White et aL: Reproductive seasonality, fecundity, and spawning frequency of Tautoga onitis 



431 



each fish. Number of spawnings per female was calculated 

 by dividing the number of days in the spawning season by 

 estimated annual spawning frequency. The relationship 

 between mean annual fecundity per 50-mm length inter- 

 val and total length was analyzed with both linear and 

 exponential regression. 



Results 



Description of microscopic gonad stages 



Tautog ovarian development was described by eight micro- 

 scopic gonad stages (Table 1) characteristic of multiple 

 spawning species. Each stage can be differentiated by 

 a unique suite of histological characteristics. Immature 

 ovaries (Fig. 2A) are characterized by the presence of only 

 oogonia and primary growth oocytes within a thin ovar- 

 ian membrane and a relatively high volume of connective 

 tissue. Developing stage ovaries (Fig. 2B ) are characterized 

 by the presence of primary growth, cortical alveoli, and 

 partially yolked oocytes. The fully developed ovary (Fig. 

 2C ) is characterized by the presence of primary growth to 

 advanced yolked ooc3rtes and the absence of oocytes in final 

 oocyte maturation (FOM) classes, POFs, or remnant HOs. 

 Hydrated ovaries (Fig. 2D) are distinguished by the promi- 

 nence of hydrated oocytes still inside the ovarian follicles 

 and may also contain degenerating POFs from an earlier 

 spawning, but they noticeably lack oocytes in the germinal 

 vesicle breakdown state (GVBD). The running ripe stage 

 (Fig. 2E) is classified by the presence of an expanded ovar- 

 ian lumen, ovulated hydrated oocytes free in the lumen 

 (although hydrated oocytes are frequently washed out of 

 the sample during the staining procedure), a large number 

 of fresh POFs, and germinal vessicle migration (GVM) 

 oocytes that tend to be the most advanced stage present 

 within ovigerous folds. Partially spent/redeveloping ovaries 

 (Fig. 2F) are classified by the lack of an ovarian lumen and 

 presence of occasional remnant hydrated oocytes, primary 



growth to GVBD oocytes, and an abundance of POFs. The 

 spent stage (Fig. 2G) is characterized by resorption of 

 yolked oocytes (atresia), and sometimes the presence of 

 macrophage aggregates (MAs), which are groups of cells 

 containing the pigments lipofuscin, ceroid, and melanin 

 (Wolke, 1992). These cells appear to be a collection of 

 scavenging cells that remove cellular debris and foreign 

 substances by phagocytosis when stimulated by excessive 

 degenerating tissue (Wolke, 1992). In tautog ovaries, MAs 

 are assumed to be associated with the resorption of yolked 

 oocytes after the spawning season. Resting stage ovaries 

 (Fig. 2H) contain primary growth and cortical alveoli 

 oocytes, and atresia may be present. Resting ovaries can be 

 distinguished from immature stage ovaries by a thickened 

 ovarian membrane, relatively more oogonia than connec- 

 tive tissue and the presence of MAs. 



The reliability of macroscopic staging to predict actual 

 reproductive stage as detected by microscopic analysis was 

 examined for 484 females. We considered any level of agree- 

 ment above 80% to be acceptable. The ability of macroscopic 

 staging to predict actual microscopic stage varied consider- 

 ably, from 13Vr to 79% agreement for individual stages (Table 

 2), indicating that macroscopic staging was generally unreli- 

 able for estimating actual reproductive stage for many ovar- 

 ian stages. Percent agreement between macroscopic and mi- 

 croscopic staging was only 51% overall. The best agreement 

 was for the fully developed, partially spent/redeveloping, 

 and resting ovarian stages (agreement 69%, 79%, 75%, 

 respectively). Intermediate values were obtained for im- 

 mature, and running ripe stages, (agreement 63% and 67%, 

 respectively). The poorest agreement occurred in assigning 

 developing and spent stages (42%, 13%, respectively). 



Sex ratios 



Of the 938 tautog sexed, 522 (56%) were females and 416 

 were males. Overall, sex ratios varied significantly from an 

 expected 1:1 ratio, with more females than males (1.25:1, 

 ^2=11.98, P<0.01). At lengths <400 mm, females were 



