Journal of Fish Biology (2014) 85, 421–445 doi:10.1111/jfb.12431, available online at wileyonlinelibrary.com
Time course of oocyte development in winter flounder Pseudopleuronectes americanus and spawning seasonality for the Gulf of Maine, Georges Bank and southern New England stocks Y. K. Press*, R. S. McBride † ‡ and M. J. Wuenschel† *Integrated Statistics, 16 Sumner Street, Woods Hole, MA 02543, U.S.A. and †Northeast Fisheries Science Center, National Marine Fisheries Service, 166 Water Street, Woods Hole, MA 02543, U.S.A. (Received 16 September 2013, Accepted 23 April 2014) Winter flounder Pseudopleuronectes americanus were collected at monthly intervals from December 2009 to May 2011, to describe the pattern and seasonality of oocyte development, including: (1) the group-synchronous transition from primary to secondary oocytes that initiates immediately after spawning, (2) the slow (months) development of vitellogenic oocytes followed by the rapid (weeks) maturation of oocytes, (3) the synchronous nature of mature oocytes ovulating, but the discrete releases of benthic eggs in batches, (4) the protracted (months) degradation of postovulatory follicles and (5) the occurrence of follicular atresia. Although fish were collected across only c. 2∘ latitudinal range, the spawning season was c. 1 month later in the Gulf of Maine (GOM) than on Georges Bank and in southern New England. This is probably due to lower temperatures in the GOM. These stock-specific data regarding the time course of oogenesis are of practical value. This information is discussed in relation to measuring and interpreting elements of reproductive potential such as maturation, skipped spawning and fecundity, the response of reproductive traits by this widely distributed species to changing climate and the response by this common, marine-estuarine species to urbanization, particularly environmental pollutants and dredging. Published 2014. This article is a U.S. Government work and is in the public domain in the USA.
Key words: north-east U.S.A.; oogenesis; Pleuronectidae; reproduction.
INTRODUCTION The general patterns and processes governing oogenesis are well understood for a broad array of fish taxa (Tyler & Sumpter, 1996; Burton, 1998; Carrier et al., 2004; Grier et al., 2009; Lubzens et al., 2010). Oocytes develop into eggs following a common blueprint (pattern of stages), yet the seasonality of development can differ considerably among species and produce remarkable variety in patterns of maturity and spawning (Marza, 1938; Murua & Saborido-Rey, 2003; Murua et al., 2003; McBride et al., in press). In extreme examples, age at maturation may occur as quickly as a few weeks or may take several decades (Cortés, 2000; Errea & Danulat, 2001). Once mature, ‡Author to whom correspondence [email protected]
should
be
addressed.
Tel.:
+1
421 Published 2014. This article is a U.S. Government work and is in the public domain in the USA.
508
495
2000;
email:
422
Y. K . P R E S S E T A L.
oocytes may be produced daily (McBride & Thurman, 2003; McBride et al., 2012; Schismenou et al., 2012), initiate only seasonally (Burton et al., 1997; Kurita et al., 2003) or even require over a year to mature (Kennedy et al., 2011; Trinnie et al., 2012). Non-annual spawning has been documented in fishes, a consequence when oogenesis is either slow for all individuals in a population (Doroshov et al., 1997; Trinnie et al., 2012) or because some individuals do not have sufficient energy to advance a clutch of oocytes (Rideout et al., 2005; Rideout & Tomkiewicz, 2011). Thus, the time course of oogenesis explains inter- and even intra-specific variation in maturation and spawning rates. Winter flounder Pseudopleuronectes americanus (Walbaum 1792), a fishery species ranging along the north-east coast of North America, in both the U.S.A. and Canada (Klein-MacPhee, 2002), has been the subject of many laboratory and field studies (Davies et al., 1982; Litvak, 1999). Consequently, its reproductive biology is well known: it is iteroparous, commonly maturing at 2–4 years (McBride et al., 2013); the development of the leading cohort of oocytes is group synchronous (Dunn & Tyler, 1969; Dunn, 1970; Burton & Idler, 1984); down-regulation of yolked oocytes occurs early in vitellogenesis, but the annual, individual fecundity becomes determinate prior to spawning (McElroy et al., 2013); after spawning, there is a nutritionally critical period when fish in poor condition may not advance a cohort of oocytes, leading to skipped spawning the following year (Tyler & Dunn, 1976; Burton & Idler, 1984, 1987). Nonetheless, oogenesis in P. americanus has not been described in detail. Dunn & Tyler (1969) and Burton & Idler (1984) used histological methods in their studies, targeting specific times of year leading up to and just past the spawning period; however, all stages of oogenesis were not detailed. McBride et al. (2013) examined gonad histology in relation to the macroscopic appearance of the gonad, but they collected only in the spring and autumn. In addition, the spawning pattern is presented in different ways by different authors. Burton & Idler (1984) and Burton (1998) describe female P. americanus as single event spawners (i.e. total spawner) that deposit eggs over a very short time, but Murua & Saborido-Rey (2003) classify P. americanus as a batch spawner. Stoner et al. (1999) reported that ovaries appeared homogeneous, indicative of a total spawner, but they also observed individual females spawning as many as 40 times over a period of at least 1 week. These reports can be reconciled if P. americanus are total ovulators, ovulating the entire cohort of mature oocytes in synchrony, but perhaps that requires several days and is followed by the release of eggs from the oviduct in multiple spawning events. If so, then there would be no interruption of oocyte stages through maturation of the oocyte during the spawning season, at least until the individual enters an unambiguous post-spawning, spent condition. In addition to describing the pattern and seasonality of oocyte development, another primary goal of this study was to examine intraspecific variation in spawning seasonality. Many studies report on spawning seasonality within a specific stock: at the northern range, offshore of Newfoundland (Burton & Idler, 1984; 1987), on Georges Bank (GB) (Sibunka et al., 2006) or the southern range, offshore of New Jersey (Stoner et al., 1999; Wuenschel et al., 2009). Summaries of regional differences are available (Colton et al., 1979; Smith, 1985; O’Brien et al., 1993); however, the spatial resolution of these studies is large-grain and the data examined are 25–30 years old, so it may not reflect current spawning patterns. Here, recent collections occurred synoptically among Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
423
P S E U D O P L E U RO N E C T E S A M E R I C A N U S O O G E N E S I S
43° N GOM
42° N
GB
41° N
SNE
0
40° N 72° W
71° W
70° W
69° W
25 50
68° W
100 km
67° W
Fig. 1. Capture locations ( ) of female Pseudopleuronectes americanus examined in this study. Multiple fish were obtained from most locations. Boundaries for the three stock areas, Gulf of Maine (GOM), Georges Bank (GB) and southern New England (SNE), are indicated ( ). The 50 m ( ) and 100 m ( ) contours are also indicated.
three stocks in a central part of this species’ range, but across a latitudinal range of at least 2∘ latitude, in an area of c. 200 km × 500 km (Fig. 1). Thus, the spatial and temporal variability in P. americanus oogenesis was investigated with monthly collections of reproductive and non-reproductive P. americanus females during an 18 month period from three stock areas: southern Gulf of Maine (GOM), GB and southern New England (SNE). The ovary was examined at three complementary levels, organ, cell and sub-cell, to determine the complete time course of oocyte development and subsequent recrudescence or skipped spawning. Some practical aspects of describing the pattern, seasonality and the spatial variation in these traits of oogenesis will also be discussed.
MATERIALS AND METHODS FISH Fish were obtained by bottom trawling, primarily from commercial vessels participating in the Northeast Fisheries Science Center (NEFSC) Cooperative Research Study Fleet programme. Supplemental bottom trawl samples were collected from: (1) other special projects by the NEFSC Cooperative Research Programme, (2) spring and autumn resource monitoring by the NEFSC Ecosystem Survey Branch, (3) spring and autumn resource monitoring by the Massachusetts Department of Fish and Game, Division of Marine Fisheries and (4) winter and spring resource monitoring by the University of Rhode Island, Graduate School of Oceanography. Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
424
Y. K . P R E S S E T A L.
Fish were collected in the three north-eastern U.S.A. stock areas: GOM, GB and SNE (Fig. 1). A total of 1325 female P. americanus were examined; 1032 were mature and 293 immature (Table I). The SNE stock had the largest sample size (n = 606), followed by the GOM stock (n = 476) and the GB stock (n = 243). Every effort was made to collect 30 fish as a mixture of mature and immature fish per month, when samples were pooled among years (Table I). Water temperature data for each region was assembled from NOAA buoys (www.neracoos. org/gomoos). An average daily temperature value (1 m, surface) was downloaded for the period: December 2009 to May 2011, which overlapped with fish sampling. One buoy was selected in each stock area based on relative location and completeness of data during this period [GOM (buoy 44 013; 42⋅35∘ N; 70⋅69∘ W), GB (buoy 44 008; 40⋅5∘ N; 69⋅43∘ W) and SNE (BUZM3; 41⋅4∘ N; 71⋅03∘ W)]. If temperature for a specific day of the year was measured in more than 1 year, then the values were averaged. Some data gaps were evident, particularly in June for the GOM and October to November for SNE, but these do not restrict the general interpretation of the relationship between temperature and spawning. In the laboratory, fish were immediately processed to ensure the quality of the reproductive tissue. For each female, total length (LT ; ±1 mm) and mass (M T ; ±0⋅1 g) were determined. Ovary mass (M O ; ±0⋅001g) was recorded. The gonado-somatic index (I G ) was calculated as follows: I G = 100 M O (M T − M O )−1 , to evaluate the spawning period and peak for each stock area. Approximately 1 ml of tissue was taken from the middle of one ovarian lobe and fixed in 10% buffered formalin. Ovary samples were fixed for a minimum of 1 month before determining oocyte size frequencies to minimize potential changes in oocyte size during fixation (Lowerre-Barbieri & Barbieri, 1993; Heins & Baker, 1999). These fixed samples were also used for histology.
HISTOLOGY Fixed ovary samples were prepared according to standard paraffin embedding techniques (Press et al., 2010). A Schiffs–Mallory trichrome (SMT) stain was initially compared to the standard haematoxylin and eosin (H&E) stain, and it was determined that SMT improved resolution and interpretation, especially for identifying postovulatory follicles (POFs); therefore, SMT was subsequently used for all materials (Press et al., 2010). Histology slides were viewed (×70–700) on a large monitor using a microscope and digital camera system. Preliminary categorization of oocyte stages, including POFs and atresia, was modified from Grier et al. (2009), Lubzens et al. (2010) and Lowerre-Barbieri et al. (2011) (Table II). Oocyte development followed this pattern: (1) primary growth, (2) early cortical alveolar, (3) late cortical alveolar, (4) early vitellogenesis, (5) late vitellogenesis, (6) germinal vesicle migration, (7) germinal vesicle breakdown and (8) ovulated egg. POFs were ranked as: (1) recent, (2) older and (3) oldest, and follicular atresia was ranked as: (1) alpha and (2) beta, based on the level of cellular degradation (Table II). The amount of atresia was also noted to help interpret the processes of down-regulation of late vitellogenic oocytes before, during or after spawning. Encysted cells, secondary growth oocytes and ovulated, but unspawned (remnant) eggs that did not appear to be undergoing a normal rate of degradation were also noted, as these occurred in females with evidence of recent spawning (POFs present). Tunica (gonad wall) thickness was measured from digital images of sections and measurements < or >150 μm were considered thin and thick, respectively, as demonstrated by McBride et al. (2013). Stroma within the ovary were recorded as either thin, with a wispy appearance, or thick if the diameter across was more than double the width of perinucleolar oocytes. M AT U R I T Y A N D S K I P P E D S PAW N I N G Mature and skipped spawning individuals were identified using histological criteria described in recent studies of P. americanus from waters offshore of the north-east U.S.A. (McBride et al., 2012; McElroy et al., 2013). Briefly, all females with a most advanced oocyte stage (MAOS) past primary growth were considered in preparation for spawning (either for the first time or as repeat spawners). Post-spawning evidence indicating maturity and spawning activity included presence of POFs, encysted cells and the tunica thickness. Skipped spawning females have, by Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
2 2 324–328 317–399
1 370
1 315
– –
– –
– –
58 309–485
4 21 351–426 339–439
– –
29 46 77 39 49 20 12 13 21 33 23 276–439 301–479 323–485 286–449 257–456 321–441 309–475 306–393 281–457 290–440 304–434
– –
– –
22 168–333
35 35 16 57 24 26 395–582 350–523 319–405 331–584 350–504 353–471
1 362
September October November December
35 42 9 16 14 8 15 8 98–325 110–337 210–369 160–345 162–310 174–358 191–301 202–325
– –
18 372–492
– –
4 7 6 220–250 230–321 90–329
– –
– –
August
10 312–401
July
69 43 45 34 74 2 17 17 30 15 20 300–400 273–458 296–419 255–455 216–479 307–349 294–461 287–457 299–491 231–386 256–441
June
2 319–332
May
9 3 6 6 30 2 5 8 14 10 5 304–380 304–329 299–362 236–279 99–365 254–279 267–336 270–365 136–340 155–291 274–303
April
GOM Immature n LT range (mm) Mature n LT range (mm) GB Immature n LT range (mm) Mature n LT range (mm) SNE Immature n LT range (mm) Mature n LT range (mm)
March
January February
Stock Maturity
420
186
236
7
376
100
Total (n)
Table I. Number of fish (n) and total length (LT ) range of immature and mature female Pseudopleuronectes americanus. Fish were collected from December 2009 to May 2011, but are pooled here by month and stock area: Gulf of Maine (GOM; n = 476), Georges Bank (GB; n = 243) and southern New England (SNE; n = 606)
P S E U D O P L E U RO N E C T E S A M E R I C A N U S O O G E N E S I S
Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
425
Germinal vesicle breakdown (578⋅7 ± 34⋅8)*
Oocyte maturation Germinal vesicle migration (523⋅7 ± 54⋅9)
Late vitellogenesis (421⋅5 ± 105⋅2)
Early vitellogenesis (183⋅6 ± 24⋅6)
Late cortical alveolar (139⋅4 ± 16⋅7)
Secondary growth Early cortical alveolar (117⋅9 ± 14⋅9)
Primary growth Primary growth (78⋅4 ± 28⋅8)
Histology stage
After the lipoprotein yolk globules filled the cytoplasm, the germinal vesicle migrates towards the animal pole of the zona pellucida) While the germ cell remained in the follicle, the germinal vesicle broke down and the lipoprotein yolk globules fused. Cell diameter increased due to hydration
Cortical alveoli first appeared in the cytoplasm surrounding the germinal vesicle as small white, circular inclusions Dark inclusions appear within the vesicles (observed with Schiffs–Mallory trichrome, not observed with haematoxylin and eosin stain; Press, pers. obs.). The cortical alveoli are organized around the germinal vesicle in the cytoplasm. The zona pellucida thickened by this stage Lipoprotein yolk globules appeared around the periphery of the oocyte and advanced inward towards the germinal vesicle. Cortical alveoli were still visible in the cytoplasm Lipoprotein yolk globules extended more than halfway across the cytoplasm, eventually filling the cytoplasm. Cortical alveoli have been displaced, but were still visible along the edge of the zona pellucida
Chromatin nucleolus and perinucleolar oocytes were typically observed together and classified as primary growth. Chromatin nucleolus oocytes had a single or a few prominent nucleoli in the germinal vesicle and are basophilic. Perinucleolar oocytes increased in diameter and had multiple nucleoli oriented around the periphery of the germinal vesicle
Criteria
Table II. Histology staging for female Pseudopleuronectes americanus. Eight oocyte stages comprise four different developmental periods of oogenesis. Postovulatory follicles (POFs) and follicular atresia are also described. Mean±s.d. oocyte diameter (μm) taken from histology slides are given in parentheses
426 Y. K . P R E S S E T A L.
Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
The germinal vesicle disintegrated, the zona pellucida broke down, the basement membrane remained intact and lipoprotein yolk globules were present in early and late vitellogenic oocytes Internal oocyte components (lipoprotein yolk globules and cortical alveoli) were digested through phagocytosis leaving a vacuous, translucent appearance A pronounced appearance (>50%) of late vitellogenesis oocytes undergoing alpha and beta atresia. This was rare
A complex structure consisting of granulosa cells (inner layer) and theca cells (outer layer). The columnar granulosa cells were typically separated from the theca, forming an inner ring, and both layers stain light purple. The collapsing follicle was loosely arranged and irregular in shape, with a lumen larger than a perinucleolar oocyte The granulosa and theca cell layers remained distinguishable. The granulosa cells, which were no longer columnar, stained deep purple and the theca cells stained light purple. The follicle was now more compact, approximately the size of a perinucleolar oocyte. The lumen was still visible, but much reduced compared to recent POFs The two-layered structure may have been identifiable in some instances, but cell integrity had greatly deteriorated and stained dark purple. The follicle was almost collapsed without a distinguishable lumen and smaller in size than a perinucleolar oocyte
Follicle cells and overlying germinal epithelium parted, releasing the oocyte (now an egg outside the follicle) into the ovarian lumen, leaving behind the POF
Criteria
*Oocyte sizes measured from histological preparations are biased low due to distortion from the dehydration step during preparation and this is most obvious for the hydrated and ovulated stages.
Massive atresia
Beta
Follicular atresia Alpha
Oldest
Older
POFs Recent
Ovulation Egg (558⋅7 ± 91)*
Histology stage
Table II. Continued
P S E U D O P L E U RO N E C T E S A M E R I C A N U S O O G E N E S I S
427
428
Y. K . P R E S S E T A L.
definition, spawned in the past (Rideout et al., 2005). A thick tunica and extensive stroma in the section are indicators of prior spawning (Rideout et al., 2005; McBride et al., 2013). Collections from all months were examined to determine an optimal period that skipped spawning could be identified for all three stocks. Such a window would open during the non-spawning period, after a point when mature, active (not skipping) females would be unambiguously developing a new cohort of secondary oocytes (i.e. late cortical alveolar and vitellogenic oocytes). This window would extend into the spawning season, at least until a point when fish would either show signs of imminent spawning (nucleus migration, breakdown or ovulation) or recent spawning (i.e. POFs, encysted oocytes or eggs). The window would close when unambiguous signs of recent spawning began to disappear.
OOCYTE SIZE FREQUENCIES Although the GOM, GB and SNE showed slightly different spawning periods, each stock area had a single broad and overlapping period of oocyte development per year. Therefore, monthly oocyte size frequencies of mature females were analysed from the GOM stock only, which is assumed to be representative although temporally offset from the other stocks. One individual with representative I G was selected per month to quantify the seasonal cycle of oocytes developing from the reservoir of primary growth oocytes into vitellogenic oocytes and eggs. Oocytes from sub-samples of fixed ovary (c. 0⋅05–0⋅10 g) were manually separated. The sizes of 300 random oocytes per fish were recorded to the nearest 1 μm, by capturing images with a Leica MZ6 scope and DFC295 camera (www.leica-microsystems.com) and measuring the diameters along a single plane using ImageJ software (v. 1.44n, National Institute of Health; http://imagej.nih.gov/) and the ObjectJ (v. 1.01i, University of Amsterdam; http://simon.bio.uva.nl/objectj/) plug-in. Because oocytes >100 μm were easier to separate from each other, smaller oocytes are not well represented in the graphical depictions. In addition to whole oocyte measurements, diameters of 60 cells per oocyte stage were measured from histology slides as a stage-specific size reference. These measurements were limited to oocytes with the nucleus in the histology section to reduce bias associated with sectioning. Diameters from whole and histological preparations were comparable for all oocyte stages except those undergoing hydration, which were non-uniform in shape due to the dehydration step in histological preparation.
RESULTS PAT T E R N O F O O G E N E S I S
All stages of oogenesis, from oogonia proliferation to the ovulation of eggs, as well as the degradation of POFs and atretic oocytes, were observed (Fig. 2). Cell nests, which consisted principally of oogonia [Fig. 2(a)], were observed along the germinal epithelium in all of the fish examined (size range: 90–584 mm LT ). Chromatin nucleolar and perinucleolar stages were categorized as primary growth oocytes. The chromatin nucleolus stage, which marked the beginning of folliculogenesis, was characterized as having a germinal vesicle with a single or a few prominent nucleoli, together with a highly basophilic cytoplasm [Fig. 2(b)]. As these nucleoli moved to the periphery of the germinal vesicle, the perinucleolar stage, the cell diameter increased markedly [Fig. 2(b)]. Initiation of secondary growth oocytes began with the formation of cortical alveoli, which was divided into two stages. During the early cortical alveolar stage, white, circular inclusions (the cortical alveoli) first appeared around the germinal vesicle. In the late cortical alveolar stage, prominent dark dots appeared in the cortical alveoli and the zona pellucida thickened [Fig. 2(c)]. Cortical alveoli remained evident during Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
429
P S E U D O P L E U RO N E C T E S A M E R I C A N U S O O G E N E S I S
(a)
(b)
Pn
(c) LCa
CN
(d)
(e)
EV
(f)
LV
(g)
(h)
GM (i) Egg
Hy
GVBd
(j)
(k) POF
(l)
POF
POF
(m)
(n) At
(o) At
At At
Fig. 2. Stages of Pseudopleuronectes americanus oogenesis (follows Table II). (a) Oogonia nested among primary growth oocytes, (b) primary growth oocytes, with perinucleolar as the most advanced stage, (c) oocytes with late-staged cortical alveoli in the ooplasm, (d) early vitellogenic oocytes, showing how yolk forms first along the periphery of the ooplasm, (e) late vitellogenic oocytes, (f) oocytes with a migrating germinal vesicle, (g–h) oocytes exhibiting yolk fusion and break down of the germinal vesicle, (i) ovulation of the egg from the follicle, (j) recent POFs, (k) older POFs, (l) oldest POFs, (m) alpha atresia of a late vitellogenic oocytes, (n) beta atresia of a late vitellogenic oocytes and (o) a rare example of mass alpha atresia of late vitellogenic oocytes. Scale bars: (a–d) 25 μm, (j–n) 50 μm and (e–i) and (o) 250 μm. At, atretic oocyte; CN, cell nest; EV, early vitellogenic; GVBd, germinal vesicle breakdown; GM, germinal vesicle migration; Hy, hydration; LCa, late cortical alveoli; LV, late vitellogenic; POF, postovulatory follicle; Pn, perinucleolar.
Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
430
Y. K . P R E S S E T A L.
the early vitellogenesis stage when lipoprotein yolk globules began to accumulate at the periphery of the oocyte [Fig. 2(d)]. When lipoprotein yolk globules filled the cytoplasm completely, classified as late vitellogenesis, the oocyte diameter was significantly larger (Table II) and cortical alveoli were displaced towards the periphery [Fig. 2(e)]. Germinal vesicle migration marked the onset of oocyte maturation. As the germinal vesicle migrated in the direction of the animal pole, the oocyte continued to increase in diameter [Table II and Fig. 2(f)]. Germinal vesicle breakdown was characterized by two morphological events: (1) shortly after the germinal vesicle reached the animal pole it broke down and the lipoprotein yolk globules fused together [Fig. 2(g)] and (2) subsequently, while remaining in the follicle, the diameter of the oocyte increased as hydration occurred [Table II and Fig. 2(h)]. Ovulation was identified when the follicle cells and overlying germinal epithelium parted, releasing the oocyte (an egg, once outside the follicle) into the ovarian lumen, leaving behind the POF [Fig. 2(i)]. Mature, post-spawned females were characterized by three stages of POFs. Recent POFs were irregular in shape, loosely arranged with a central lumen and larger than a perinucleolar oocyte (i.e. >35–100 μm; Table II). The complex structure of a recent POF consisted of the inner layer of columnar granulosa cells and the outer layer of theca cells, both staining light purple [Fig. 2(j)]. Older POFs were compact, approximately the size of a perinucleolar oocyte. The central lumen was visible, but much reduced in size compared to recent POFs, and the granulosa and theca cell layers remained distinguishable. The inner granulosa cell layer, which was no longer columnar, stained deep purple and the outer theca cell layer light purple colour [Fig 2(k)]. Oldest POFs were most compact, smaller in size than a perinucleolar oocyte and they were without a distinguishable central lumen. Although greatly deteriorated, the structure of oldest POFs remained identifiable, staining dark purple throughout [Fig 2(l)]. Follicular atresia was not restricted to any one oocyte stage of development, although it was more frequently seen in secondary growth oocytes. Two stages of atresia were evident. Alpha atresia was observed as the initial disintegration of the germinal vesicle and the perforation, folding and collapse of the zona pellucida [Fig. 2(m)]. The basement membrane remained intact and lipoprotein yolk globules were still present. Atresia was categorized as beta after phagocytosis had more or less destroyed evidence of the lipoprotein yolk globules, cortical alveoli and other cytoplasmic components, leaving behind an amorphous appearance [Fig. 2(n)]. High levels of atresia were rare. Specifically, a histology section having >50% of late vitellogenic oocytes in either alpha or beta atresia [Fig. 2(o)] was evident in only seven individuals, representing
Time course of oocyte development in winter flounder Pseudopleuronectes americanus and spawning seasonality for the Gulf of Maine, Georges Bank and southern New England stocks Y. K. Press*, R. S. McBride † ‡ and M. J. Wuenschel† *Integrated Statistics, 16 Sumner Street, Woods Hole, MA 02543, U.S.A. and †Northeast Fisheries Science Center, National Marine Fisheries Service, 166 Water Street, Woods Hole, MA 02543, U.S.A. (Received 16 September 2013, Accepted 23 April 2014) Winter flounder Pseudopleuronectes americanus were collected at monthly intervals from December 2009 to May 2011, to describe the pattern and seasonality of oocyte development, including: (1) the group-synchronous transition from primary to secondary oocytes that initiates immediately after spawning, (2) the slow (months) development of vitellogenic oocytes followed by the rapid (weeks) maturation of oocytes, (3) the synchronous nature of mature oocytes ovulating, but the discrete releases of benthic eggs in batches, (4) the protracted (months) degradation of postovulatory follicles and (5) the occurrence of follicular atresia. Although fish were collected across only c. 2∘ latitudinal range, the spawning season was c. 1 month later in the Gulf of Maine (GOM) than on Georges Bank and in southern New England. This is probably due to lower temperatures in the GOM. These stock-specific data regarding the time course of oogenesis are of practical value. This information is discussed in relation to measuring and interpreting elements of reproductive potential such as maturation, skipped spawning and fecundity, the response of reproductive traits by this widely distributed species to changing climate and the response by this common, marine-estuarine species to urbanization, particularly environmental pollutants and dredging. Published 2014. This article is a U.S. Government work and is in the public domain in the USA.
Key words: north-east U.S.A.; oogenesis; Pleuronectidae; reproduction.
INTRODUCTION The general patterns and processes governing oogenesis are well understood for a broad array of fish taxa (Tyler & Sumpter, 1996; Burton, 1998; Carrier et al., 2004; Grier et al., 2009; Lubzens et al., 2010). Oocytes develop into eggs following a common blueprint (pattern of stages), yet the seasonality of development can differ considerably among species and produce remarkable variety in patterns of maturity and spawning (Marza, 1938; Murua & Saborido-Rey, 2003; Murua et al., 2003; McBride et al., in press). In extreme examples, age at maturation may occur as quickly as a few weeks or may take several decades (Cortés, 2000; Errea & Danulat, 2001). Once mature, ‡Author to whom correspondence [email protected]
should
be
addressed.
Tel.:
+1
421 Published 2014. This article is a U.S. Government work and is in the public domain in the USA.
508
495
2000;
email:
422
Y. K . P R E S S E T A L.
oocytes may be produced daily (McBride & Thurman, 2003; McBride et al., 2012; Schismenou et al., 2012), initiate only seasonally (Burton et al., 1997; Kurita et al., 2003) or even require over a year to mature (Kennedy et al., 2011; Trinnie et al., 2012). Non-annual spawning has been documented in fishes, a consequence when oogenesis is either slow for all individuals in a population (Doroshov et al., 1997; Trinnie et al., 2012) or because some individuals do not have sufficient energy to advance a clutch of oocytes (Rideout et al., 2005; Rideout & Tomkiewicz, 2011). Thus, the time course of oogenesis explains inter- and even intra-specific variation in maturation and spawning rates. Winter flounder Pseudopleuronectes americanus (Walbaum 1792), a fishery species ranging along the north-east coast of North America, in both the U.S.A. and Canada (Klein-MacPhee, 2002), has been the subject of many laboratory and field studies (Davies et al., 1982; Litvak, 1999). Consequently, its reproductive biology is well known: it is iteroparous, commonly maturing at 2–4 years (McBride et al., 2013); the development of the leading cohort of oocytes is group synchronous (Dunn & Tyler, 1969; Dunn, 1970; Burton & Idler, 1984); down-regulation of yolked oocytes occurs early in vitellogenesis, but the annual, individual fecundity becomes determinate prior to spawning (McElroy et al., 2013); after spawning, there is a nutritionally critical period when fish in poor condition may not advance a cohort of oocytes, leading to skipped spawning the following year (Tyler & Dunn, 1976; Burton & Idler, 1984, 1987). Nonetheless, oogenesis in P. americanus has not been described in detail. Dunn & Tyler (1969) and Burton & Idler (1984) used histological methods in their studies, targeting specific times of year leading up to and just past the spawning period; however, all stages of oogenesis were not detailed. McBride et al. (2013) examined gonad histology in relation to the macroscopic appearance of the gonad, but they collected only in the spring and autumn. In addition, the spawning pattern is presented in different ways by different authors. Burton & Idler (1984) and Burton (1998) describe female P. americanus as single event spawners (i.e. total spawner) that deposit eggs over a very short time, but Murua & Saborido-Rey (2003) classify P. americanus as a batch spawner. Stoner et al. (1999) reported that ovaries appeared homogeneous, indicative of a total spawner, but they also observed individual females spawning as many as 40 times over a period of at least 1 week. These reports can be reconciled if P. americanus are total ovulators, ovulating the entire cohort of mature oocytes in synchrony, but perhaps that requires several days and is followed by the release of eggs from the oviduct in multiple spawning events. If so, then there would be no interruption of oocyte stages through maturation of the oocyte during the spawning season, at least until the individual enters an unambiguous post-spawning, spent condition. In addition to describing the pattern and seasonality of oocyte development, another primary goal of this study was to examine intraspecific variation in spawning seasonality. Many studies report on spawning seasonality within a specific stock: at the northern range, offshore of Newfoundland (Burton & Idler, 1984; 1987), on Georges Bank (GB) (Sibunka et al., 2006) or the southern range, offshore of New Jersey (Stoner et al., 1999; Wuenschel et al., 2009). Summaries of regional differences are available (Colton et al., 1979; Smith, 1985; O’Brien et al., 1993); however, the spatial resolution of these studies is large-grain and the data examined are 25–30 years old, so it may not reflect current spawning patterns. Here, recent collections occurred synoptically among Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
423
P S E U D O P L E U RO N E C T E S A M E R I C A N U S O O G E N E S I S
43° N GOM
42° N
GB
41° N
SNE
0
40° N 72° W
71° W
70° W
69° W
25 50
68° W
100 km
67° W
Fig. 1. Capture locations ( ) of female Pseudopleuronectes americanus examined in this study. Multiple fish were obtained from most locations. Boundaries for the three stock areas, Gulf of Maine (GOM), Georges Bank (GB) and southern New England (SNE), are indicated ( ). The 50 m ( ) and 100 m ( ) contours are also indicated.
three stocks in a central part of this species’ range, but across a latitudinal range of at least 2∘ latitude, in an area of c. 200 km × 500 km (Fig. 1). Thus, the spatial and temporal variability in P. americanus oogenesis was investigated with monthly collections of reproductive and non-reproductive P. americanus females during an 18 month period from three stock areas: southern Gulf of Maine (GOM), GB and southern New England (SNE). The ovary was examined at three complementary levels, organ, cell and sub-cell, to determine the complete time course of oocyte development and subsequent recrudescence or skipped spawning. Some practical aspects of describing the pattern, seasonality and the spatial variation in these traits of oogenesis will also be discussed.
MATERIALS AND METHODS FISH Fish were obtained by bottom trawling, primarily from commercial vessels participating in the Northeast Fisheries Science Center (NEFSC) Cooperative Research Study Fleet programme. Supplemental bottom trawl samples were collected from: (1) other special projects by the NEFSC Cooperative Research Programme, (2) spring and autumn resource monitoring by the NEFSC Ecosystem Survey Branch, (3) spring and autumn resource monitoring by the Massachusetts Department of Fish and Game, Division of Marine Fisheries and (4) winter and spring resource monitoring by the University of Rhode Island, Graduate School of Oceanography. Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
424
Y. K . P R E S S E T A L.
Fish were collected in the three north-eastern U.S.A. stock areas: GOM, GB and SNE (Fig. 1). A total of 1325 female P. americanus were examined; 1032 were mature and 293 immature (Table I). The SNE stock had the largest sample size (n = 606), followed by the GOM stock (n = 476) and the GB stock (n = 243). Every effort was made to collect 30 fish as a mixture of mature and immature fish per month, when samples were pooled among years (Table I). Water temperature data for each region was assembled from NOAA buoys (www.neracoos. org/gomoos). An average daily temperature value (1 m, surface) was downloaded for the period: December 2009 to May 2011, which overlapped with fish sampling. One buoy was selected in each stock area based on relative location and completeness of data during this period [GOM (buoy 44 013; 42⋅35∘ N; 70⋅69∘ W), GB (buoy 44 008; 40⋅5∘ N; 69⋅43∘ W) and SNE (BUZM3; 41⋅4∘ N; 71⋅03∘ W)]. If temperature for a specific day of the year was measured in more than 1 year, then the values were averaged. Some data gaps were evident, particularly in June for the GOM and October to November for SNE, but these do not restrict the general interpretation of the relationship between temperature and spawning. In the laboratory, fish were immediately processed to ensure the quality of the reproductive tissue. For each female, total length (LT ; ±1 mm) and mass (M T ; ±0⋅1 g) were determined. Ovary mass (M O ; ±0⋅001g) was recorded. The gonado-somatic index (I G ) was calculated as follows: I G = 100 M O (M T − M O )−1 , to evaluate the spawning period and peak for each stock area. Approximately 1 ml of tissue was taken from the middle of one ovarian lobe and fixed in 10% buffered formalin. Ovary samples were fixed for a minimum of 1 month before determining oocyte size frequencies to minimize potential changes in oocyte size during fixation (Lowerre-Barbieri & Barbieri, 1993; Heins & Baker, 1999). These fixed samples were also used for histology.
HISTOLOGY Fixed ovary samples were prepared according to standard paraffin embedding techniques (Press et al., 2010). A Schiffs–Mallory trichrome (SMT) stain was initially compared to the standard haematoxylin and eosin (H&E) stain, and it was determined that SMT improved resolution and interpretation, especially for identifying postovulatory follicles (POFs); therefore, SMT was subsequently used for all materials (Press et al., 2010). Histology slides were viewed (×70–700) on a large monitor using a microscope and digital camera system. Preliminary categorization of oocyte stages, including POFs and atresia, was modified from Grier et al. (2009), Lubzens et al. (2010) and Lowerre-Barbieri et al. (2011) (Table II). Oocyte development followed this pattern: (1) primary growth, (2) early cortical alveolar, (3) late cortical alveolar, (4) early vitellogenesis, (5) late vitellogenesis, (6) germinal vesicle migration, (7) germinal vesicle breakdown and (8) ovulated egg. POFs were ranked as: (1) recent, (2) older and (3) oldest, and follicular atresia was ranked as: (1) alpha and (2) beta, based on the level of cellular degradation (Table II). The amount of atresia was also noted to help interpret the processes of down-regulation of late vitellogenic oocytes before, during or after spawning. Encysted cells, secondary growth oocytes and ovulated, but unspawned (remnant) eggs that did not appear to be undergoing a normal rate of degradation were also noted, as these occurred in females with evidence of recent spawning (POFs present). Tunica (gonad wall) thickness was measured from digital images of sections and measurements < or >150 μm were considered thin and thick, respectively, as demonstrated by McBride et al. (2013). Stroma within the ovary were recorded as either thin, with a wispy appearance, or thick if the diameter across was more than double the width of perinucleolar oocytes. M AT U R I T Y A N D S K I P P E D S PAW N I N G Mature and skipped spawning individuals were identified using histological criteria described in recent studies of P. americanus from waters offshore of the north-east U.S.A. (McBride et al., 2012; McElroy et al., 2013). Briefly, all females with a most advanced oocyte stage (MAOS) past primary growth were considered in preparation for spawning (either for the first time or as repeat spawners). Post-spawning evidence indicating maturity and spawning activity included presence of POFs, encysted cells and the tunica thickness. Skipped spawning females have, by Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
2 2 324–328 317–399
1 370
1 315
– –
– –
– –
58 309–485
4 21 351–426 339–439
– –
29 46 77 39 49 20 12 13 21 33 23 276–439 301–479 323–485 286–449 257–456 321–441 309–475 306–393 281–457 290–440 304–434
– –
– –
22 168–333
35 35 16 57 24 26 395–582 350–523 319–405 331–584 350–504 353–471
1 362
September October November December
35 42 9 16 14 8 15 8 98–325 110–337 210–369 160–345 162–310 174–358 191–301 202–325
– –
18 372–492
– –
4 7 6 220–250 230–321 90–329
– –
– –
August
10 312–401
July
69 43 45 34 74 2 17 17 30 15 20 300–400 273–458 296–419 255–455 216–479 307–349 294–461 287–457 299–491 231–386 256–441
June
2 319–332
May
9 3 6 6 30 2 5 8 14 10 5 304–380 304–329 299–362 236–279 99–365 254–279 267–336 270–365 136–340 155–291 274–303
April
GOM Immature n LT range (mm) Mature n LT range (mm) GB Immature n LT range (mm) Mature n LT range (mm) SNE Immature n LT range (mm) Mature n LT range (mm)
March
January February
Stock Maturity
420
186
236
7
376
100
Total (n)
Table I. Number of fish (n) and total length (LT ) range of immature and mature female Pseudopleuronectes americanus. Fish were collected from December 2009 to May 2011, but are pooled here by month and stock area: Gulf of Maine (GOM; n = 476), Georges Bank (GB; n = 243) and southern New England (SNE; n = 606)
P S E U D O P L E U RO N E C T E S A M E R I C A N U S O O G E N E S I S
Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
425
Germinal vesicle breakdown (578⋅7 ± 34⋅8)*
Oocyte maturation Germinal vesicle migration (523⋅7 ± 54⋅9)
Late vitellogenesis (421⋅5 ± 105⋅2)
Early vitellogenesis (183⋅6 ± 24⋅6)
Late cortical alveolar (139⋅4 ± 16⋅7)
Secondary growth Early cortical alveolar (117⋅9 ± 14⋅9)
Primary growth Primary growth (78⋅4 ± 28⋅8)
Histology stage
After the lipoprotein yolk globules filled the cytoplasm, the germinal vesicle migrates towards the animal pole of the zona pellucida) While the germ cell remained in the follicle, the germinal vesicle broke down and the lipoprotein yolk globules fused. Cell diameter increased due to hydration
Cortical alveoli first appeared in the cytoplasm surrounding the germinal vesicle as small white, circular inclusions Dark inclusions appear within the vesicles (observed with Schiffs–Mallory trichrome, not observed with haematoxylin and eosin stain; Press, pers. obs.). The cortical alveoli are organized around the germinal vesicle in the cytoplasm. The zona pellucida thickened by this stage Lipoprotein yolk globules appeared around the periphery of the oocyte and advanced inward towards the germinal vesicle. Cortical alveoli were still visible in the cytoplasm Lipoprotein yolk globules extended more than halfway across the cytoplasm, eventually filling the cytoplasm. Cortical alveoli have been displaced, but were still visible along the edge of the zona pellucida
Chromatin nucleolus and perinucleolar oocytes were typically observed together and classified as primary growth. Chromatin nucleolus oocytes had a single or a few prominent nucleoli in the germinal vesicle and are basophilic. Perinucleolar oocytes increased in diameter and had multiple nucleoli oriented around the periphery of the germinal vesicle
Criteria
Table II. Histology staging for female Pseudopleuronectes americanus. Eight oocyte stages comprise four different developmental periods of oogenesis. Postovulatory follicles (POFs) and follicular atresia are also described. Mean±s.d. oocyte diameter (μm) taken from histology slides are given in parentheses
426 Y. K . P R E S S E T A L.
Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
The germinal vesicle disintegrated, the zona pellucida broke down, the basement membrane remained intact and lipoprotein yolk globules were present in early and late vitellogenic oocytes Internal oocyte components (lipoprotein yolk globules and cortical alveoli) were digested through phagocytosis leaving a vacuous, translucent appearance A pronounced appearance (>50%) of late vitellogenesis oocytes undergoing alpha and beta atresia. This was rare
A complex structure consisting of granulosa cells (inner layer) and theca cells (outer layer). The columnar granulosa cells were typically separated from the theca, forming an inner ring, and both layers stain light purple. The collapsing follicle was loosely arranged and irregular in shape, with a lumen larger than a perinucleolar oocyte The granulosa and theca cell layers remained distinguishable. The granulosa cells, which were no longer columnar, stained deep purple and the theca cells stained light purple. The follicle was now more compact, approximately the size of a perinucleolar oocyte. The lumen was still visible, but much reduced compared to recent POFs The two-layered structure may have been identifiable in some instances, but cell integrity had greatly deteriorated and stained dark purple. The follicle was almost collapsed without a distinguishable lumen and smaller in size than a perinucleolar oocyte
Follicle cells and overlying germinal epithelium parted, releasing the oocyte (now an egg outside the follicle) into the ovarian lumen, leaving behind the POF
Criteria
*Oocyte sizes measured from histological preparations are biased low due to distortion from the dehydration step during preparation and this is most obvious for the hydrated and ovulated stages.
Massive atresia
Beta
Follicular atresia Alpha
Oldest
Older
POFs Recent
Ovulation Egg (558⋅7 ± 91)*
Histology stage
Table II. Continued
P S E U D O P L E U RO N E C T E S A M E R I C A N U S O O G E N E S I S
427
428
Y. K . P R E S S E T A L.
definition, spawned in the past (Rideout et al., 2005). A thick tunica and extensive stroma in the section are indicators of prior spawning (Rideout et al., 2005; McBride et al., 2013). Collections from all months were examined to determine an optimal period that skipped spawning could be identified for all three stocks. Such a window would open during the non-spawning period, after a point when mature, active (not skipping) females would be unambiguously developing a new cohort of secondary oocytes (i.e. late cortical alveolar and vitellogenic oocytes). This window would extend into the spawning season, at least until a point when fish would either show signs of imminent spawning (nucleus migration, breakdown or ovulation) or recent spawning (i.e. POFs, encysted oocytes or eggs). The window would close when unambiguous signs of recent spawning began to disappear.
OOCYTE SIZE FREQUENCIES Although the GOM, GB and SNE showed slightly different spawning periods, each stock area had a single broad and overlapping period of oocyte development per year. Therefore, monthly oocyte size frequencies of mature females were analysed from the GOM stock only, which is assumed to be representative although temporally offset from the other stocks. One individual with representative I G was selected per month to quantify the seasonal cycle of oocytes developing from the reservoir of primary growth oocytes into vitellogenic oocytes and eggs. Oocytes from sub-samples of fixed ovary (c. 0⋅05–0⋅10 g) were manually separated. The sizes of 300 random oocytes per fish were recorded to the nearest 1 μm, by capturing images with a Leica MZ6 scope and DFC295 camera (www.leica-microsystems.com) and measuring the diameters along a single plane using ImageJ software (v. 1.44n, National Institute of Health; http://imagej.nih.gov/) and the ObjectJ (v. 1.01i, University of Amsterdam; http://simon.bio.uva.nl/objectj/) plug-in. Because oocytes >100 μm were easier to separate from each other, smaller oocytes are not well represented in the graphical depictions. In addition to whole oocyte measurements, diameters of 60 cells per oocyte stage were measured from histology slides as a stage-specific size reference. These measurements were limited to oocytes with the nucleus in the histology section to reduce bias associated with sectioning. Diameters from whole and histological preparations were comparable for all oocyte stages except those undergoing hydration, which were non-uniform in shape due to the dehydration step in histological preparation.
RESULTS PAT T E R N O F O O G E N E S I S
All stages of oogenesis, from oogonia proliferation to the ovulation of eggs, as well as the degradation of POFs and atretic oocytes, were observed (Fig. 2). Cell nests, which consisted principally of oogonia [Fig. 2(a)], were observed along the germinal epithelium in all of the fish examined (size range: 90–584 mm LT ). Chromatin nucleolar and perinucleolar stages were categorized as primary growth oocytes. The chromatin nucleolus stage, which marked the beginning of folliculogenesis, was characterized as having a germinal vesicle with a single or a few prominent nucleoli, together with a highly basophilic cytoplasm [Fig. 2(b)]. As these nucleoli moved to the periphery of the germinal vesicle, the perinucleolar stage, the cell diameter increased markedly [Fig. 2(b)]. Initiation of secondary growth oocytes began with the formation of cortical alveoli, which was divided into two stages. During the early cortical alveolar stage, white, circular inclusions (the cortical alveoli) first appeared around the germinal vesicle. In the late cortical alveolar stage, prominent dark dots appeared in the cortical alveoli and the zona pellucida thickened [Fig. 2(c)]. Cortical alveoli remained evident during Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
429
P S E U D O P L E U RO N E C T E S A M E R I C A N U S O O G E N E S I S
(a)
(b)
Pn
(c) LCa
CN
(d)
(e)
EV
(f)
LV
(g)
(h)
GM (i) Egg
Hy
GVBd
(j)
(k) POF
(l)
POF
POF
(m)
(n) At
(o) At
At At
Fig. 2. Stages of Pseudopleuronectes americanus oogenesis (follows Table II). (a) Oogonia nested among primary growth oocytes, (b) primary growth oocytes, with perinucleolar as the most advanced stage, (c) oocytes with late-staged cortical alveoli in the ooplasm, (d) early vitellogenic oocytes, showing how yolk forms first along the periphery of the ooplasm, (e) late vitellogenic oocytes, (f) oocytes with a migrating germinal vesicle, (g–h) oocytes exhibiting yolk fusion and break down of the germinal vesicle, (i) ovulation of the egg from the follicle, (j) recent POFs, (k) older POFs, (l) oldest POFs, (m) alpha atresia of a late vitellogenic oocytes, (n) beta atresia of a late vitellogenic oocytes and (o) a rare example of mass alpha atresia of late vitellogenic oocytes. Scale bars: (a–d) 25 μm, (j–n) 50 μm and (e–i) and (o) 250 μm. At, atretic oocyte; CN, cell nest; EV, early vitellogenic; GVBd, germinal vesicle breakdown; GM, germinal vesicle migration; Hy, hydration; LCa, late cortical alveoli; LV, late vitellogenic; POF, postovulatory follicle; Pn, perinucleolar.
Published 2014. This article is a U.S. Government work and is in the public domain in the USA. Journal of Fish Biology 2014, 85, 421–445
430
Y. K . P R E S S E T A L.
the early vitellogenesis stage when lipoprotein yolk globules began to accumulate at the periphery of the oocyte [Fig. 2(d)]. When lipoprotein yolk globules filled the cytoplasm completely, classified as late vitellogenesis, the oocyte diameter was significantly larger (Table II) and cortical alveoli were displaced towards the periphery [Fig. 2(e)]. Germinal vesicle migration marked the onset of oocyte maturation. As the germinal vesicle migrated in the direction of the animal pole, the oocyte continued to increase in diameter [Table II and Fig. 2(f)]. Germinal vesicle breakdown was characterized by two morphological events: (1) shortly after the germinal vesicle reached the animal pole it broke down and the lipoprotein yolk globules fused together [Fig. 2(g)] and (2) subsequently, while remaining in the follicle, the diameter of the oocyte increased as hydration occurred [Table II and Fig. 2(h)]. Ovulation was identified when the follicle cells and overlying germinal epithelium parted, releasing the oocyte (an egg, once outside the follicle) into the ovarian lumen, leaving behind the POF [Fig. 2(i)]. Mature, post-spawned females were characterized by three stages of POFs. Recent POFs were irregular in shape, loosely arranged with a central lumen and larger than a perinucleolar oocyte (i.e. >35–100 μm; Table II). The complex structure of a recent POF consisted of the inner layer of columnar granulosa cells and the outer layer of theca cells, both staining light purple [Fig. 2(j)]. Older POFs were compact, approximately the size of a perinucleolar oocyte. The central lumen was visible, but much reduced in size compared to recent POFs, and the granulosa and theca cell layers remained distinguishable. The inner granulosa cell layer, which was no longer columnar, stained deep purple and the outer theca cell layer light purple colour [Fig 2(k)]. Oldest POFs were most compact, smaller in size than a perinucleolar oocyte and they were without a distinguishable central lumen. Although greatly deteriorated, the structure of oldest POFs remained identifiable, staining dark purple throughout [Fig 2(l)]. Follicular atresia was not restricted to any one oocyte stage of development, although it was more frequently seen in secondary growth oocytes. Two stages of atresia were evident. Alpha atresia was observed as the initial disintegration of the germinal vesicle and the perforation, folding and collapse of the zona pellucida [Fig. 2(m)]. The basement membrane remained intact and lipoprotein yolk globules were still present. Atresia was categorized as beta after phagocytosis had more or less destroyed evidence of the lipoprotein yolk globules, cortical alveoli and other cytoplasmic components, leaving behind an amorphous appearance [Fig. 2(n)]. High levels of atresia were rare. Specifically, a histology section having >50% of late vitellogenic oocytes in either alpha or beta atresia [Fig. 2(o)] was evident in only seven individuals, representing
