13)

was 253.3 ± 0.7 days (range: 238–264 days). Only 11% of fur seals that were detected pregnant at the time of ultrasound examination and observed the next year, returned to the rookeries without a pup and were presumed to have lost their pregnancies. Fur seals that experienced pregnancy loss did not differ in mass or age (upper or lower estimate) from those that carried an embryo to term (Table 1). The likelihood of detecting an EV (Binomial GLM: v2 = 5.7, P = 0.017) followed by an embryo proper (v2 = 27.7, P < 0.001) and heartbeat (v2 = 12.5, P < 0.001; Fig. 1; Fig. 2A) increased with calendar date. Adding animal mass to the probability of detecting an EV across the period of ultrasound examinations (Nov. 11–24) improved model fit (v2 = 7.7, P = 0.008). However, this was likely an artefact of the strong relationship between animal mass and age (lower age estimate: F1,155 = 170.7, P < 0.001; upper age estimate: F1,155 = 158.7, P < 0.001); fur seals that were relatively large for their age (i.e., mass
age residuals) were not more likely to be detected pregnant during the sampling period. Older fur seals were more likely to be detected pregnant (Fig. 2B; lower age estimate: v2 = 7.0, P = 0.011; upper age estimate: v2 = 8.9, P = 0.008). Corre-

spondingly, older fur seals had earlier estimated dates of embryo reactivation (i.e., EV size
date residuals; Fig. 2C, lower age estimate: F1,111 = 7.7, P = 0.007; upper age estimate: F1,111 = 5.9, P = 0.016). Fur seals with a larger EV (linear regression: F1,68 = 7.6, P = 0.008), and those with earlier estimated dates of embryo reactivation (Fig. 3; F1,68 = 17.7, P < 0.001) gave birth earlier during the following year’s pupping season. The linear slope relating embryo reactivation date to parturition date the following year was significantly greater than 1 (0.226 ± 0.11 95% CI) showing that on average, for each day later that embryo reactivation occurred, active gestation was
4.4 days longer. Fur seals that were likely to have been in a period of embryonic diapause at the time of examination (i.e., were not detected pregnant by ultrasound, but were seen with a pup the following year), did not give birth later than fur seals that were detected pregnant at the time of transrectal ultrasound examination (t43.7 = 1.41, P = 0.167). Similarly, older fur seals gave birth slightly, but not significantly, earlier than younger seals (lower age estimate: F1,88 = 3.7, P = 0.059; upper age estimate: F1,88 = 2.0, P = 0.162). Older animals had slightly

Fig. 1. (A) Ultrasound images showing development of fur seal embryos from a small embryonic vesicle (‘‘EV”) to those with an embryo proper (‘‘EP”) and heart beat (‘‘Hrt”). Scale bars are in 1 cm increments; cranial is to the right and ventral is downward. (B) Modified embryonic growth curve fit with a Bayesian model (from Testa et al., 2010) showing backward extrapolations (dashed arrows) to estimated date of embryo reactivation when the EV would be 1.5 mm in diameter (points are jittered); shaded regions show when the embryo proper first becomes detectable, followed by a heartbeat. (C) Density distribution of estimated dates of embryo reactivation; dotted line indicates median date.

82

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89

Fig. 3. Linear regression showing fur seals that had an earlier estimated date of embryo reactivation (greater EV size
date residuals) gave birth at an earlier date the next year. Detected pregnant fur seals (closed circles) did not give birth earlier than those that were not detectably pregnant during ultrasound examinations (open circles). Points are jittered. Sample size: EV detected, n = 65; no EV detected, n = 28.

Fig. 2. (A) The probability of detecting an EV, EP, and heart beat increased with calendar date, and (B) the probability of detecting an EV increased with female age, shown using binomial GLM. (C) Of fur seals that were detected pregnant, older animals had earlier estimated dates of embryo reactivation (i.e., greater EV size
Date residuals; open circles = lower, closed = upper age estimate. Adding a quadratic term did not improve model fit).

(lower age estimate: F1,61 = 3.0, P = 0.090; upper age estimate: F1,61 = 3.4, P = 0.071), and larger animals had significantly (F1,62 = 4.2, P = 0.044), shorter estimated active gestation lengths.

the uterine horn containing the conceptus in 84.9% of females in which both the side of the EV and CL could be confirmed (n = 119), significantly more frequently than if by chance (Fisher’s exact test: P < 0.001). A CL was not detected in either ovary in 9.0%, and 2 CL were detected in 7.8% of fur seals. While CL size did not vary with calendar date, the CL was larger in fur seals that were detected pregnant at the time of ultrasound examination, than in those that were not (Fig. 4; Gaussian GLM; F1,150 = 9.8, P = 0.002). Further, fur seals with a larger CL had a slightly earlier estimated date of embryo reactivation (F1,118 = 3.5, P = 0.065) and significantly shorter estimated active gestation lengths (F1,60 = 5.9, P = 0.018). Nine fur seals (5 pregnant; 4 non-pregnant) had an apparent corpus albicans (CA) during ultrasound examination. In 5 of them, the apparent CA was in the ovary opposite to that bearing the CL and in 2 it was in the same ovary; 2 fur seals had no detectable CL but an apparent CA. An ovary-by-day interaction in the number of small (2–5 mm) follicles detected was attributed to more follicles in the ovary without the CL, and numbers declining with calendar date (Fig. 5A). An ovary-by-pregnancy status interaction in the number of large (>5 mm) follicles was a result of pregnant animals having significantly fewer large follicles in the ovary without a CL (Fig. 5B). Therefore, the mean size of the largest follicle in the CL-bearing ovary was larger than in the non CL-bearing ovary (Fig. 5C). Adding date or pregnancy status did not improve model fit. In combination, calendar date, pregnancy status, and ovary all influenced the total number of follicles observed (Fig. 5D). The total number of follicles was lower in the CL-bearing ovary than in the non-bearing ovary. With respect to the ovary without the CL, the total number of follicles was lower in fur seals that were detected pregnant than in those not detected pregnant; there was a negative relationship with calendar date only for pregnant fur seals. Further, the majority of fur seals with a CL contralateral to the embryo (9 of 11) or that had two CL (8 of 13) also had more large follicles in the CL-bearing ovary, or the ovary with the largest CL, respectively.

3.2. Ovarian structures

3.3. Hormone concentrations & links with ovarian morphology and pregnancy status

The majority of fur seals (83.1%) had a single CL at the time of ultrasound examination. The ovary with the CL was ipsilateral to

Serum estradiol and estrone concentrations decreased, while progesterone and relaxin concentrations increased with calendar

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89

83

Fig. 4. (A) Ultrasound images of a pair of ovaries from a northern fur seal showing antral follicles (‘‘F”) of varying diameter (2–5 mm) in both the ovary without (left) and the ovary with the corpus luteum (‘‘CL”; right). Scale bars are in 1 cm increments. (B) The CL was larger in seals that were detectably pregnant at the time of ultrasound examination (P < 0.01).

date (Gamma GLM; Fig. 6). The number of 2–5 mm follicles and the total number of follicles in both ovaries accounted for a significant amount of variation in estradiol and estrone concentrations, respectively (Table 2). Further, adding CL size to the model relating progesterone to calendar date improved model fit. Fur seals with higher relaxin concentrations than would be expected relative to handling date, also had a larger maximum follicle and earlier estimated dates of embryo reactivation (F1,87 = 5.7, P = 0.019), but CL size did not improve model fit. None of the reproductive hormones measured in this study exhibited significant differences by pregnancy status, although fur seals that were detected pregnant tended to have higher serum estradiol and estrone concentrations (both P < 0.1). Similar patterns were present between serum hormone concentrations and EV diameter. Estrone concentrations were negatively correlated with EV size, while progesterone and relaxin levels exhibited a positive relationship with EV diameter. However, there was no significant relationship between EV size and estradiol (Fig. 7), and adding CL and follicle morphology did not improve the models. Hormone concentrations exhibited a stronger relationship with calendar date than with size of the embryo (see Table S1), and adding age, hour of sampling, or interactive effects did not improve model fit. Fur seals that were observed with a pup the following year had slightly, but not significantly, higher relaxin concentrations relative to date of sampling than fur seals that failed to give birth (Relaxin
Date residuals: t18.1 = 1.8, P = 0.091). These trends were not present with serum estrogen or progesterone concentrations. 4. Discussion The stability of the northern fur seal population, as in any wildlife population, is a function of reproductive success and offspring survival to reproductive age. Ultrasonography as a minimallyinvasive and non-lethal technique, provided a previously inaccessible opportunity to examine the reproductive status of northern fur seals in the wild at a time coinciding with the end of embryonic diapause, and to relate the findings to reproductive outcomes the following year. The use of ultrasonography, instead of dissection of embryos from sacrificed fur seals (Craig, 1964; Daniel, 1974, 1980, 1971, 1975), enabled direct longitudinal assessment of the major components of fecundity; i.e., pregnancy rate, loss rate, and birth rate. In the present study, the majority (73%) of females were detected pregnant using ultrasonography in November. The vast majority of fur seals for which an embryo was not detected were indeed pregnant (87%), as evinced by birthing the following summer. Therefore, at the time of ultrasound examination, the

undetected embryo of these seals was likely still in diapause or so recently after diapause that it was too small to be visualized by ultrasonography (2 mm). Pregnancy rates in this study may be slightly underestimated if any non-detectably pregnant females lost their conceptus before ultrasound examination. Loss rates may also be overestimated if any females experienced perinatal loss, or if observers were unable to associate a female with its pup due to cryptic behavior or movement from the intensive study site. These potential errors would be low, given the high reproductive success rates of the northern fur seals in this study. Utilizing ultrasonographic pregnancy diagnosis in combination with reproductive outcomes, the estimated pregnancy rate of northern fur seals was remarkably high (96%; 101 of 105 females conceived; similar to Testa et al., 2010), and pregnancy loss rates were low (11%; 8 of 73 pregnant animals were not observed with a pup the following year when re-sighted). Radiotelemetry data for a few of the females seen without pups suggest possible maternal foraging behavior, and the loss rate may even be as low as 3%. Embryonic diapause ensures that parturition occurs during times of the year when environmental conditions are most conducive to offspring survival (Fenelon et al., 2014; Renfree and Shaw, 2000), and in northern fur seals this is coincident with the only non-pelagic period in their annual life history cycle (Bartholomew and Hoel, 1953; Gentry, 1998). Without this period of embryonic quiescence, the metabolic demands of lactation would overlap that of active gestation, and in northern fur seals fetal growth would be 50–70% complete by the time the previous year’s pup is weaned, and birthing would occur mid-winter. Based on embryo/fetal morphometry in other mammalian models (Ginther, 2007; Kähn, 2004; Kim and Son, 2007; Pretzer, 2008), the size and anatomical features of the fur seal embryos in November suggest that they ranged from 2 to 4 weeks active gestational age. This estimate also corroborates the calculated dates of embryo reactivation in this study. That fur seals with a smaller embryo at the time of ultrasound examination gave birth later the following year supports the notion that the pregnancies were of younger gestational age at the time of examination. For the subset of fur seals in the present study that were observed both at the start of active gestation and again at parturition, results document that embryo reactivation (as a reflection of attachment) varied among individuals by >20 days, and that the duration of active gestation was 8.4 months (with variation of 3.7 weeks; 10.3%). Contrary to the general assumption that embryonic diapause exhibits plasticity in its duration (Reijnders et al., 2010) while active gestation does not (Daniel, 1980), results of this longitudinal study design revealed that variation in northern fur seal active gestation length was similar to the variation in estimated date of embryo reactivation.

84

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89

Fig. 5. Ovarian follicle characteristics (mean ± SE) in northern fur seals examined by transrectal ultrasonography during the period from Nov. 11–24. Figures show trends when cofactors improved model fit using Zero-Inflated GLMM (Poisson distribution). (A) Calendar date and ovarian status impacted the number of small, 2–5 mm follicles. Numbers decreased in the ovary without a CL (left) but not in the CL-bearing ovary (right). (B) The number of large follicles >5 mm diameter were lower in the ovary without a CL, only in detectably pregnant fur seals. (C) The CL-bearing ovary also had the largest follicle, with no effect of date or pregnancy status. (D) The total number of follicles decreased with calendar date only in the ovary without a CL (left) and in detectably pregnant fur seals (solid line). Overall, pregnant fur seals had fewer antral follicles, driven by a progressive decrease in numbers in the ovary without a CL (left).

85

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89

Fig. 6. Serum hormone concentrations in northern fur seals in which an embryo was or was not detected by ultrasonographic examination during Nov. 11–24 (General Linear Model). (A) estradiol, (B) estrone, (C) progesterone, and (D) relaxin. Dashed lines show 95% confidence intervals (points are jittered). Sample sizes (embryonic vesicle detected, no detectable embryonic vesicle): estradiol (n = 92, 32); estrone (n = 85, 28); progesterone (n = 97, 36); relaxin (n = 89, 31).

Table 2 Ovarian status improved model fit between circulating hormone concentrations and calendar date in northern fur seals, as shown by a decrease in Akaike Information Criterion (corrected for small sample size; AICc) and increased log likelihood (LL). Best models are in bold, and the coefficient ± SE, F-values, and P statistic of the added morphological measurement from the ovaries are provided. Model

[Estradiol]

[Estrone]

Fit


Day
Day – Total Number of Follicles ~Day – Number of 2–5 mm Follicles
Day
Day – Total Number of Follicles

Statistics for Added Variable

DAICc

AICc

LL

df

Coefficient

F; P-value

— 8.9 12.8

1,113.9 1,105.0 1,101.1

553.8 548.3 546.4

3 4 4

— 0.201 ± 0.074 0.243 ± 0.079

— F1,111 = 4.7, P = 0.008 F1,111 = 7.7,P = 0.003

— 4.0

802.3 798.3

398.0 394.9

3 4

— 0.104 ± 0.041

— F1,101 = 5.0,P = 0.011

[Progesterone]


Day ~Day + CL size

— 7.3

674.5 667.2

334.1 329.4

3 4

— +0.038 ± 0.012

— F1,109 = 10.4,P = 0.002

[Relaxin]


Day ~Day + Maximum Follicle Diameter

— 5.2

317.3 312.1

155.5 51.9

3 4

— +0.025 ± 0.009

— F1,110 = 7.8,P = 0.006

Integrating ultrasound imaging with endocrine measures provides a broader understanding of the mechanisms critical to establishing pregnancy in northern fur seals. Previous histological analysis of northern fur seal reproductive tracts has shown that the CL remains functional for the first month following ovulation; however, luteal cells become vacuolated with little capability of steroid secretion for the remainder of embryonic diapause (Boshier, 1981; Craig, 1964; Daniel, 1971; Harrison et al., 1952). Coincident with luteal quiescence, folliculogenesis occurs primarily in the CL-bearing ovary (Craig, 1964; Skinner and Westlin-van Aarde, 1989). The end of diapause is associated with a transient elevation in circulating estrogen concentrations (Atkinson, 1997;

Daniel, 1974), agreeing with the higher estradiol and estrone concentrations measured at the earliest sampling dates in this study, and may be necessary for re-activating embryonic growth (Geisert et al., 2015). At the earliest stages of active gestation, fur seals with a detectable embryo already had a significantly larger CL than those that had yet to start active gestation. There is some evidence that marine mammals produce a chorionic gonadotropin similar to primates and equids, which would have a luteotrophic effect aiding CL resurgence during early pregnancy (Hobson and Boyd, 1984; Hobson and Wide, 1986). However, it is likely that the functional components of the CL were still developing since circulating pro-

86

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89

Fig. 7. Serum hormone concentrations in northern fur seals relative to the diameter of the embryonic vesicle detected by ultrasonographic examination (General Linear Models). (A) estradiol, (B) estrone, (C) progesterone, and (D) relaxin. Dashed lines show 95% confidence intervals (points are jittered). Sample sizes: estradiol (n = 92); estrone (n = 85); progesterone (n = 97); relaxin (n = 89).

gesterone concentrations at the onset of active gestation in the present study were not as high as those reported during late gestation in the northern fur seal (Boshier, 1981; Daniel, 1975). Within individuals, a strong correlation between circulating progesterone concentrations and CL size has been documented by serial ultrasonography in other species (Adams et al., 1991; Bergfelt et al., 1989; Kastelic et al., 1990), but the correlation is not as strong from single-point sampling in cross-sectional studies (Honnens et al., 2008; Mann, 2009). After attachment, the CL degenerates and the placenta becomes the primary source of progesterone, maintaining late term pregnancies (Craig, 1964). In domestic species, anovulatory follicular waves continue during the early gestation period concomitant with increasing progesterone concentrations (Adams et al., 2008; Adams and Singh, 2014; Baerwald et al., 2003; Ginther et al., 2001; Guilbault et al., 1986), consistent with the observed follicular dynamics in northern fur seals. An alternative mechanism by which the dam may become physiologically aware of the embryo would be through preferential proliferation of the endometrium of the gravid uterine horn as described in marsupials with diapause (Renfree, 2000), or through acute-phase inflammatory processes driven by embryo attachment similar to some ursids and canids (Vannucchi et al., 2002; Willis et al., 2011). Further, the embryo may utilize a local uteroovarian pathway in its strategy to invoke maternal recognition of pregnancy, as described in domestic ruminants and pigs (Ford and Chenault, 1981; Ginther, 1974, 1976, 1981; Towers et al., 1986). Products of embryonic origin are conveyed to the ipsilateral ovary via intimate apposition of uterine venous drainage with ovarian arterial supply. In northern fur seals, the strong tendency for the CL to be ipsilateral to the gravid uterine horn suggests a similar local utero-ovarian strategy for signaling the presence of the growing conceptus. Maximizing contact between the embryo

and endometrium is another critical step in establishing maternal recognition of pregnancy; however, this is achieved in a speciesspecific manner. The spherical shape of the early fur seal EV is similar to that of other polytocous carnivores (i.e., canids and felids) wherein numerous offspring appose the uterus across both horns (Tsutsui et al., 2002), and equids where trans-uterine mobility of a singular embryo is essential for maintaining pregnancy (Ginther, 1983, 1985; McDowell et al., 1988). In pinnipeds, observations of trans-uterine mobility and utero-ovarian correlations with circulating concentrations of reproductive hormones will require more critical study using serial examination to determine the causal relationships that translate maternal recognition of pregnancy and embryo reactivation (i.e., whether the embryo drives changes in the endometrium and CL, or vice versa). Along with morphological changes in the reproductive tract, temporal changes in circulating hormone concentrations were detected in the present study. However, none could be specifically ascribed as the trigger for embryo reactivation. Reproductive hormones correlated more strongly with calendar date than with embryo presence or size (Figs. 6 and 7; Table S1), suggesting that the temporal patterns of embryo reactivation in the northern fur seal and other pinnipeds may be driven by environmental factors (Boyd, 1991; Temte, 1985a,b). Photoperiodic cues are translated through the pineal gland, with decreasing day length then reducing melatonin production and releasing inhibitory effects on progesterone and estrogen production (Malpaux et al., 2001). Yet, previous work has shown that neither experimental manipulation of photoperiod nor exogenous administration of steroid hormones directly influenced the timing of embryo attachment in northern fur seals (Daniel, 1980), indicating that there may be additional unidentified factor(s) playing a role in the reactivation process such as environmental temperature or nutrient availability

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89

(Boyd, 1996, 1991; Lopes et al., 2004). Indeed, female size and age influenced the timing of embryo reactivation and the start of active gestation in the northern fur seal, and similarly affects the probability of pregnancy in other large mammals (Boltnev and York, 2001; Boyd, 1984, 1991; Guinet et al., 1998; Testa and Adams, 1998). Similar hormone profiles between pregnant and non-pregnant individuals (sometimes referred to as ‘‘pseudopregnancy;” Renouf et al., 1994) until the last few months of gestation, or even the entire pregnancy, is markedly similar to that of the closelyrelated canine bitch (Kowalewski et al., 2015; McKenzie et al., 2005; Reijnders, 1990; Steinetz et al., 1989). In criticallyexamined carnivore models, the only endocrine marker that is a reliable indicator of pregnancy status is relaxin due to its placental origin (Anand-Ivell and Ivell, 2014; Bergfelt et al., 2010; Steinetz et al., 2009, 1989). Interestingly, circulating relaxin concentrations increased in pregnant and non-pregnant fur seals in the present study, presumably before the start of placentation, and correlated with antral follicle size, suggesting that fur seal ovaries may also be a source of relaxin, as in a few other species (Einspanier et al., 1997; Evans et al., 1983). In the present study, relaxin concentrations did not distinguish between pregnant and non-pregnant fur seals during early pregnancy. However, in aquarium-based northern fur seals, serial sampling revealed a progressive increase in relaxin concentrations across gestation, such that serum relaxin was significantly higher during late gestation (3–6 months into active gestation; Bergfelt et al., 2010; Steinetz et al., 2009). The placenta was determined to be the predominant source of relaxin in the late gestation fur seals (Bergfelt et al., 2010), suggesting that relaxin may not be produced in sufficient quantities to be diagnostic of pregnancy until after placentation has occurred. Nonetheless, the high pregnancy rates in this study precluded a comparison of physiological status with non-bred animals; therefore, additional study involving serial examination is warranted to assess the earliest stage that endocrine markers differ between pregnant and truly non-pregnant fur seals. Maintaining a primed uterus and ensuring the functionality of the CL through the periimplantation period, regardless of the presence of an embryo (i.e., prolonged diestrus) may be an evolutionary adaptation ensuring that fewer embryos are lost via mis-timed CL regression. A protracted ‘‘implantation window” and uterine receptivity may be important contributing factors to the northern fur seal’s high reproductive rates noted here and in previous studies (Gentry, 1998; Testa et al., 2010). Ultrasonography in combination with physiological measures provides a powerful tool to better understand the mechanisms governing successful pregnancies in wildlife populations, and to determine the root causes of population declines. Results of the present study do not support the hypothesis that stagnation of the northern fur seal population is attributable to a loss of fertility, or excessive pregnancy loss. This critically important finding allows future fur seal conservation efforts to focus on factors associated with survival to reproductive age, rather than fertility itself, to determine and mitigate causes of an unstable fur seal population. The northern fur seal’s long diestrus makes it difficult to establish a reliable hormone marker for early pregnancy diagnosis, similar to the canine model. Therefore, the hormone markers used for early pregnancy diagnosis in many other domestic species cannot be used to distinguish between pregnant and non-pregnant northern fur seals during the peri-implantation period. Further study is needed with serial sampling, spanning later into active gestation, to elucidate if, and how, the morphological changes observed in ovarian status may be translated into systemic markers to diagnose pregnancy using other minimally-invasive means. A better understanding of the establishment of pregnancy and ovarian dynamics at the time of embryo attachment sets the stage

87

for making informed decisions in conservation management and breeding programs, as well as developing assisted reproductive technology protocols for species in decline (Holt et al., 2003; Kersey and Dehnhard, 2014). Acknowledgments We wish to thank J. Baker, D. Deghetto, E.C. Goertz, C. Kuhn, E. Kunisch, B. Page, P. Pomeroy, J. Sterling, J. Thomason and M. Williams for their assistance in the field. Technical support for analysis of reproductive hormones was provided by D. L. Thompson, Jr at Louisiana State University and B.G. Steinetz at New York University School of Medicine. Fur seal handlings and sample collection were conducted under Marine Mammal Protection Act Permit No. 7821708-00 issued to the National Marine Mammal Laboratory. Funding Overall support was provided by the National Marine Mammal Laboratory, Alaska Fisheries Science Center. Reference to trade names does not imply endorsement by the National Marine Fisheries Service, U.S. National Oceanic and Atmospheric Administration. Support for DRB was provided by the U.S. Environmental Protection Agency. GPA was supported by the University of Saskatchewan, and grants from the Natural Sciences and Engineering Research Council of Canada. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.ygcen.2017.10.007. References Adams, G.P., Jaiswal, R., Singh, J., Malhi, P., 2008. Progress in understanding ovarian follicular dynamics in cattle. Theriogenology 69, 72–80. Adams, G.P., Singh, J., 2014. Ovarian Follicular and Luteal Dynamics in Cattle, Bovine Reproduction. John Wiley & Sons, Inc, pp. 219–244. Adams, G.P., Sumar, J., Ginther, O.J., 1991. Form and function of the corpus luteum in llamas. Anim. Reprod. Sci. 24, 127–138. Adams, G.P., Testa, J.W., Goertz, C.E.C., Ream, R.R., Sterling, J.E., 2007. Ultrasonographic characterization of reproductive anatomy and early embryonic detection in the northern fur seal (Callorhinus ursinus) in the field. Mar. Mammal Sci. 23, 445–452. Anand-Ivell, R., Ivell, R., 2014. Regulation of the reproductive cycle and early pregnancy by relaxin family peptides. Mol. Cell. Endocrinol. 382, 472–479. Atkinson, S., 1997. Repoductive biology of seals. Rev. Reprod. 2, 175–194. Baba, N., Kiyota, M., Yoshida, K., 1991. Relationship between whisker color and age of northern fur seals (Callorhinus ursinus) collected in the western North Pacific and Okhotsk Sea, Annual Meeting of the International North Pacific Fisheries Commission. National Research Institute of Far Seas Fisheries, Shimizu, Tokyo, Japan, p. 7. Baerwald, A.R., Adams, G.P., Pierson, R.A., 2003. A new model for ovarian follicular development during the human menstrual cycle. Fertil. Steril. 80, 116–122. Bartholomew, G.A., Hoel, P.G., 1953. Reproductive behavior of the Alaska fur seal, Callorhinus ursinus. J. Mammal. 34, 417–436. Bazer, F.W., 2013. Pregnancy recognition signaling mechanisms in ruminants and pigs. J. Anim. Sci. Biotechnol. 4, 1–10. Bazer, F.W., Wu, G., Spencer, T.E., Johnson, G.A., Burghardt, R.C., Bayless, K., 2010. Novel pathways for implantation and establishment and maintenance of pregnancy in mammals. Mol. Hum. Reprod. 16, 135–152. Bergfelt, D.R., Pierson, R.A., Ginther, O.J., 1989. Resurgence of the primary corpus luteum during pregnancy in the mare. Anim. Reprod. Sci. 21, 261–270. Bergfelt, D.R., Steinetz, B.G., Dunn, J.L., Atkinson, S., Testa, J.W., Adams, G.P., 2010. Validation of a homologous canine relaxin radioimmunoassay and application with pregnant and non-pregnant Northern fur seals (Callorhinus ursinus). Gen. Comp. Endocrinol. 165, 19–24. Boltnev, A.I., York, A.E., 2001. Maternal investment in northern fur seals (Callorhinus ursinus): interrelationships among mothers’ age, size, parturition date, offspring size and sex ratios. J. Zool. 254, 219–228. Boshier, D., 1981. Structural changes in the corpus luteum and endometrium of seals before implantation. J. Reprod. Fertil. Suppl. 29, 143.

88

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89

Boyd, I., 1996. Individual variation in the duration of pregnancy and birth date in Antarctic fur seals: the role of environment, age, and sex of fetus. J. Mammal. 77, 124–133. Boyd, I.L., 1984. The relationship between body condition and the timing of implantation in pregnant grey seals (Halichoerus grypus). J. Zool. 203, 113–123. Boyd, I.L., 1991. Environmental and physiological factors controlling the reproductive cycles of pinnipeds. Can. J. Zool. 69, 1135–1148. Browne, P., Conley, A.J., Spraker, T., Ream, R.R., Lasley, B.L., 2006. Sex steroid concentrations and localization of steroidogenic enzyme expression in freeranging female northern fur seals (Callorhinus ursinus). Gen. Comp. Endocrinol. 147, 175–183. Clemente, M., de La Fuente, J., Fair, T., Al Naib, A., Gutierrez-Adan, A., Roche, J., Rizos, D., Lonergan, P., 2009. Progesterone and conceptus elongation in cattle: a direct effect on the embryo or an indirect effect via the endometrium? Reproduction 138, 507–517. Comizzoli, P., Crosier, A.E., Gunther, M.S., Howard, J.G., Wildt, D.E., 2009. Advances in reproductive science for wild carnivore conservation. Reprod. Dom. Anim. 44, 47–52. Craig, A., 1964. Histology of reproduction and the estrus cycle in the female fur seal, Callorhinus ursinus. J. Fish. Board Can. 21, 773–811. Cross, J.C., Werb, Z., Fisher, S.J., 1994. Implantation and the placenta: key pieces of the development puzzle. Science 266, 1508. Daniel Jr, J., 1974. Circulating levels of oestradiol-17b during early pregnancy in the Alaskan fur seal showing an oestrogen surge preceding implantation. J. Reprod. Fertil. 37, 425–428. Daniel Jr, J., 1980. Delayed implantation in the northern fur seal (Callorhinus ursinus) and other pinnipeds. J. Reprod. Fertil. Suppl. 29, 35–50. Daniel Jr, J.C., 1971. Growth of the preimplantation embryo of the northern fur seal and its correlation with changes in uterine protein. Dev. Biol. 26, 316–322. Daniel Jr, J.C., 1975. Concentrations of circulating progesterone during early pregnancy in the northern fur seal, Callorhinus ursinus. J. Fish. Board Can. 32, 65–66. Einspanier, A., Zarreh-Hoshyari-Khah, M.R., Balvers, M., Kerr, L., Fuhrmann, K., Ivell, R., 1997. Local relaxin biosynthesis in the ovary and uterus through the oestrous cycle and early pregnancy in the female marmoset monkey (Callithrix jacchus). Hum. Reprod. 12, 1325–1337. Evans, G., Wathes, D.C., King, G.J., Armstrong, D.T., Porter, D.G., 1983. Changes in relaxin production by the theca during the preovulatory period of the pig. J. Reprod. Fertil. 69, 677–683. Fenelon, J.C., Banerjee, A., Murphy, B.D., 2014. Embryonic diapause: development on hold. Int. J. Dev. Biol. 58, 163–174. Ford, S., Chenault, J., 1981. Blood flow to the corpus luteum-bearing ovary and ipsilateral uterine horn of cows during the oestrous cycle and early pregnancy. J. Reprod. Fertil. 62, 555–562. Gales, N.J., Williamson, P., Higgins, L.V., Blackberry, M.A., James, I., 1997. Evidence for a prolonged postimplantation period in the Australian sea lion (Neophoca cinerea). J. Reprod. Fertil. 111, 159–163. Geisert, R.D., Bazer, F.W., 2015. Regulation of Implantation and Establishment of Pregnancy in Mammals. Springer International Publishing AG, Switzerland. Geisert, R.D., Johnson, G.A., Burghardt, R.C., 2015. Implantation and Establishment of Pregnancy in the Pig, Regulation of Implantation and Establishment of Pregnancy in Mammals. Springer, pp. 137–163. Gelatt, T., Ream, R., Johnson, D., 2015. Callorhinus ursinus The IUCN Red List of Threatened Species 2015, e. T3590A45224953. Gentry, R.L., 1998. Behavior and Ecology of the Northern Fur Seal. Princeton University Press, Princeton, NJ. Gentry, R.L., Holt, J.R., 1982. Equipment and techniques for handling northern fur seals. NOAA technical report NMFS SSRF-United States. National Marine Fisheries Service (USA). Ginther, O., 1974. Internal regulation of physiological processes through local venoarterial pathways: a review. J. Anim. Sci. 39, 550–564. Ginther, O., 1976. Comparative anatomy of uteroovarian vasculature. Vet. Scope 20 (1). Ginther, O., 1981. Local versus systemic uteroovarian relationships in farm animals. Acta Vet. Scand. Suppl. 77, 103. Ginther, O., Beg, M., Bergfelt, D., Donadeu, F., Kot, K., 2001. Follicle selection in monovular species. Biol. Reprod. 65, 638–647. Ginther, O.J., 1983. Mobility of the early equine conceptus. Theriogenology 19, 603– 611. Ginther, O.J., 1985. Dynamic physical interactions between the equine embryo and uterus. Equine Vet. J. 17, 41–47. Ginther, O.J., 2007. Ultrasonic imaging and animal reproduction: book 4, colorDoppler ultrasonography. Equiservices Publishing, Cross Plains, WI. Guilbault, L., Dufour, J., Thatcher, W., Drost, M., Haibel, G., 1986. Ovarian follicular development during early pregnancy in cattle. J. Reprod. Fertil. 78, 127–135. Guinet, C., Roux, J.P., Bonnet, M., Mison, V., 1998. Effect of body size, body mass, and body condition on reproduction of female South African fur seals (Arctocephalus pusillus) in Namibia. Can. J. Zool. 76, 1418–1424. Harrison, R.J., 1948. The development and fate of the corpus luteum in the vertebrate series. Biol. Rev. 23, 296–331. Harrison, R.J., Matthews, L.H., Roberts, J.M., 1952. Reproduction in some Pinnipedia. Trans. Zool. Soc. London 27, 437–540. Hobson, B., Boyd, I., 1984. Gonadotrophin and progesterone concentrations in placentae of grey seals (Halichoerus grypus). J. Reprod. Fertil. 72, 521–528.

Hobson, B., Wide, L., 1986. Gonadotrophin in the term placenta of the dolphin (Tursiops truncatus), the Californian sea lion (Zalophus californianus), the grey seal (Halichoerus grypus) and man. J. Reprod. Fertil. 76, 637–644. Holt, W.V., Pickard, A.R., Rodger, J.C., Wildt, D.E., 2003. Reproductive Science and Integrated Conservation. Cambridge University Press. Honnens, A., Voss, C., Herzog, K., Niemann, H., Rath, D., Bollwein, H., 2008. Uterine blood flow during the first 3 weeks of pregnancy in dairy cows. Theriogenology 70, 1048–1056. Kähn, W., 2004. Veterinary Reproductive Ultrasonography. Schlutersche Verlagsgesellschaft mbH & Co. KG, Hans-Bockler-Allee 7, 30173, Hannover. Kastelic, J.P., Bergfelt, D.R., Ginther, O.J., 1990. Relationship between ultrasonic assessment of the corpus luteum and plasma progesterone concentration in heifers. Theriogenology 33, 1269–1278. Kersey, D.C., Dehnhard, M., 2014. The use of noninvasive and minimally invasive methods in endocrinology for threatened mammalian species conservation. Gen. Comp. Endocrinol. 203, 296–306. Kim, B.-S., Son, C.-H., 2007. Time of initial detection of fetal and extra-fetal structures by ultrasonographic examination in Miniature Schnauzer bitches. J. Vet. Sci. 8, 289–293. Kowalewski, M.P., Gram, A., Kautz, E., Graubner, F.R., 2015. The Dog: Nonconformist, Not Only in Maternal Recognition Signaling, Regulation of Implantation and Establishment of Pregnancy in Mammals. Springer, pp. 215–237. Lopes, F.L., Desmarais, J.A., Murphy, B.D., 2004. Embryonic diapause and its regulation. Reproduction 128, 669–678. Macklon, N.S., Geraedts, J.P.M., Fauser, B.C.J.M., 2002. Conception to ongoing pregnancy: the ‘‘black box” of early pregnancy loss. Hum. Reprod. Update 8, 333–343. Malpaux, B., Migaud, M., Tricoire, H., Chemineau, P., 2001. Biology of mammalian photoperiodism and the critical role of the pineal gland and melatonin. J. Biol. Rhythms 16, 336–347. Mann, G., 2009. Corpus luteum size and plasma progesterone concentration in cows. Anim. Reprod. Sci. 115, 296–299. McDowell, K., Sharp, D., Grubaugh, W., Thatcher, W., Wilcox, C., 1988. Restricted conceptus mobility results in failure of pregnancy maintenance in mares. Biol. Reprod. 39, 340–348. McKenzie, J., Parry, L.J., Page, B., Goldsworthy, S.D., 2005. Estimation of pregnancy rates and reproductive failure in New Zealand fur seals (Arctocephalus forsteri). J. Mammal. 86, 1237–1246. Pretzer, S., 2008. Canine embryonic and fetal development: a review. Theriogenology 70, 300–303. R. Core Team, 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Reijnders, P.J.H., 1990. Progesterone and oestradiol-17B concentration profiles throughout the reproductive cycle in harbour seals (Phoca vitulina). J. Reprod. Fertil. 90, 403–409. Reijnders, P.J.H., Brasseur, S.M.J.M., Meesters, E.H.W.G., 2010. Earlier pupping in harbour seals, Phoca vitulina. Biol. Lett. 6, 854–857. Renfree, M.B., 2000. Maternal recognition of pregnancy in marsupials. Rev. Reprod. 5, 6–11. Renfree, M.B., Shaw, G., 2000. Diapause. Annu. Rev. Physiol. 62, 353–375. Renfree, M.B., Shaw, G., 2014. Embryo-endometrial interactions during early development after embryonic diapause in the marsupial tammar wallaby. Int. J. Dev. Biol. 58, 175–181. Renouf, D., Taylor, R., Gales, R., 1994. Pseudopregnancy in harp seals (Phoca groenlandica). J. Reprod. Fertil. 101, 31–36. Roberts, R.M., Xie, S., Mathialagan, N., 1996. Maternal recognition of pregnancy. Biol. Reprod. 54, 294–302. Shero, M.R., Adams, G.P., Burns, J.M., 2015. Field use of ultrasonography to characterize the reproductive tract and early pregnancy in a phocid, the Weddell seal (Leptonychotes weddellii). Anat. Rec. 298, 1970–1977. Skaug, H., Fournier, D., Nielsen, A., Magnusson, A., Bolker, B., 2012. Generalized linear mixed models using AD Model Builder. R package version 0.7. 2.12. See http://glmmadmb/.r-forge.r-project.org. Skinner, J., Westlin-van Aarde, L.M., 1989. Aspects of reproduction in female Ross seals (Ommatophoca rossii). J. Reprod. Fertil. 87, 67–72. Steinetz, B., Lasano, S., Haas, De, van Dorsser, F., Glickman, S., Bergfelt, D., Santymire, R., Songsassen, N., Swanson, W., 2009. Relaxin concentrations in serum and urine of endangered and crazy mixed-up species. Ann. N.Y. Acad. Sci. 1160, 179–185. Steinetz, B.G., Bullesbach, E.E., Goldsmith, L.T., Schwabe, C., Lust, G., 1996. Use of synthetic canine relaxin to develop a rapid homologous radioimmunoassay. Biol. Reprod. 54, 1252–1260. Steinetz, B.G., Goldsmith, L.T., Harvey, H.J., Lust, G., 1989. Serum relaxin and progesterone concentrations in pregnant, pseudopregnant, and ovariectomized, progestin-treated pregnant bitches: detection of relaxin as a marker of pregnancy. Am. J. Vet. Res. 50, 68–71. Teklenburg, G., Salker, M., Heijnen, C., Macklon, N.S., Brosens, J.J., 2010. The molecular basis of recurrent pregnancy loss: impaired natural embryo selection. Mol. Hum. Reprod. 16, 886–895. Temte, J.L., 1985a. Photoperiod and delayed implantation in the northern fur seal (Callorhinus ursinus). J. Reprod. Fertil. 73, 127–131. Temte, J.L., 1985b. Photoperiod and the timing of pupping in the pacific harbor seal (Phoca vitulina richardsi) with notes on reproduction in northern fur seals and dall porpoise. Oregon State Univ., 1–85.

M.R. Shero et al. / General and Comparative Endocrinology 255 (2018) 78–89 Testa, J.W., Adams, G.P., 1998. Body Condition and adjustments to reproductive effort in female moose, Alces alces. J. Mammal. 79, 1345–1354. Testa, J.W., Adams, G.P., Bergfelt, D.R., Johnson, D.S., Ream, R.R., Gelatt, T.S., 2010. Replicating necropsy data without lethal collections: using ultrasonography to understand the decline in northern fur seals. J. Appl. Ecol. 47, 1199–1206. Towers, P., Shaw, G., Renfree, M., 1986. Urogenital vasculature and local steroid concentrations in the uterine branch of the ovarian vein of the female tammar wallaby (Macropus eugenii). J. Reprod. Fertil. 78, 37–47. Trites, A.W., York, A.E., 1993. Unexpected changes in reproductive rates and mean age at first birth during the decline of the Pribilof northern fur seal (Callorhinus ursinus). Can. J. Fish. Aquat. Sci. 50, 858–864. Tsutsui, T., Shimizu, T., Hori, T., Kawakami, E., 2002. Factors affecting transuterine migration of canine embryos. J. Vet. Med. Sci. 64, 1117–1121.

89

Vannucchi, C.I., Mirandola, R.M., Oliveira, C.M., 2002. Acute-phase protein profile during gestation and diestrous: proposal for an early pregnancy test in bitches. Anim. Reprod. Sci. 74, 87–99. Wildt, D.E., Comizzoli, P., Pukazhenthi, B., Songsasen, N., 2010. Lessons from biodiversity – the value of nontraditional species to advance reproductive science, conservation, and human health. Mol. Reprod. Dev. 77, 397–409. Willis, E.L., Kersey, D.C., Durrant, B.S., Kouba, A.J., 2011. The acute phase protein ceruloplasmin as a non-invasive marker of pseudopregnancy, pregnancy, and pregnancy loss in the giant panda. PLoS One 6, e21159. York, A.E., Scheffer, V.B., 1997. Timing of implantation in the northern fur seal, Callorhinus ursinus. J. Mammal. 78, 675–683. Zuur, A.F., Hilbe, J.M., Ieno, E.N., 2013. A Beginner’s Guide to GLM and GLMM with R: A Frequentist and Bayesian Perspective for Ecologists. Highland Statistics Newburgh.