Mustafizur M. Rahman, Amira F. Samat

aInstitute of Oceanography and Maritime Studies, International Islamic University Malaysia (IIUM), Kg. Cherok Paloh, Kuantan, Pahang, 26160, Malaysia

bDepartment of Marine Science, IIUM, Jalan Sultan Ahmad Shah, Bandar Indera Mahkota, Kuantan, Pahang, 25200, Malaysia

cUniversiti Malaysia Perlis, Sungai Chuchuh, Padang Besar, Perlis, 02100, Malaysia

Keywords:

ABSTRACT Understanding the reproductive biology particularly sexual maturity, reproductive cycle and fecundity of fish is useful for fisheries management. Unfortunately, there is no information about reproductive biology of Nemipterus furcosus. This study addresses this deficit for N. furcosus. We conducted a 12-month study to understand the reproductive biology particularly the sex ratio, sexual maturity, fecundity and reproductive cycle of N. furcosus at the east coast of peninsular Malaysia. Results indicated that ratios of male and female were statistically similar up to 19.0 cm TL (total length). Male outnumbered female above 19.0 cm TL (P <0.05). The fecundity ranged between 54,970 and 236,938 with a mean fecundity of 102,477 ±43,580 (standard deviation). Male and female fish reached first sexual maturation at almost same size, 15.8 and 15.6 cm, respectively. This information is needed to determine the lowest acceptable harvest length of this fish under the management aim to enter all female fish for reproduction minimum one time in their life. We estimated the reproductive period of N. furcosus applying oocytes diameter, histological examination of ovaries and monthly mean GSI. All of these methods gave a consistent result about spawning season, which appeared to be from March to July with peaks in March and June. Based on this result, N. furcosus spawners can be protected during March to July for entering the reproductive stock. Apart from artificial breeding, the results of this study might facilitate future efforts at fine-tuning fishing effort for a proper management of N. furcosus.

1.Introduction

The genus

Nemipterus

encompasses more than 50 nominal species and are distributed all over the Indo-West Pacific region (Russell, 1993).Most species of this genus are normally found near the sandy or muddy bottoms at depths from 5 to 80 m (Kerdgari, Valinassab, Jamili, Fatemi,& Kaymaram, 2009; Russell, 1993). Generally, nemipterids are commercially caught by bottom trawl. They are one of the most commercially important groups of marine fishes particularly in the east China sea (Tonnie et al., 2018). Due to their importance in fisheries,many studies have been conducted regarding various aspects of biology of nemipterids, particularly

N. peronii

(Said, Mohsin, & Ambak, 1994;Wu, Weng, Liu, & Su, 2008),

N. virgatus

(Takahashi, Kobayashi, &Suzuki, 1989),

N. Bathibius

(Granada, Masuda, & Matsuoka, 2004;Takahashi et al., 1989),

N. japonicus

(Raje, 2002, Rajkumar, Rao, &Kingsly, 2003; Manojkumar, 2004; Kerdgari et al., 2009; ElHaweet,2013; Nettely et al., 2016) and

N. randalli

(Erguden, Turan, Gurlek,Yaglioglu, & Gungor, 2010; Innal et al., 2015). Unfortunately, information regarding reproductive biology of

N. furcosus

is very limited(Amira et al., 2016; FAO, 2012, p. 209). Persada, Utami, and Rosalina(2016) conducted a three-month (March–May 2015) study on a few reproductive parameters (sex ratio, gonadosomatic index, facundity,etc.) of

N. furcosus

landed at the Sungailiat archipelago fishery Port,Indonesia.

Nemipterus furcosus

, frequently called forktail threadfin bream, is one of the most popular and economically important food fish in many Asian countries, where it is an important coastal demersal fisheries resource(Amira et al., 2016). It is distributed in numerous Asian and Australian coasts. Due to high demand and market price, it is one of the chosen fishes for Malaysian fisher and therefore, total production of

N. furcosus

on the east coast of Malaysian peninsular is decreasing day by day(Amira et al., 2016; FAO, 2012, p. 209). Presently, it is also a target fish for aquaculture to meet the consumer demand. However, artificial breeding of this fish is not successful yet due to insufficient knowledge regarding reproductive biology especially sexual maturity, fecundity,gonadal change and spawning period.Size at first sexual maturity (SFSM) is applied to estimate the lowest acceptable harvest length of fish to enter female fish for reproduction minimum one time in their life (Rahman, 2017). Understanding fecundity is useful to evaluate the commercial possibilities of fish populations(Tracey, Lyle, & Haddon, 2007). It helps estimating the revival potential of overexploited fish populations (Hughes & Stewart, 2006). Rules and legislations need to be applied to prohibit catching spawners during the reproductive period for maintaining sustainability of the fish production. Therefore, besides artificial breeding, understanding the reproductive parameters of fish is crucial for fishery managers, who usually depend on the SFSM, and the starting and duration of reproduction for fisheries management (Jakobsen, Fogarty, Megrey, & Moksness, 2009;Trindade-Santos & Freire, 2015). The lack of this information in some countries like Malaysia often leads to overfishing and management failure (Alves & Minte-Vera, 2012; Kjesbu, 2016). Unfortunately, information about the reproductive parameters of

N. Furcosus

is scarce,particularly in Asian countries including Malaysia. This study sought to contribute towards addressing the deficit of information by investigating the sex ratio, reproductive cycle, sexual maturity, fecundity and the correlation of fecundity with length and weight of

N. furcosus

for a period of 12 months at the east coast of peninsular Malaysia. The information of this paper can be served as an important reference for the

N. furcosus

management decision making.

2.Materials and methods

2.1.Sampling and measurement

All fish were sampled randomly from fishing vessels, which caught fish at the east coast of peninsular Malaysia (Fig. 1) using trawl nets. We used random sampling method for avoiding biasness in the homogenous group of fish. After arriving the fishing vessels at the landing center, all fish are kept in a pile on the clean concrete made floor for manually sorting them based on size. We sampled a portion (approximately 10–20 kg) of the fish before sorting them out. Fish sampling was conducted monthly over 12 months (Aug 2012 to Jul 2013) to incorporate a full seasonal cycle. A total of 1446 forktail threadfin bream specimens were collected for this research. Sample size in October and November were comparatively low due to very rough coastal weather. After collection,specimens were immediately iced and transported to the Marine Science Laboratory, IIUM. Confirmation of species was done using an identification key (Mohsin & Ambak,

1983)

. Total length (TL) and total body weight (TW) were determined using a centimeter scale and a weighing scale (precision 0.01 g), respectively. Gonad of each fish were collected by dissection and were weighed using the same digital balance. Sex was determined via the observation of gonads, by naked eye or by microscope. Sexual dimorphism was checked by naked eyes. In each month,randomly selected 50% (n =25 (Aug), 20 (Sep), (9) Oct, 4 (Nov), 20(Dec), 17 (Jan), 23 (Feb), 24 (Mar), 23 (Apr), 43 (May), 45 (Jun), 35(Jul) of the total ovary were preserved in 10% phosphate buffered formalin for histological examination and other ovaries (n =24 (Aug),19 (Sep), 9 (Oct), 4 (Nov), 20 (Dec), 16 (Jan), 22 (Feb), 23 (Mar), 23(Apr), 42 (May), 45 (Jun), 34 (Jul)) were preserved in Gilson fluid to estimate oocyte diameter. Gonadosomatic index (GSI), SFSM, fecundity,size of oocyte and histological examination of ovary were estimated as female reproductive parameters, while GSI and SFSM were estimated as male reproductive parameters.

2.2.Estimation of gonadosomatic index, size at first sexual maturity,fecundity and oocyte diameter

GSI was determined applying the formula,

GSI

(%)=×100 with

GW

=gonad weight (g),

BW

=body weight (g). The SFSM was estimated by plotting total length (X axis) versus gonadosomatic index(Y axis) (Ohtomi, Tashiro, Atsuchi, & Kohno, 2003). It was further calculated using the empirical equation, LogL=−0.1189 +0.9157×logL(Binohlan & Froese, 2009) with L=the SFSM, L=the maximum total length.All ovaries preserved in Gilson fluid were used to estimate oocyte diameter. Preserved (in Gilson fluid) ovaries with a GSI ≥2.2% were considered as matured ovaries and were used for fecundity estimation. A total of 104 ovaries were used for fecundity estimation. For this, a subsample of 0.05 g was taken from each of 6 portions (the anterior, middle,and posterior of both lobes) and counted number of eggs under a stereo microscope. The fecundity (

F

) was calculated using the equation,

F

=((

N

×

W

)/

W

)with

N

=number of oocytes in 6 subsamples

, W

=total weight the ovary,

W

=total weight of 6 subsamples.The relationships of fecundity with body length (L) and fecundity (F)with body weight (W) were determined using the equations,

F

=

aL

and

F

=

aW

, respectively. The constants b and

R

were determined using a linear regression on the logarithmic transformed equations

LogF

=log

a

+

b logL

and

LogF

=log

a

+

b logW,

respectively. ‘

Fig. 1.Map showing N. furcosus capture location at the East coast of Peninsular Malaysia.

The size of oocytes was determined using a microscope connected with a digital camera mounted on a computer and software.

2.3.Histological examination of ovary

Four major steps of tissue processing (dehydration, clearing, infiltration, and impregnation and embedding) were done to process ovaries according to Morrison (1990). Firstly, the ovaries were dehydrated by running them with a series of alcohol in increasing concentrations from 50% to 100% and cleared in an automated tissue processor. The cleared tissues were pre-embedded with toluene-paraffin and fully embedded in paraffin wax. The embedded tissues were poured into a mold and cooled on the cold plate. After cooling, they were sectioned at 6 μm thickness by a microtome. The sliced tissues were put on slides and stained with haematoxylin and eosin. The stained tissues were mounted in DPX mountant and observed under a microscope connected with a digital camera. Gonad maturation was categorized (six stages) by histological inspection according to Bancroft and Stevens (1996).

2.4.Data analyses

The significant difference between male and female ratio was determined through the χtest. A one-way ANOVA (significant at

P

≤0.05) performed to find the temporal variation of the GSI. ANOVA was followed by Tukey test if the effect was significant. All data were examined for suitability of parametric test before performing the ANOVA.

3.Results

3.1.Sex ratio

The sex ratio of

N. furcosus

was statistically similar up to 19.0 cm total length (

P

>0.05) (Table 1). Male significantly more than female above 19.0 cm TL (

P

<0.05). The ratio of male and female were statistically different in all months except April and July (Table 1). The ratio of male was higher than female in January, February, March,August, September, October, November and December, whereas theopposite result was observed in May and June. Overall male

N. furcosus

(n =877) was significantly higher (

P

<0.05) in number compared to the female

N. furcosus

(n =569).

Table 1 Ratio of female N. furcosus by size and month during August 2012–July 2013.

3.2.Fecundity

The fecundity varied from 54,970 to 236,938 with an overall mean of 102,477 ±43,580 (standard deviation). Although not reported here,monthly mean fecundity of

N. furcosus

was statistically similar (

P

>0.05). Fecundity (F) of fish increased with total length (TL) and total body weight (W): F =80.1 ×TL(R=0.94,

P

<0.01, n =104) and F =11.2 ×10×W(R=0.95,

P

<0.01, n =104) (Fig. 2).

3.3.Size of oocytes

Fig. 2.Relationship of fecundity with body length (a) and total weight (b)of N. furcosus.

Fig. 3.Monthly mean number (%) of various size oocysts observed in the female fish ovaries during August 2012–July 2013.

The size of oocyte was highly variable and ranged from 0.09 to 1.18 mm in diameter. Fig. 3 represents the variation in the oocyte size over 12 months to incorporate a full seasonal cycle. Large-sized oocytes >0.5 mm in diameter were recorded mostly during January to August, with highest sized group 1.11–1.20 mm during March to July. Small-sized group <0.5 mm was recorded mostly during August to December. The size of oocyte was related with the gonadosomatic index. The size of oocyte from ovaries with an average gonadosomatic index of >2.12 was almost always >0.5 mm in diameter. The size of oocyte from ovaries with an average gonadosomatic index of <1.1 was almost always <0.5 mm in diameter. Once oocytes became ≥0.5 mm in diameter were considered matured (Wu et al., 2008).in the immature stage. A few oogonia were also observed in the maturing stage. The mature stage of ovaries was observed in January and February when most of the ovaries (65–70%, Fig. 5) contained secondary vitellogenic oocyte and post vitellogenic oocyte (Fig. 4c).These are large and mature oocytes (0.51–0.70 mm in diameter).Oogonia were also observed scarcely in the mature stage. After the mature stage, ovaries entered the ripe and running stage during March to July when most of the ovaries (54–61%, Fig. 5) consisted mainly of post vitellogenic oocytes consisting a nucleoli at the center of the nucleus (Fig. 4d). At this stage, more yolk accumulation was observed compared to mature stage. At this stage, most of the oocytes were from 0.91 to 1.20 mm in dimeter. The fish spawn a number of times during this stage. The spent stage of ovaries was observed in August (Fig. 4e). At this stage, most of the ovaries (44%) had also oogonia, which were scattered between secondary vitellogenic oocytes. In addition, ovaries had large empty spaces where post vitellogenic oocytes were located before. Finally, the ovaries entered the resting stage in September(Fig. 4f) during which most of the ovaries (55%, Fig. 5) contained scattered eggs but not as large as very recently spawned ovaries. Large space, and few primary oocytes (atretic oocytes; 0.11–0.20 mm in dimeter) and oogonia (unyolked oocytes; 0.02–0.10 mm in dimeter)were also observed in the resting stage of ovaries.

3.4.Development of oocytes

Gonads of female

N. furcosus

followed six maturity stages: immature,maturing, mature, ripe and running, spent and resting (Fig. 4). Immature ovary was observed from August to December with the highest percentage (67%) in October (Fig. 5). Immature ovary composed mainly of oogonia cells (unyolked oocyte) and primary oocytes (atretic oocytes)(Fig. 4a). Most of the oocytes in the immature ovary were from 0.02 to 0.10 mm in dimeter. The ovaries entered the maturing or developing stage in November and December during which most of the ovaries(50–65%, Fig. 5) contained primary oocytes and primary vitellogenic oocytes (PVO) with yolk granule (Fig. 4b). Oocytes were larger(0.11–0.30 mm in diameter) and more in number compared to oocytes

3.5.Gonadosomatic index

Monthly mean gonadosomatic index of female and male

N. furcosus

is presented in Fig. 7. Female fish had the lowest mean gonadosomatic index in October (Fig. 6a). A sudden increase of female’s gonadosomatic index started after December, reaching a peak in March. Gonadosomatic index then slightly decreased in April and May until another peak in June. Mean gonadosomatic indexes of female fish in September,October, November and December were statistically same (

P

>0.05)and were statistically lower compared to remaining sampling months (

P

<0.01). Monthly mean GSIs of male were similar to female with a few exceptions (Fig. 6b). Male fish did not show any peak GSI in June.

3.6.Sexual maturity

The highest gonadosomatic index was 6.64% in females and 0.97%in males (Fig. 7a). The gonadosomatic index rose sharply at around and above 15.7 cm total length in females. Female fish with a GSI ≥2.2%could approximately be considered as sexually matured females.Therefore, the SFSM was considered to be around 15.7 cm total length in females. Male’s GSI increased sharply at around 15.8 cm total length(Fig. 7b). Therefore, 15.8 cm total length was considered as the SFSM for male

N. furcosus.

Male fish with a gonadosomatic index ≥0.28% could approximately be considered as mature males. The SFSM was calculated as 15.6 cm total length for females and 15.5 cm total length for male using the empirical equation developed by Binohlan and Froese (2009).Using minimum GSI value of mature fish, the number of mature male and female were calculated as 279 (31.8%) and 215 (37.8%),respectively.

4.Discussion

Fig. 4.Microscopic observation of various stages of N. furcosus overies. O, oogonia; PO, primary oocyte; PVO, primary vitellogenic oocyte; SVO, secondary vitellogenic oocyte; PsVO, post vitellogenic oocyte; ES, empty spaces. Stage I (a); stage II (b); c, Stage III (c); stage IV (d); stage V (e); stage VI (f).

Fig. 5.Percent composition of various reproduction stages observed in female N. furcosus during August 2012–July 2013.

This study provides an important contribution towards the primary literature describing sex ratio, fecundity, relationship of fecundity with body length and body weight, size of oocyte, oocytes development,spawning season and sexual maturity of

N. furcosus

. Especially, by sampling across a full one-year cycle in

N. furcosus

fishery and collecting a plethora of biological data, we have delineated their reproductive biology. This information can be discussed according to possible lifehistory requirements for the

N. furcosus

and used to not only progress regional but also broader fisheries management. Analysis of sex ratio indicated that males dominated in large size fishes. This might be due to the low survival of female fish and higher life span of male fish (Granada et al., 2004; Wu et al., 2008; Rajkumar, Rahman, Reni Prabha, & Phukan, 2013). In addition, sexual dimorphism and differences in growth and survival between male and female fish might also be contributing parameters. In the current observation, we did not observe any sexual dimorphism in

N. furcosus

. The growth of male might be greater than female as the growth of female might be slowed at the start of maturity.Slowed growth and higher mortality in large size females have to date been generally accepted possibly due to greater costs associated with spawning (Farley, Williams, Hoyle, Davies, & Nicol, 2013; Rahman,2017). Granada et al. (2004) found higher percentage of males than females in large size

N. bathybius

at the Kagoshima Bay (Japan) due to a higher growth rate of male compared to female. The sex ratio of large size fish may also be related to fecundity (Wu et al., 2008). Some studies found that for those with high fecundity such as the mullet (

Mugil cephalus

), male outnumbered female in large size fish (Chiou & Chen,1993; Su, 1989). Published data comparing the ratio of male and female

N. furcosus

from the east coast of Malaysian peninsular or elsewhere is lacking, although a number of studies were reported on other species of

Nemipterus

genus from the other geographical regions. For example,Granada et al. (2004) observed the sex ratio of

N. bathybius

at the Kagoshima Bay, Japan with males dominating at large-sized fish.

Fig. 6.Monthly variation of mean (±95% CI) gonadosomatic index (%) in female (a) and male (b) N. furcosus population during August 2012–July 2013.Mean with no common letter indicates statistically different (P <0.05) based on Tukey test.

Nemipterus furcosus

also showed a temporal variation in sex ratio with male predominating in most of the sampling months. Almost similar temporal variation in sex ratio was observed in

N. japonicus

by Raje (2002) in Indian coastal water. However, the overall sex ratio observed in this study is comparable with Persada at al. (2016), who observed male

N. furcosus

outnumbered than female

N. furcosus

at the Jakarta Bay, Indonesia.In the present study, the SFSM was estimated using the GSI-TL (total length) plot and the empirical equation by Binohlan and Froese (2009).The size at first sexual maturity of fishes can also be obtained from logistic equation derived from plots of the percentage occurrence of mature females against length class (Wu et al., 2008). However, many studies reported (e.g., Hossain, Ahmed, Islam, Jasmine, & Ohtomi,2010; Hossain & Ohtomi, 2008; Ohtomi et al., 2003) a low accuracy of the logistic equation in estimating the size of sexual maturity of fishes.Nowadays, using the GSI-TL plot to determine the size at first sexual maturity is a common and well-established method. However, in the present study, both GSI-TL plot and the empirical equation gave almost an identical result showing the first maturation of male and female

N. furcosus

at almost same size (female: 15.7 cm (GSI-TL plot), 15.6 cm(empirical equation) and male: 15.8 cm (GSI-TL plot), 15.5 cm (empirical equation)). The observed SFSM of male and female

N. furcosus

cannot be compared to any other studies as no such study has been found in the literature. However, Krishnamurthi (1971) determined 16.5 cm total length as the SFSM of

N. japonicus

in the Andhra-Orissa coast, India.Vivekanandan and James (1986) reported lower length (14.5 cm) at first maturity of

N. japonicus

in the Madras coastal water, India. Maturity of fsh can be differed based on the environmental conditions (Jennings,Kaiser, & Reynolds, 2001; Rahman & Balcombe, 2018; Rahman &Verdegem, 2007). Water temperature, fishing intensity and abundance of food are major parameters that influence maturity of fish altough the information about these variables are not presented in this study.

Fig. 7.Relationship of gonadosomatic index (GSI) with total length of (a) female (n =569) and (b) male (n =877) N. furcosus.

We recorded that the fecundity of

N. furcosus

varied from 54,970 to 236,938 with a mean of 102,477 ±43,580 (standard deviation). In comparison,

N. furcosus

from the Jakarta Bay, Indonesia had a range of fecundity of only 39,728–40,921 (Persada et al., 2016). In coastal water of India, a range of fecundity of 10,260–184,946 for

N. japonicas

was recorded by Raje (2002), while an even higher range of fecundity of 13,758–398,859 for

N. peronii

had been recorded from the waters of southwestern Taiwan by Wu et al. (2008). Thus, the fecundity of fish is highly variable depending on many factors including species, size and maturity stage of fish. However, the documented fecundity of fish may not be comparable among studies since size and maturity stage of fish were different. We used matured fish for a full 12-month cycle, while Persada et al. (2016) used 2–3 years old fish without considering maturity for a period of only three months. Our observed fecundity was fluctuated with the length and weight of fish. A significant relationship was observed for total length and total body weight with fecundity in

N. furcosus

. This concurs with many other studies on

Nemipterus

genus(Manojkumar, 2004; Nettely et al., 2016; Raje, 2002; Wu et al., 2008).The reproductive period of fish is commonly determined by their gonadosomatic index although the study of gonad histology is time intensive and considered a powerful tool for reproductive studies in fish as it can precisely determine their spawning season (Kopf, Davie,Bromhead, & Young, 2012; Zeyl, Oliver, Love, & Higgs, 2014). In the present study, oocytes diameter, histological examination of ovaries and monthly mean GSI were used to determine the spawning season. All these methods gave consistent result about the spawning season of

N. furcosus

. Oocytes diameter measurements showed that the largest sized oocytes 0.91–1.20 mm in diameter were appeared from March to July. Histological examination of ovaries indicated that the ripe and running stage of oocytes was appeared during the same period(March–July). Female fish showed comparatively higher GSI during March to July with two peaks in March and June. Although fish had an increasing trend of GSI during the December–February period, this period should not be included in the main spawning season as gonads were mostly in the maturing stage. Ripe and running stage of gonad was hardly observed at this period. Fish do not spawn when their gonads are in the maturing stage because their ovaries contain mostly primary oocytes (Kokokiris, Stamoulis, Monokrousos, & Doulgeraki, 2014;Nettely et al., 2016). Based on the above observations, the spawning season of

N. furcosus

on the east coast of Malaysian peninsular appears to be from March to July with peaks in March and June. Similar extended spawning season is generally observed in other

Nemipterus

species. For example, Wu et al. (2008) reported the reproductive period of

N. peronii

at southwestern Taiwanese coast from February to July with peaks in April and May. In another study, Granada et al. (2004) determined the reproductive period of

N. bathybius

in Kagoshima Bay, southern Japan from June to October with a peak-spawning period in July to September.In conclusion. the result regarding sexual maturity would be helpful to determine the lowest acceptable harvest length of this species under the management aim to enter female fish for reproduction minimum one time in their life. It would also be useful to guide activities that routinely estimates variations in SFSM associated to various factors including overfishing (Rahman, 2017). It is necessary to continue the observation for understanding the yearly maturation variability that associated with the fish size and the time of the year. Based on this result,

N. furcosus

spawners can be protected during March to July to allow all spawners to spawn minimum one time in their life. The data collected here have implications for optimal fine−scale harvesting of

N. furcosus

for their proper management and also for their artificial breeding. There is further research potential to examine the reproductive period and SFSM at various spatial scale to understand the impact of fisheries upon wild fish populations in various aquatic environments.

CRediT authorship contribution statement

Mustafizur M. Rahman: Methodology, Writing - original draft,Formal analysis, Data curation. Amira F. Samat: Data curation.

Acknowledgements

We would like to thank the International Islamic University Malaysia and the Ministry of Higher Education, Malaysia for the financial assistance through P-RIGS18-032-0032 and FRGS19-096-0705 projects,respectively. No author has any conflict of interest with the content of the manuscript.