Sustainable Exploitation of Dominant Fishes in the Largest Estuary in Southeastern China

Abstract

Globally, marine fisheries have declined under multiple stresses including overfishing, climate change, and habitat degradation. The Min River Estuary, as the largest estuary in southeastern China, has confronted this situation over recent decades. In this study, the dominant species of fish stocks in the Min River Estuary, including Coilia mystus, Cynoglossus abbreviates, Collichthys lucidus, Amblychaeturichthys hexanema, Polydactylus sextarius, Harpodon nehereus, and Secutor ruconius, were evaluated by the length-based Bayesian biomass estimator method (LBB). Outcomes could be grouped into three categories as healthy, showing the lowest exploitation rate (E: 0.31–0.43) and highest relative biomass (B/B_msy: 1.30–1.90), including S. ruconius, C. mystus, and H. nehereus; overfished, with a medium E (0.50–0.58) and B/B_msy (0.68–0.79), including A. hexanema and C. abbreviates; and collapsed, with the highest E (0.89–0.92) and lowest B/B_msy (0.03–0.21), including C. lucidus and P. sextarius. Corresponding imperative countermeasures such as using larger-sized mesh gears and reducing fishing intensity should be deployed according to the current status of each species for sustainable fishery exploitation and fish conservation.

1. Introduction

Worldwide fisheries are severely damaged under overfishing [1], and higher requirements have been put forward for stock management [2]. The reasonable estimation of resource exploitation and reliable stock assessment of target fish species are greatly encouraged in rebuilding the fishery resources. Corresponding methods were developed by the data support such as catch trends, life history, and the ratio of the empirical estimation of biomass in the final year to the unfished resource (B/B₀) as a priori input [3,4,5,6,7].

The length-based Bayesian biomass estimation method (LBB), combining species length frequencies data and Bayesian Monte Carlo Markov chain (MCMC), could rapidly estimate the status of fish exploitation [8,9]. The LBB method rejects the dependency of catch data and life history information but mitigates the subjectivity caused by prior estimation of the remaining biomass in the final year. Most indicators, including estimated relative biomass (B/B₀), length at first capture (L_c), optimum length in the first capture (L_{c_opt}), and the ratio of current relative biomass to the biomass capable of producing maximum sustainable yields (B/B_msy), could be extracted from LBB for fishery management and applied as inputs in other stock assessment models. LBB has a promising application by providing better performance and more reliable results that are closer to the “true” values [8].

China possessed the largest marine fisheries in the world with a total catch of 11.1 million tons in 2017 [10], which accounted for 19.2% of the global capture [11]. However, the catch composition has shifted from traditional large-sized species with high trophic levels to small-sized and low-valued fish species (“feed-grade fish”). These low-valued fishes accounted for 35% of the total catch in China’s exclusive economic zones [12], reflecting a common phenomenon of “fishing down” [13]. The excessive and aimless exploitation without fishery stock assessment resulted in a short-term increase but subsequently a long-term decline in both the yield and the economic value of Chinese coastal fisheries [14].

The Min River Estuary, as the largest estuary in southeastern China, plays a vital role in the local economy, culture, and ecology [15,16,17]. Characterized by a complex and unique ecosystem structure under the synergic effect of discharge dilution, peripheral seawater salinization, and the Taiwan Warm Current [18], the Min River Estuary is highly productive and has been used as spawning grounds, feeding grounds, and natural habitats for more than 222 fish species [19]. The most important economic and dominant species in the estuary showed a high spatio-temporal niche breadth, supporting a longer time of inhabiting and a wider range of distribution [15,20]. Through the top-down control, bottom-up control, or wasp-waist control among different species, variations in economic species abundance would lead to the reaction of the whole fish community structure and even the ecosystem. Recent investigations had discovered that fishery resources in the estuary tended to decrease [21,22,23]. Unfortunately, there were few reports regarding the fish stock assessment in this area, which impeded the application of effective management.

In this study, stocks of dominant fish species in the Min River Estuary were evaluated based on investigations from 2015 to 2017. Fish length frequency data during the three years were applied for LBB modeling, based on which the current status of exploitation in dominant species was evaluated. The results could provide insights into fish stock assessment in the estuary, further contributing to fish conservation and fishery management.

2. Material and Method

2.1. Ethical Statement

All animal experimental procedures used in this study were approved by the Animal Care and Ethics Committee of the Ocean University of China.

2.2. Study Area, Sampling, and Dominant Species

The Min River Estuary, located in the east of Fujian Province at a range of latitude 25°45′00″–26°30′00″ and longitude 119°30′00″–120°00′00″ (Figure 1), is the largest in southeastern China, with an annual discharge of 57.5 billion m³ [24]. Twelve consecutively seasonal surveys were conducted over three years in January, May, August, and November from 2015 to 2017. Fish samples were collected by a commercial fishing vessel through a benthic trawling net with a mesh size of 25 mm. A total of 11 stations were set up (Figure 1), and trawling lasted approximately 30 min per station at an average speed of three knots. The collected fishes were identified to the lowest taxonomic level according to ‘The Fishes of Fujian Province’ [25]. The taxonomy was then revised according to Fishbase (www.fishbase.org) to avoid invalid species, synonyms, and homonyms. Data including abundance, body length, and body weight were measured. Index of relative importance (IRI) was calculated as in Formula (1) [26]:

IRI_i = (N_i + W_i) × F_i

(1)

where N_i signifies the relative abundance of species i in all sampling sites; W_i designates the relative weight of species i out of the total fish weight; and F_i indicates the occurrence of species i out of the number of total stations. Species with IRI > 500 were defined as dominant species [27].

2.3. LBB Modeling

LBB modeling was performed using the “TropFishR” package [28] in R 3.6.1 [29]. The basic assumption in this model was that fish grew continuously and reached the maximum body length at the maximum age, which ensured that somatic growth and natural mortality of adults could be expressed in a ratio [8,30]. In LBB, relative natural mortality (M/K) and relative fishing mortality (F/K) were induced to estimate the ratio of fishing mortality to natural mortality (F/M) and the ratio of current exploited biomass to original stock without fishing (B/B₀). In this way, the number of input parameters can be effectively reduced [8].

The model firstly estimated the asymptotic length (L_∞), the length where 50% of the fish were captured (L_c), M/K, and F/K. However, the practical value of L_∞ was recommended if it was directly available and reliable, to reduce the model uncertainty. The body length of the fish was confirmed by the von Bertalanffy growth function (2) [31]:

$$L_{\mathrm{t}}=L_{\infty }[1-e^{-K(t-{\mathrm{t}}_0)}]$$

(2)

where L_t is the length at age t, L_∞ is the asymptotic length, K is the growth rate (year⁻¹), t₀ is the age of fish at zero length, and t is the age of the fish.

Parameters L_∞, L_c, M/K, and F/K were estimated by LBB following a series of equations [8]. With the input of L_∞, L_c, M/K, and F/K, the length corresponding to the maximum unexploited cohort biomass (L_opt) was measured by Equation (3) [32]:

$$L_{\mathrm{o}\mathrm{p}\mathrm{t}}=L_{\infty }(\raisebox{1ex}{$3$}\!\left/ \!\raisebox{-1ex}{$3+M/K$}\right.)$$

(3)

The optimum length in first capture (L_{c_opt}) was obtained by function (4) [33]:

$$L_{\mathrm{c}\_\mathrm{o}\mathrm{p}\mathrm{t}}=\raisebox{1ex}{$L_{\infty }(2+3F/M)$}\!\left/ \!\raisebox{-1ex}{$(1+F/M)(3+M/K)$}\right.$$

(4)

The yield per recruit (Y′/R) could be given as Equation (5) [34]:

$$Y\prime /R=\raisebox{1ex}{$F/M$}\!\left/ \!\raisebox{-1ex}{$(1+F/M)$}\right.{(1-L_{\mathrm{c}}/L_{\infty })}^{M/K}(1-\frac{3(1-L_{\mathrm{c}}/L_{\infty })}{1+(\frac{1}{M/K+F/K})}+\frac{3{(1-L_{\mathrm{c}}/L_{\infty })}^2}{1+(\frac{2}{M/K+F/K})}+\frac{{(1-L_{\mathrm{c}}/L_{\infty })}^3}{1+(\frac{3}{M/K+F/K})})$$

(5)

The catch per unit of effort (CPUE’/R) was given by Equation (6) [34]:

$$\mathit{CPUE}\prime /=\raisebox{1ex}{$Y’/R$}\!\left/ \!\raisebox{-1ex}{$F/M$}\right.=\frac{1}{1+F/M}{(1-L_{\mathrm{c}}/L_{\infty })}^{M/K}(1-\frac{3(1-L_{\mathrm{c}}/L_{\infty })}{1+(\frac{1}{M/K+F/K})}+\frac{3(1-L_{\mathrm{c}}/L_{\infty }{)}^2}{1+(\frac{2}{M/K+F/K})}+\frac{{(1-L_{\mathrm{c}}/L_{\infty })}^3}{1+(\frac{3}{M/K+F/K})})$$

(6)

The relative biomass in the exploited phase of the population without fishing activity (B_0’/R) was given by Equation (7) [34] (B_0′> L_c):

$$B_{0\prime }/R={(1-L_{\mathrm{c}}/L_{\infty })}^{M/K}(1-\frac{3(1-L_c/L_{\infty })}{1+\frac{1}{M/K}}+\frac{3(1-L_c/L_{\infty }{)}^2}{1+\frac{2}{M/K}}+\frac{{(1-L_c/L_{\infty })}^3}{1+\frac{3}{M/K}})$$

(7)

Finally, B/B₀ was estimated via Equation (8) [34]:

$$B/B_0=\raisebox{1ex}{$\mathit{CPUE}’/R$}\!\left/ \!\raisebox{-1ex}{$B_{0’}/R$}\right.$$

(8)

When re-running these equations with F = M and L_c = L_{c_opt}, the relative biomass capable of producing maximum sustainable yields (B_msy/B₀) was estimated. The current biomass relative to the biomass capable of producing maximum sustainable yields (B/B_msy) was obtained via Equation (9) [8]:

$$B/B_{\mathrm{msy}}=\raisebox{1ex}{$B/B_0$}\!\left/ \!\raisebox{-1ex}{$B_{\mathrm{msy}}/B_0$}\right.$$

(9)

3. Results

3.1. Relative Importance and Length Frequency of the Fishes

Species with IRI > 500 in at least two seasons in the Min River Estuary were identified (

Table 1. The index of relative importance (IRI) for seven dominant fish species in the Min River Estuary in different seasons in 2015.

Species	Spring	Summer	Autumn	Winter
Coilia mystus	516.0	6.9	506.2	6086.6
Collichthys lucidus	709.5	5.7	579.1	3693.1
Amblychaeturichthys hexanema	867.5	12.0	703.4	506.2
Secutor ruconius	1866.3	831.1	224.1	231.2
Cynoglossus abbreviatus	1440.1	223.8	334.3	1878.5
Polydactylus sextarius	-	6502.4	3208.8	-
Harpodon nehereus	155.2	2222.5	4211.9	2413.6

Note: Values in bold and italic represent the dominant species with IRI > 500; the symbol “-” means absence of corresponding species.

), including Osbeck’s grenadier anchovy Coilia mystus (Feng-ji in Chinese), three-lined tongue sole Cynoglossus abbreviatus (Duan-she-ta in Chinese), big head croaker Collichthys lucidus (Ji-tou-mei-tong-yu in Chinese), pinkgray goby Amblychaeturichthys hexanema (Liu-si-dun-wei-xia-hu-yu in Chinese), blackspot threadfin Polydactylus sextarius (Liu-zhi-ma-ba in Chinese), Bombay duck Harpodon nehereus (Long-tou-yu in Chinese), and deep pugnose ponyfish Secutor ruconius (Lu-ban-yang-kou-bi in Chinese). The highest IRI was 6502.4 of species P. sextarius in summer, followed by 6086.6 of species C. mystus in winter.

The probability density curve of total body length was unimodal for all fishes except S. ruconius. The widest range of body length (72.0–350.0 mm) was determined in C. abbreviatus, while the minimum range (21.0–107.0 mm) was in S. ruconius (Figure 2, Supplementary Table S1).

3.2. Current Exploitation Status of the Fishes

The fishes were classified into three categories of exploitation status. Species S. ruconius, C. mystus, and H. nehereus were grouped in healthy; A. hexanema and C. abbreviates were in overfished; and C. lucidus and P. sextarius were in collapsed (Figure 3). Specific indices regarding the exploitation are listed in

Table 2. Descriptive variables concerning the exploitation of seven dominant fish species in the Min River Estuary in 2015–2017.

Variables	Coilia mystus	Collichthys lucidus	Amblychaeturichthys hexanema	Secutor ruconius	Cynoglossus abbreviatus	Polydactylus sextarius	Harpodon nehereus
quantity	2104	1355	1175	968	897	805	487
length range	47–235	22–205	36–190	21–107	72–350	29–284	50–350
L∞ by user (mm)	234.0 [16,35]	213.0 [36]	174.0 [37]	101.0 [38]	329.0 [16]	300.0 [39]	323.0 [16,21]
L∞ by LBB (mm)	239.0	214.0	180.0	104.0	323.0	305.0	314.0
Lc (mm)	147.6	174.0	83.4	72.0	117.6	65.0	156.0
M/K	1.71	1.42	1.50	2.07	1.10	1.30	1.27
F/K	1.30	17.00	2.10	0.95	1.10	11.00	0.67
Z/K	3.01	18.40	3.60	3.02	2.20	12.30	1.94
F/M	0.78	12.00	1.40	0.48	0.98	7.40	0.55
E (i.e., F/Z)	0.43	0.92	0.58	0.31	0.50	0.89	0.35
Lmean (mm)	168.0	185.0	107.0	77.5	181.0	127.0	202.0
Lopt (mm)	152.0	144.0	119.0	62.0	235.0	205.0	222.0
Lmean/Lopt	1.10	1.30	0.88	1.30	0.72	0.47	0.89
Lc_opt (mm)	123.0	140.0	103.0	48.0	196.0	197.0	174.0
Lc/Lc_opt	1.20	1.20	0.81	1.50	0.60	0.33	0.90
B/B0	0.47	0.08	0.25	0.64	0.30	0.01	0.51
B/B0 (when F = M)	0.357	0.367	0.364	0.345	0.383	0.368	0.377
B/Bmsy	1.30	0.21	0.68	1.90	0.79	0.03	1.40
Y′/R	0.0300	0.0180	0.0440	0.0140	0.0640	0.0007	0.0480
status [40]	healthy	collapsed	overfished	healthy	overfished	collapsed	healthy

Notes: L_∞, asymptotic length; L_c, length at first capture; M, natural mortality; K, growth rate; F, fishing mortality; Z, total mortality; E, exploitation rate; L_mean, the mean length of exploited stock; L_opt, the length where the unexploited cohort biomass is maximum; L_{c_opt}, the optimum length in first capture; B, current exploited biomass; B₀, unfished biomass; B_msy, the biomass capable of producing maximum sustainable yields; Y′/R, yield per recruit.

. Species S. ruconius showed the lowest exploitation rate (0.31), the highest L_mean/L_opt (1.30) and L_c/L_{c_opt} (1.50), the highest B/B₀ (0.64) and B/B_msy (1.90), but the lowest F/M (0.48). Species C. mystus and H. nehereus showed a lower exploitation rate (0.53 and 0.35, respectively) and fishing mortality (0.78 and 0.55, respectively), but higher B/B₀ (0.47 and 0.51, respectively), B/B_msy (1.30 and 1.40, respectively), L_mean/L_opt (1.10 and 0.89, respectively), and L_c/L_{c_opt} (1.20 and 0.90, respectively). Species A. hexanema and C. abbreviatus were detected as slightly overfished in view of the E values (0.50 and 0.58, respectively) and B/B_msy (0.68 and 0.79, respectively). Species C. lucidus and P. sextarius had extremely high E values close to “1” (0.92 and 0.89, respectively) and the lowest B/B_msy (0.21 and 0.03, respectively).

Three species, C. mystus, C. abbreviatus, and P. sextarius, were selected as the surrogates of these three categories, i.e., healthy, overfished, and collapsed, to demonstrate the modeling procedure of LBB (Figure 4). The LBB model was well fitted to all three fishes with Z/K values of 3.01, 2.20, and 12.30, respectively.

4. Discussion

In recent years, the dramatic decrease in coastal fish resources and significant changes in catch composition in China caused by continuously increasing fishing pressure have attracted a lot of concerns [12,41,42]. For instance, a drastic decline was reported in CPUE in 2011 down to 0.86% of 1959 in the Bohai Sea [12,43] and to 53% of 1991 in the Yellow Sea and the East China Sea in 2000 [44]. Meanwhile, local high-valued and large-sized catches disappeared while small fishes with lower trophic levels and minor economic value became dominant. For example, from 1991 to 2000, dominant fish species in the northern East China Sea changed from black scraper Thamnaconus modestus, red bigeye Priacanthus macracanthus, and silver croaker Pennahia argentata to small-sized Indian perch Jaydia lineata, Snyder’s gaper Champsodon snyderi, and skinnycheek lanterfish Benthosema pterotum. Besides, overfishing also contributed to reducing the proportion of elder fishes in the catch and the average weight of individuals [44], suggesting trends of miniaturization and early maturing of the species. For example, in the Yellow Sea, the average first maturity length of small yellow croaker Larimichthys polyactis dramatically decreased from 152.8 in 1960 to 105.3 mm in 2003, accompanying the population collapse [45].

When L_c > L₅₀ (first maturity length, where 50% of the individuals are mature), the fishing gears give up catching immature individuals, leaving the fish a chance to be mature and reproduce at least once before being caught. This is beneficial to the healthy and sustainable development of fish stocks [30]. In the Min River Estuary, the fishing mortalities of species S. ruconius, C. mystus, and H. nehereus were all lower than the respective natural mortalities, and exploitation rates (E) were relatively low. Besides, L_mean/L_opt and L_c/L_{c_opt} were close to or even bigger than 1, suggesting little harm from the commercial fishing activity on the juveniles and the supplement was abundant for the population. These three species were relatively healthy in terms of natural resources stock and possessing potential economic value for further exploitation and utilization.

Though the L₅₀ for C. mystus was unavailable, mature female individuals with a body length of less than 71.7 mm were discovered in the Yangtze River Estuary [46], much smaller than its L_c (147.6 mm) in the Min River Estuary. Therefore, a considerable proportion of mature female individuals in the Min River Estuary presumably were not caught, leaving opportunities for the recovery of the population. Benefitting from its largest spatio-temporal niche breadth in the Min River Estuary, C. mystus showed the strongest competitive ability, the widest distribution [20], and higher growth rate (K = 1.30) to ensure the strong resilience of the population [22]. It was reported that the stock of C. mystus in the Min River Estuary significantly increased from 1990 to 2006, as well as its dominance in the community [47].

The L₅₀ of S. ruconius at 65.3 mm [38] was smaller than the L_c at 72.0 mm in this study. Similar to the status of C. mystus, the stock of S. ruconius in the Min River Estuary had steadily increased in recent years [22]. As a small economic fish feeding on plankton, S. ruconius had expanded its population, accounted by the decreasing prey pressure caused by the decline in large species and support of sufficient food.

For H. nehereus, L_c (156 mm) was shorter than L₅₀ (200–220 mm) [48,49,50,51], causing difficulties in leaving enough mature individuals out of the catch. However, as a ferocious, cannibalistic, and highly active hunter [48,50], with a rapid growth rate (K = 0.94) [21] and strong resistance to external interference, H. nehereus had expanded its population in the East China Sea in recent decades when the traditional commercial species population declined [52]. Besides, in the East China Sea, the peak spawning season for H. nehereus from July to September [53] totally fell into the official “closed fishing season”, ensuring its population supplement [21]. Even so, the fact that its L_c in current exploitation was still too small should be taken into cautious consideration in the long-term management for sustainable exploitation.

For A. hexanema and C. abbreviatus, if only concerning the F/M and E, the ratio of fishing mortality to natural mortality was close to “1” and the exploitation rates were around “0.5”, both indicating an acceptable state of exploitation and no need of human countermeasures for extra resource protection [54]. However, their stocks were unsustainable when taking L_mean/L_opt and L_c/L_{c_opt} into consideration. L_mean and L_c were smaller than L_opt and L_{c_opt}, respectively, which indicated that too many immature individuals were included in commercial catches. No limitation on fish size in fishing would further reduce the fish biomass [55].

As estuarine benthic species, A. hexanema and C. abbreviatus showed a wide niche breadth and high competitiveness in the Min River Estuary [20]. However, in recent years, the extensive use of bottom trawls seriously damaged the population, as well as the seabed environment [56]. Responsive management strategies should be implemented immediately before the population collapse of these species. Based on the results from the LBB analysis, L_{c_opt} values for these two species (103 and 196 mm, respectively) were determined as appropriate sizes of the first-capture individuals.

There was no doubt that the worst conditions had happened to C. lucidus and P. sextarius. Fishing mortalities for these two species were much higher than their natural mortalities, reflecting drastic declines in fish stocks (B/B₀ = 0.08 and 0.01, respectively). Unfortunately, C. lucidus and P. sextarius showed relatively smaller growth rates [22] and weak interspecific competitiveness [20]. In addition, C. lucidus was one of the most favorable aquatic products for coastal residents [57], which further aggravated the loss of its population. Catches of C. lucidus in Fujian Province recorded by the Chinese Fishery Statistical Yearbook had declined from 30 thousand tons in 2005 to 21 thousand tons in 2017 [10], accompanied by a trend of species miniaturization [22]. The present extensive capturing capability of the fishing gears with small mesh sizes and the overloaded fishing efforts not only captured a large number of immature individuals [12,58] but also ruined the fish habitats [14]. Under this scenario, urgent and effective countermeasures should be deployed so as to prevent further resource degradation or even species extinction in the Min River Estuary. To relieve the pressure put on the stocks, the most basic but very effective step by using larger-sized mesh gears should be suggested to allow individuals to reproduce at least once before being caught. Furthermore, more specific and delicate measures including closed fishing systems (i.e., “summer closed fishing”) and establishment of protected areas should be effectively and strictly executed.

5. Conclusions

In this paper, LBB was used to evaluate the exploitation status of seven dominant fish species in the Min River Estuary. LBB successfully classified these fishes into three categories: healthy stocks, including S. ruconius, C. mystus, and H. nehereus, with B/B_msy between 1.30 and 1.90; overfished stocks, including A. hexanema and C. abbreviates, with B/B_msy between 0.68 and 0.79; and collapsed stocks, including C. lucidus and P. sextarius, with B/B_msy between 0.03 and 0.21. The L_{c_opt} estimated by LBB provided references for the exploitation and stock recovery of these species. Correspondingly, management measures, such as reducing fishing intensity, cutting down the number of fishing boats, and limiting fishing time, should be considered for fisheries sustainability.