Croakers and drums or sciaenids (family Sciaenidae) have long been important species of coastal fisheries in warm temperate and tropical nations. Estimated annual global sciaenid catches increased from approximately 241,300 tonnes (t) in 1950 to over 1,000,000 t in 1995 for the first time and over 1,500,000 t in 2006–2019 (www.fao.org/fishery/statistics/global-capture-production/en). Many sciaenid stocks are facing declines and some are of conservation concern by the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) and the International Union for Conservation of Nature (IUCN) (FAO, 2018; Oliveira and Oliveira, 2018; FAO, 2020; Cisneros-Mata et al., 2021). The global assessments of sciaenids (N = 286) using the 2001 IUCN Red List Categories and Criteria (version 3.1) revealed that 5.6% of sciaenids are threatened, including vulnerable (VU), endangered (EN) or critically endangered (CR), and 1.4% near threatened (NT) (www.iucnredlist.org). The fishery operations of sciaenids are mainly associated with their biological features, including targeting their nearshore and river margin spawning and nursery aggregations (Chao et al., 2015).

China (mainland, excluding Hong Kong, Macao and Taiwan, unless otherwise specified) is the largest capture fisheries country in the world (Kang et al., 2018). Sciaenid capture fisheries have been of significance in domestic marine fisheries. Among the 26 marine fish species and species groups available for statistical catch volumes, six are from sciaenids, including the large yellow croaker Larimichthys crocea, the small yellow croaker Larimichthys polyactis, the Mi-iuy croaker Miichthys miiuy, Pennahia species, Nibea species, and Collichthys species (MARA, 2021). The estimated capture productions of sciaenids varied from 223,121 t in 1956 to 734,285 t in 2020, exceeding one million t in 2012, 2013, and 2015, constituting an average of 9.2% (between 36.1% in 1968 and 1.0% in 1989) of the annual total marine fish capture fishery production (MOA, 1950–2018; MARA, 2019–2021) ( Figure 1 ). L. crocea and L. polyactis have the longest statistical datasets since 1956, indicating their commercially importance in Chinese domestic marine fisheries.

L. crocea is distributed in southern Yellow Sea, East China Sea, and northern South China Sea; therefore, it is largely endemic in Chinese waters (Liu et al., 2020a). In China, the annual catch volumes of L. crocea were over 50,000 t in the 1950s–early 1980s, exceeding 150,000 t in some years ( Figure 1 ). The productions came mainly from the exploitation of the spawning and over-wintering aggregations in its entire distribution region (Liu and Sadovy de Mitcheson, 2008). The L. crocea fishery collapsed in the late 1980s, and the annual catches were less than 20,000 t; the aggregations of L. crocea were no longer significance and subsequently disappeared (Liu and Sadovy de Mitcheson, 2008). Various measures have been introduced for L. crocea management, including the prohibition of the drag seine nets since the mid-1950s, which particularly targeting the spawning aggregations, and the protection of nearshore spawning grounds by establishing protected areas, reducing fishing pressure, and controlling fishing gears since the 1980s (Liu and Sadovy de Mitcheson, 2008). Furthermore, L. crocea mariculture and juvenile restocking have been promoted, and the national fishing moratorium regulation in May to August in all Chinese seas has been enforced since the 1990s (MOA, 2000; Liu and Sadovy de Mitcheson, 2008; Liu et al., 2020b). Irrespective of the increase of L. crocea capture productions that has been reported in the past two decades ( Figure 1 ), direct evidence for supporting stock recovery has not been reported since the 1990s, such as the reappearance of spawning aggregations, and the collection of fully mature females and males in the historical spawning grounds (Ye et al., 2020; Xu et al., 2021). Understanding the reproductive dynamics of commercially important fishes can provide valuable information for stock assessment and management measure evaluation (Sadovy and Domeier, 2005; Lowerre-Barbieri et al., 2011; Farley et al., 2013; Sadovy de Mitcheson et al., 2013).

Sansha Bay (26.40−27.00° N, 119.50−120.20° E), a typical semi-enclosed bay, is located in northern Fujian Province of China. Within Sansha Bay, there is a well-known L. crocea spawning ground, i.e., Guanjingyang (GJY) spawning ground, the only semi-enclosed bay type spawning ground among the 15 spawning grounds identified in its distribution region (Liu and Sadovy de Mitcheson, 2008; Zhang et al., 2011) ( Figure 2 ). On the basis of the catch data, two spawning seasons were reported for L. crocea in GJY, i.e., in May to June and in September to October, referring to the spring and autumn spawning seasons, respectively; furthermore, the spring spawning season was considered as the main one (Chu and Wu, 1985; Zhang and Hong, 2015). The reductions of the minimum size [standard length (SL)] and body weight (BW) at female sexual maturity and the increase of growth rate in L. crocea were observed in GJY from the late 1950s to the late 1980s over three decades (Xu et al., 1980; Lin et al., 1992; Liu and Sadovy de Mitcheson, 2008). Although there were biological data available aforementioned, detailed studies through year-round sampling to determine the natural spawning season and the peak and the size at 50% maturity using the cost-effective gonad histology method have not been conducted for L. crocea in GJY, neither throughout the distribution region.

This study was conducted in GJY L. crocea spawning ground, and the objectives were 1) to determine the current spawning season and its peak(s), 2) to examine the minimum sizes of sexual maturity and the sizes at 50% maturity for both females and males, and 3) to evaluate the changes of reproductive patterns over decades. The results will help us understand the current status of GJY spawning ground and discuss the national fishing moratorium regulation that applied since the 1990s.

Four biological changes on L. crocea were noticed over years in GJY spawning ground. First, the declines of the maximum size were observed over six decades. The maximum size in catches (N = 173) in May to June 1959 (spring spawning season) was 515 mm SL, with a high proportion (11%) larger than 400 mm SL (Xu et al., 1980). In 1986–1990, the maximum size in catches (N = 210) was 430 mm SL (Lin et al., 1992). In 2019–2021 (this study), the maximum sizes were 391 mm SL in total catches (N = 1,006) and 362 mm SL in May to June catches (N = 320) ( Table 2 ). Unsustainable exploitation has confirmed to play a major role in the shift of size structure toward small individuals (Shin et al., 2005; Tu et al., 2018). This study showed that the maximum size declined nearly 30% in spring spawning season over six decades for L. crocea in GJY.

Second, the reductions of the sizes at female and male sexual maturity were identified on L. crocea in GJY ( Table 3 ). In nearly three decades from 1959 to 1986–1990, the minimum SL for female maturity declined 12.5%, with a further decline of 8.6% over the past three decades from 1986–1990 to 2019–2021. For males, the reduction of the minimum size at maturity was greater, nearly 32% over the past three decades from 1986–1990 to 2019–2021. Although the methods for determining SL₅₀ (so called the majority proportion for maturity) were not standardized, the declines of SL₅₀ were clear over the past three decades from 1986–1990 to 2019–2021: 16.5% and 19% for females and males, respectively. Furthermore, the estimated SL₅₀ for female maturity on L. crocea was nearly 40 mm larger than that of males in 2019–2021; similar results were found in 1986–1990 ( Table 3 ). For some sciaenids studied, the estimated SL₅₀ of females were all larger than that of males (Tuuli et al., 2011; Militelli et al., 2012; Zhang et al., 2019). Larger mature females indicate higher fecundity, and this would have the most substantial impact on stock resistance and recovery (Farley et al., 2013; Bris et al., 2015; Sabrah et al., 2017). However, in other sciaenids, such as Plagioscion magdalenae and P. squamosissimus, SL₅₀ in males was larger than for females (Castro, 1999; Marciano et al., 2005; Santos et al., 2010). This may be associated with specific characteristics or differential responses to fishery exploitation in different species (Santos et al., 2010). The over-exploitation of spawning and over-wintering aggregations and the loss of genetic diversity are likely to contribute to the size reduction at maturity which would have long-term impact on population structure and reproductive pattern (Liu and Sadovy de Mitcheson, 2008).

Third, a nearly year-round spawning pattern was observed for L. crocea females and males in GJY, the first time for the species. The same phenomenon of year-round spawning pattern has also been reported in other sciaenids, e.g., females of the tiger tooth croaker Otolithes ruber and the bigeye croaker Pennahia anea (Menon et al., 2015; Farkhondeh et al., 2018; Lanzuela et al., 2020). On the basis of the locations of these different studies, the year-round spawning pattern in sciaenids can be found in Indian and West Pacific Oceans, and from tropical to temperate.

Fourth, the shifts of spawning peaks were observed for the first time in L. crocea. The well-known two spawning peaks in GJY spawning ground were in May to June and in September to October with the spring was a major (Chu and Wu, 1985; Zhang and Hong, 2015) ( Table 3 ). At least 2 months earlier in spring peak and 1 month later in autumn peak were noticed in this study based on cost-effective, gonad histology method; no any other reports mentioned the March and November spawning peaks in GJY.

The significant findings on year-round spawning activity and the shift of spawning peak in L. crocea merit further investigations. Temperature is likely to be the dominant factor influencing the variability of migration, spawning, and recruitment on animals (Gibson et al., 1993; Marshall and Elliott, 1998; Pankhurst and Munday, 2011; Golpour et al., 2021). The global warming is a non-negligible factor which can affect the reproductive dynamics of fishes. Elevated seawater temperature would stimulate the earlier spawning activity of spring and summer spawners while delaying the onset of sexual maturation of autumn spawners and extending the spawning duration of marine and freshwater fishes (Pankhurst and King, 2010; Yamamoto and Shiah, 2012; Rogers and Dougherty, 2019; Kawai et al., 2020). In GJY, the annual average temperature has increased by 1.65°C in the past four decades (Meteorology Bureau of Fujian Province, 1981–2010; Ningde Bureau of Statistics, 2011–2020) ( Figure 9A ). L. crocea can spawn in the wild at temperature of 18°C–24°C, and stop spawning at the temperature above 26°C (Liu, 2004; Yamada et al., 2007; Xu, 2018). The highest temperature in GJY was found in July, > 28°C (Xu, 2018). In this study, females did not spawn in July and August ( Figure 7A ) and were likely to be associated with high temperature, > 26°C, in summer ( Figure 9B ).

This study provided two pieces of evidence that GJY still functions as the spawning ground for L. crocea. First, eggs of L. crocea were collected in April to June, August, October, and November (Dai, 2006; Shen, 2011; Xu, 2018; Jiang et al., 2021), matching largely to the two spawning peaks (spring and autumn) determined in this study except August. The L. crocea eggs mainly distributed in the eastern Sandu Island, GJY waters with extensions to the entry of Sansha Bay and Dongwuyang (Xu, 2018; Jiang et al., 2021) ( Figure 10 ), indicating the spawning areas nearby in terms of the short embryonic development period (30–52 h under temperature of 18°C–23°C) (Sha, 1962; Liu, 1999). Second, females of L. crocea with HO and/or POF were collected in this study overlapped with the areas where eggs collected ( Figure 10 ). Further studies are needed to investigate whether L. crocea forms spawning migration and enters Sansha Bay to reproduce, and the scale of the aggregations. At least from this study, part of the spawning stock is likely to remain in the Bay year-round and does not form spawning migration.

However, L. crocea larvae, as an important stage of life cycle, were rare and sporadic in plankton collections in GJY (Xu, 2018; Jiang et al., 2021). The changes in hydrological and ecological condition in Sansha Bay could have negative impact on survival rate of L. crocea larvae. For instance, the Noctiluca scintillans blooms and the red tides caused by the eutrophication would have detrimental effect on the survival of L. crocea larvae directly, such as hypoxia and toxins, or indirectly through the shortage of zooplankton as diets (Jiang et al., 2021). The environmental and ecological factors influence the survival of L. crocea larvae merit further investigation.

The management measures for L. crocea in Sansha Bay are diverse, with the prohibition of the drag seine nets in the 1950s, the establishment of the protected area for spawning aggregations in the 1980s, the conduction of long-term restocking programs since the 1990s, the introduction of national fishing moratorium regulation in May to August since the 1990s, and, to date, the regulation on the minimum catch size control (255 mm SL) (Liu and Sadovy de Mitcheson, 2008; http://hyyyj.fujian.gov.cn/). Evaluation after the implementation of the series measures is essential. On the basis of the fundamental information on the reproductive dynamics provided by this study, extra measures need to be considered, such as the protection of autumn spawning peak, the earlier regulation for protecting spring spawning peak, the control of fishing gears (e.g., set nets), and the increase of mesh size (1–5 mm are currently commonly used).

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

The animal collection and study was reviewed and approved by Fujian Province Ocean and Fisheries Bureau of China and Xiamen University of China.

LY wrote the first draft and organized sampling trips. LY, YJ, QX, GD, and ML conducted the sample collection. LY, YJ, QX, and XC performed the histological analyses and data analyses. LY, YJ, and ML revised the manuscript. All authors contributed to the article and approved the submitted version.

This work was supported by the Fujian Province Ocean and Fisheries Bureau of China (contract nos. [3500]HTZB[GK]2019007-1-1 and 20200059) and the National Natural Science Foundation of China (grant no. 41976091). The funders had no role in data collection and analysis, decision to publish, or preparation of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.