Fig 1.
Synchrony of oocyte development with respect to vitellogenesis differs between fish with group-synchronous (left) and asynchronous (right) oocyte development.
Time is represented by discrete stanzas from the immature class (a, b), to the mature, pre-spawning class (c), to the spawning class (d, e). Prior to vitellogenesis, the pattern for each type of synchrony is the same among immature females (a,b): germ cells are initially composed of oogonia (a) and then primary oocytes emerge (b; oogonia + 1°, where the truncated dotted mode [right] represents a reservoir of both oogonia and previtellogenic oocytes that persists from year to year in iteroparous fishes). Once vitellogenesis occurs, it can occur in two patterns, each with different consequences. With group synchrony (left), annual fecundity equals the standing crop of vitellogenic (2°) oocytes once the 2° oocytes are readily distinguished (i.e., completely larger) from 1° oocytes but before spawning begins (c). Once spawning begins (d), the standing crop of 2° oocytes diminishes (e) as some cells become tertiary (= mature; 3°) oocytes, either in batches (shown here) or in one uninterrupted wave (i.e., total spawning, not shown). With asynchrony (right), annual fecundity cannot be determined at a single point in time because the standing crop of secondary oocytes is replenished during the spawning periods (i.e., de novo vitellogenesis). Oocyte size is scaled relative to a relative egg size of 1.0.
Fig 2.
A sequence of oocyte stages depicting the transition from primary to secondary oocytes.
A. Perinucleolar (PE)–thin chorion, no cytoplasmic inclusions; nucleoli visible around periphery of nucleus (arrow). B. Early cortical alveolar (C1)–thin chorion; tiny clear inclusions appear around cytoplasm (arrow). C, D. Late cortical alveolar (C2)–chorion thickening; dark inclusions appear within the white inclusions of the cytoplasm (arrow). Image D is more advanced; the cortical alveoli grow larger and proximally. E. Early vitellogenesis (V1)–first sign of yolk (red) inclusions appear (arrow). As the cell grows, yolk continues to fill the cytoplasm. F. Cell is still staged as V1 as long as yolk has not expanded to the distal edge of the cytoplasm. Black scale bar: images A–E bar = 50 μm; image F bar = 250 μm.
Table 1.
Characteristics of whole oocytes, by phase of development, and corresponding histology stages.
Appearance is as under transmitted light. Sizes are approximate (100 μm). See text and Fig 2 for description of histology stages.
Table 2.
Characteristics of individual females selected for measurements of oocyte diameters and the corresponding figure they appear in.
Data include: Sampling location along the Connecticut River, sampling date (2015), fish length (fork length [LF], mm), fish age (years, otolith method, one age not available), gonad-somatic index (IG), and histology details (i.e., most advanced oocyte stage [MAOS, see text for definitions] and whether post-ovulatory follicles [POFs] were present [+] or absent [-]).
Fig 3.
Box-whisker plots of oocyte sizes, as measured from histology preparations.
For each of ten oocyte stages (see text and Table 1 for full stage descriptions), the data are represented by a thick horizontal line (median), a box (25–75th percentile), whiskers (range), and any outliers (then the whiskers are roughly the 95% confidence limits). The lower (dotted) horizontal line corresponds to the size (100 μm) of the primary growth (‘anlagen’) oocytes, and the upper (dashed) line corresponds to the lower size (400 μm) of vitellogenic oocytes, as reported by Lehman [6] and assumed by others (e.g., [12]). Values along the top axis are numbers of cells measured by stage. See Fig 2 for images of the first four stages.
Fig 4.
Histograms of oocyte diameters from six upstream migrating, female American shad collected from the Connecticut River in 2015 (one fish per panel).
All females were developing (Class A). Fish in panels A-C were collected by gill net in the lower river, whereas fish in panels D-F were collected at three different power stations: Vernon (D), Cabot (E), and Hadley Falls (F). See Table 2A for more details about each fish. Colors are transparent and overlaid (not stacked), corresponding to small, transparent oocytes (red; primary growth phase), medium, translucent oocytes (yellow; transitional growth phase), and larger, opaque oocytes (green; secondary growth phase). Angled hatching is overlaid on the yellow bars to distinguish them more. See Table 1 for additional details about phases of development.
Fig 5.
Histograms of oocyte diameters from six downstream migrating, female American shad collected from the Connecticut River in 2015.
Fish on the left (panels A-C) were spent, whereas fish on the right were resting (D-F). At least one sample of both maturity classes shown were collected on both June 18 at Hadley Falls Power Station and June 30 at Cabot Power Station. See Table 2B for more details about each fish. Colors and hatching identifying each growth phase are the same as in Fig 4.
Fig 6.
Estimation of potential annual fecundity, as measured by a determinate fecundity method, relative to fish size (log-log transformation).
The size-specific estimate decreases once a fish shows evidence of spawning (see text). The four sampling events and associated correlation coefficients are: A) pre-spawning fish in the lower river (April 30–May 6, r = 0.36, n = 25, P = 0.08), B) pre-spawning fish collected at Hadley Falls Power Station (May 12, r = 0.65, n = 20, P = 0.002), C) spawning fish collected at Hadley Falls and Cabot Power Stations (May 12–13, r = 0.63, n = 11, P = 0.04), and D) spawning fish collected at Hadley Falls and Vernon Power Stations (May 19–20, r = 0.41, n = 10, P = 0.2). Individual fish are plotted as dots, the predictive regression is a blue line, and the 95% confidence limits are enclosed in solid gray.
Fig 7.
Estimates of reproductive characteristics for female American Shad: (A) spawning interval (SI), (B) batch fecundity (BF), (C) residence time (RT) as a proxy for spawning period, and (D) potential annual fecundity (PAF). Data are plotted as open bars together with simulated distributions (solid curves) in panels (A) and (B). RT is simulated as a normal distribution based on hatchery and tagging observations (panel C). Virgin spawner PAF estimates are plotted in panel (D) as vertical lines for the determinate method for the Connecticut River population in 2015 (303,000; dashed line), the determinate method by Leggett [27] for the Connecticut River (263,000; dotted line), and the indeterminate method by Hyle et al. [14] for the York River (511,000; solid line), together with 1,000 bootstrapped mean estimates of PAF (gray bars) from the simulations of PAF = RT/SI × BF (i.e., the indeterminate method for Connecticut River fish in 2015).
Table 3.
Spawning fraction (SF) by sample day, and spawning interval (SI; days), calculated as the inverse of SF, for Connecticut River American shad in 2015.
SF was initially calculated as the proportion of fish spawning per day from different histological markers, each adjusted to spawning: the next day (day -1; most advanced oocyte stage [MAOS] = NM1), that day (Day 0; MAOS = NM2, NM3, H, or POF0 [fresh postovulatory follicle]), yesterday (Day +1; POF1 [12–24 hour old POF]), or the previous day (Day +2; POF2 [24–24 hour old POF]). Final estimate of SF was a mean of these separate values. See text and Table 1 for additional explanation of histology codes. Sampling sites are Hadley Falls (H), Cabot (C), and Vernon (V). N = number of females sampled for histology by date.
Fig 8.
Fecundity-size relationships overlapped between recent and historic collections.
Values here are for pre-spawning females in the Connecticut River (2015)–collected in the lower river (plus [+] symbol) or the Hadley Falls Power Station (cross [×])–compared to two historic data sets: A) pre-spawning females sampled from the Hudson River 1951 (dots [.], data from [6]); B) and predictive, linear model parameters reported for three rivers as sampled in the 1960s: York River, Virginia (dashed, thin line), Connecticut River (solid, thick line), and St. John River, Canada (dotted, thin line) (parameters from [27]; table 30). In (B), raw data were not included in Leggett [27], fork length was measured in cm, and fecundity was not log-transformed (compare to Fig 7A).
Fig 9.
The decreasing trend in potential annual fecundity with increasing latitude, as observed previously with the determinate fecundity method, persists for the two rivers analyzed with the indeterminate fecundity method.
This figure compares determinate fecundity estimates from Leggett and Carscadden [28], indeterminate fecundity estimate for the York River from Hyle et al. [14], and new indeterminate fecundity estimate for the Connecticut River from the current study (new determine fecundity estimate not shown).