Introduction
Blennies (family Blenniidae) are small benthic fish that inhabit restricted areas of high complexity, such as coral and rocky reefs, usually in intertidal and subtidal areas (Orlando-Bonaca & Lipej, Reference Orlando-Bonaca and Lipej2007; Tiralongo et al., Reference Tiralongo, Tibullo, Brundo, Paladini De Mendoza, Melchiorri and Marcelli2016a). The family comprises 58 genera and 387 species found in tropical and temperate seas worldwide (Patzner et al., Reference Patzner, Hastings, Springer, Wirtz, Gonçalves, Patzner, Gonçalves and Kapoor2009; Smith-Vaniz & Rose, Reference Smith-Vaniz and Rose2012). Biologists have long been familiar with the maturation or so-called ripening of the gonads of teleost fishes and the subsequent onset of the spawning period. The reproductive biology of several blennies and their close relatives from the Atlantic Sea has been studied from a broad array of viewpoints, and several papers describe the reproductive behaviour of blennies in their natural habitat (Almada et al., Reference Almada, Oliveira, Barata, Gonçalves and Rito1990; Almada & Santos, Reference Almada and Santos1995) or in artificial structures (Possamai & Fávaro, Reference Possamai and Fávaro2015).
Blennies are generally sexually dimorphic, males are larger than females and during the breeding season, they develop special structures such as ridges on the head or eye cirrus involved in courtship. In addition, the colouration patterns in males frequently change during the reproductive stage (temporary dichromatism) (Gibson, Reference Gibson1969; Tiralongo et al., Reference Tiralongo, Tibullo, Villani, Mancini, Baldacconi, Brundo and Marcelli2016b). Blennies deposit demersal eggs that are stuck to the substrate by an adhesive disc (Santos et al., Reference Santos, Almada, Santos, Blanchard, Brain, Blanchard and Parmigiani1989; Farias et al., Reference Farías, Vargas, Tassara, Carretier, Baize, Melnick and Bataille2010; Ferreira et al., Reference Ferreira, Santos, Reis-Henriques, Vieira and Monteiro2012) or wires (Watson, Reference Watson, Patzner, Gonçalves and Kapoor2009). Adults provide parental care to the developing embryos. Blennies live more than two years and sometimes up to 10 years (e.g. Lipophrys pholis; Dunne, Reference Dunne1977; Carvalho et al., Reference Carvalho, Moreira, Queiroga, Santos and Correia2017). The breeding season spans two months or more in temperate areas, with several spawning events (Qasim, Reference Qasim1956; Santos, Reference Santos1992).
Hypleurochilus fissicornis (Quoy & Gaimard, 1824) is a small fish commonly found in rocky intertidal areas of the Buenos Aires province (Argentina) (Cousseau et al., Reference Cousseau, Figueroa, Díaz de Astarloa, Boschi and Cousseau2004; Ríos et al., Reference Ríos, Irigoyen, Galván and Delpiani2017). It inhabits reefs from the intertidal zone to 5 m depth, within pools, under rocks, in cracks or inhabiting holes formed by clams (Lithophaga patagonica d'Orbigny, 1846). It is commonly observed with half of its body out of a cave or close to it, in intertidal areas during low tide, from where it attacks its prey and quickly goes back to its refuge. Hypleurochilus fissicornis often shares the cave with the ‘torito’ (Bovichthys chilensis Regan 1913) and other intertidal fishes. In summer (January to March), during the breeding season, males show a remarkable development of the orbital cirrus and globose-like ‘anal glands’, placed in front of the anal fin (Cervigón & Bastida, Reference Cervigón and Bastida1974; Irigoyen & Delpiani, Reference Irigoyen, Delpiani, Irigoyen and Galván2010; Ríos et al., Reference Ríos, Irigoyen, Galván and Delpiani2017). Overall, egg masses can be observed in small cavities attached to the substrate or deposited on bivalves (Gehardinger et al., Reference Gerhardinger, Hostim-Silva and Barreiros2006; Irigoyen & Delpiani, Reference Irigoyen, Delpiani, Irigoyen and Galván2010; Delpiani et al., Reference Delpiani, Bruno, Díaz de Astarloa and Acuña2012; Possamai & Fávaro, Reference Possamai and Fávaro2015).
The main objective of this study is to provide information about the reproductive biology of H. fissicornis in the rocky intertidal zone of Buenos Aires province (Argentina) by means of macroscopic and histological analyses of the gonads. Gonad development was described and the spawning season and sex ratio were determined. In addition, fecundity and size at first maturity were estimated.
Materials and methods
Sample collection and measurements
Fish samples were collected monthly from four rocky intertidal areas along the Mar del Plata coast between January and December 2010, by means of a small hand net (length: 30 cm, height: 20 cm, mesh size: 2 mm bar length). Specimens were preserved in dry ice and immediately transported to the laboratory. All individuals were measured for total length (TL) with a digital caliper and weighed for total weight (TW), with a Metler Toledo electronic balance. The gonad mass (GW) and liver mass (LW) were recorded to the nearest 0.001 g. The eviscerated body was also weighed (EW) to the nearest mg. One ovary was assigned for histological study and the other for fecundity analysis. For males, only a gonad was used for histological study. Water temperature and salinity data were registered before fish sampling using a digital thermometer and a Hydrobios refractometer, respectively.
Monthly abundance estimation
The abundance of H. fissicornis was estimated by the capture per unit effort (CPUE) between January and December 2010. The CPUE was defined as the number of individuals caught per hour of sampling and calculated months (ind/Hs). In addition, a Pearson correlation calculation was made between temperature and CPUE.
Histological analysis
Sex and maturity stages were described macroscopically and the maturity stages were determined using the five stage key (I–V) proposed by Brown-Peterson et al. (Reference Brown-Peterson, Wyanski, Saborido-Rey, Macewicz and Lowerre-Barbieri2011), as follows: (I) immature, fish that have never spawned; (II) developing, fish with growing gonads but not ready to spawn; (III) spawning capable, fish with a completely developed gonad that could spawn during the cycle, including the actively spawning subphase for those found spawning; (IV) regressing, cessation of spawning; and (V) regenerating, sexually mature fish that are reproductively inactive. After the macroscopic identification, gonads used in histological analysis were dehydrated in a graded ethanol series, cleared with xylol, infiltrated and embedded with paraffin. Using a rotary microtome (Leitz Wetzlar, 1512) 6–10 sections were cut and stained with Harris's haematoxylin and eosin. Histological classification of ovaries was based on oocyte development stage (Brown-Peterson et al. Reference Brown-Peterson, Wyanski, Saborido-Rey, Macewicz and Lowerre-Barbieri2011). The presence/absence of post-ovulatory follicles and atresia was also assessed. The samples were analysed using an optical microscope (Leitz, Dialux, 20 EB) and an adapted camera linked to TSView 2.0 program (2001).
Reproductive cycle
The annual reproductive cycle was studied using macroscopic and histological maturity stages, the monthly gonadosomatic index (GSI), where GSI = 100 GW EW–1 and the presence of collected nests. Accumulation and depletion of reserves in the liver and muscles were studied by analysing the monthly changes in the hepatosomatic index (HSI), where HSI = 100 LW EW–1 (Fouda et al., Reference Fouda, Hanna and Fouda1993), and in the condition factor (CF), where CF = 100 × (EW/(TL) 3) (Kartas & Quignard, Reference Kartas and Quignard1984).
Oocyte diameter frequency distribution
The ovaries of 63 females (36–75 mm TL) in spawning capable phase were stored in ethanol 70°GL. Two hundred oocytes per ovary were measured, with an ocular micrometer.
Fecundity estimation
Estimates of batch fecundity were obtained using 15 ovaries in spawning capable maturation phase (maturity stage III), stored in ethanol 70°GL after fixation. Ovaries containing new post-ovulatory follicles (POFs) were not included to estimate batch fecundity. The whole ovary was used due to its small size. These ovaries were separated into two pieces. Later, each sample was rehydrated, weighed with analytical balance (±0.0001 g) and yolked oocytes were counted (≥0.5 mm). Consequently, batch fecundity was the product of the mean of hydrate oocytes per unit of ovary gram (Yo/g) and the ovary weight (Ow). To confirm that hydrated oocytes represented the most advanced batch of oocytes, the size frequency method (Hunter et al., 1985) was used on the gonads of two fish. The largest modal group corresponds to the size of the hydrated oocytes observed (Hunter et al., 1985). General linear regression was used to analyse the relationship between batch fecundity and total length.
Estimation of length at first maturity
To estimate length at first maturity, 208 females and 196 males were analysed for maturity phase determination. Individuals were classified as immature (Stage I) or mature (the other phases) and grouped in 5 mm length classes. A logistic model was fitted to the proportion of mature individuals by total length class using the maximum likelihood method (Roa et al., Reference Roa, Ernst and Tapia1999). The length–maturity data were compiled as a binary data set for use in the logistic model:
$${\rm \;}P_{\rm i}{\rm} = {\rm 1/}( { 1 + e^{{\rm a\ + \ b\ L}}} ) {\rm , \;}$$Where:
Pl was proportion mature at length l, a and b parameters to be estimated by maximum likelihood.
−a/b = l0.5 = length at which 50% of the females are mature.
This logistic model was evaluated and the parameters a and b were chosen to maximize the negative of the log likelihood:
$$-{\rm l}( {{\rm a, \;}\,{\rm b}} ) {\rm} = -\,{\rm sum} [ {( {{\rm hl}} ) {\rm ln}( {{\rm Pl}} ) {\rm} + ( {{\rm nl\ }\hbox{-}{\rm \ hl}} ) {\rm ln}( {{\rm 1\ }\hbox{-}{\rm \ Pl}} ) } ] $$Where:
hl was the number of mature individuals at length l, nl was the sample size at length l and Pl was Eq. (1), and where a minute constant term (0.00000001) is added to Pl which does not affect the estimation. This method follows Roa et al. (Reference Roa, Ernst and Tapia1999). The results were tested using the χ2 test (Scherrer, Reference Scherrer1984).
Results
Capture per unit effort (CPUE) and sex ratio
Overall, 435 specimens were examined, of which 212 were males and 223 females. The sex ratio (male: female) was 1:1.05. The relative abundance of H. fissicornis was estimated by the CPUE and plotted against salinity and water temperature (Figure 1). Higher CPUE values were registered in January, February, April, May and November while lower values were registered in June and October in coincidence with low water temperatures. Conversely, water salinity was less variable throughout the study period, a moderate Pearson correlation coefficient (r 2: 0.49) was detected between temperature and CPUE (female A and male B). Salinity presented less variability within the studied months, in contrast to temperature.
Fig. 1. Monthly variation of the captures per unit effort (CPUE) for female (A) and male (B) for Hypleurochilus fissicornis, temperature and salinity. Black triangles: water temperature; grey squares: water salinity; open circles: CPUE.
Reproductive cycle and maturity
During the study period, the following stages of oocyte development were observed: (A) oogonias, (A) primary growth oocyte, (B) cortical alveolus stage, (C) yolked oocytes, (D) hydrated oocytes, and (E–F) post ovulatory follicles (G–H) atretic follicle (Table 1; Figure 2).
Table 1. Descriptions of stages of oocyte development, post-ovulatory follicles and atretic follicles of Hypleurochilus fissicornis
Fig. 2. Stages of oocyte development of Hypleurochilus fissicornis: (A) oogonias (Oo), chromatin nucleolar (Cn) and primary growth oocytes (Pv); (B) cortical alveoli stage oocyte (AC); (C) yolked oocytes (Yg: yolk granule; m: radiata zone; g: granulosa cells; t: theca cells); (D) hydrated oocytes (Ho); (E) FPO 0 (l: lumen; g: granulosa cells; t: theca cells); (F) FPO 1 (l: lumen); (G) FPO 2 (l: lumen); (H) FPO 3; (I) atretic follicle type α; (J) atretic follicle type β.
Five gonad maturity phases were found during this study.
Females (Figure 3)
I. Immature. Very small ovaries, light pink coloured with a thin tunic. At the microscopic level, only oogonias and primary growth oocytes were observed. Few specimens were found in this phase, mainly between February and May (Figure 3A).
II. Developing. Characterized by increased ovarian size and blood vessels becoming more distinct, with primary growth oocytes, cortical alveoli oocyte (Figure 3B) and early yolked oocytes (Figure 3B). Maturation is related to the increase in vitellogenic oocytes from October to July.
III. Spawning capable. Ovaries attained one third of the abdominal cavity. They were violet coloured. Oocytes were noticeable to naked eye, most were yolked oocytes (Figure 3C). Specimens were observed at this phase throughout the breeding season (December–April).
IV. Regressing. Flaccid ovaries displaying prominent blood vessel. Numerous post-ovulatory follicles and germinal vesicle present from December to May (Figure 3D).
V. Regenerating. Small light violet ovaries with thick tunic. Predominantly primary growth oocytes, post-ovulatory and atretic follicles were present from May to October. (Figure 3E).
Fig. 3. Histological sections of gonads of Hypleurochilus fissicornis: (A) early developing female; (B) developing female; (C) spawning-capable female; (D) partially spawned female; (E) post-spawning female; (F) Testis and testicular gland of a male; (G, H) developing male; (I) Spawning-capable male; (J) partially spawned male (Aβ, atretic β; CA, cortical alveolar; CN, chromatin nucleolar; G, testicular gland; GVM, germinal vesicle migration; H, hydrated oocytes; Oc, ovarian cover; Oo, oogonias; PN, perinuclear; POF, post-ovulatory follicle; Sti, spermatid; SZ, spermatozoa; T, testicles; VD, vas deferens; Vit, vitellogenic).
Males
I. Immature. Testes small and transparent observed only at the microscopic level. Primary spermatocyte present with no lumen. Few specimens were found in this phase, mainly from January to June (Figure 3G).
II. Developing. Characterized by increased testicular size, spermatocytes present in the lobe, secondary spermatogonia, primary spermatocyte, and secondary spermatocyte. No sperm in the lumen. Testicular gland present increasing in size, from October to June (Figure 3H).
III. Spawning capable. White coloured fully developed testicles, with spermatocytes occurring in the lumen of the lobes or sperm ducts. All types of cells developed from spermatogenesis can be present. Spermatozoa are visible in the testicular gland. (November–May) (Figure 3I).
IV. Regressing. The testicle looks like a thin band on the testicular gland. Residual presence of sperm in the lumen of the lobes, in the vas deferens and in the testicular gland. Spermatocytes in the periphery (December–May) (Figure 3J).
V. Regenerating. Small testes. No spermatocytes. Lumen of lobule often missing. Small amount of residual spermatozoa occasionally present in lumen of lobules and in the sperm duct (April and from June to October).
The analysis of the distribution of macroscopic gonad maturity stages during an annual cycle showed the occurrence of developing males in the intertidal zone from October to June, with spawning capable individuals in November and May (Figure 4). Developing females largely occurred between October and June with spawning capable female individuals observed between December and April, with peaks in January (Figure 4). The highest values of GSI for females and males were observed between November and April in agreement with the largest number of spawning capable organisms (Figure 5). The lowest values of GSI (0.29–0.35% in males and 0.72–0.82% in females) were observed between July and October. A notable increase both in the GSI (0.91 and 3.8% for males and females, respectively) and the percentage of developing males (82%) and females (98%) occurred in November. The maximum GSI (1.02 and 8.48% in males and females, respectively) was observed in January, when 100% of individuals were spawning capable.
Fig. 4. Monthly relative frequency of gonad phases of females (A), males (B) of Hypleurochilus fissicornis in the Mar del Plata coast. N, number of samples.
Fig. 5. Monthly variation of the gonadosomatic index (GSI) of females and males of Hypleurochilus fissicornis. Males: black squares; females: black rhomboids.
Condition factor values gradually changed monthly, decreasing in winter months and increasing in spring-summer months (Figure 6). The HSI gradually increased from April to July. In December, a few weeks before the onset of spawning, both indices decreased. The maximum HSI values for females (6.8) occurred in June, and the minimum (1.9) in August (Figure 6), while in males, the maximum HSI values (5) occurred in July and the minimum (1.6) in January (Figure 6).
Fig. 6. Monthly variation of the hepatosomatic index (HSI) and condition factor (CF) of females and males of Hypleurochilus fissicornis. (A) Females; (B) males; HSI, black squares; CF, open squares; NF, number of females; NM, number of males.
Frequency distribution of oocyte diameters
The oocyte size distribution of H. fissicornis was tetramodal in the spawning capable phase (Figure 7). The first three groups, between 50 and 700 μm, with a continuous distribution corresponded to different stages of growth from primary growth oocyte phase (smaller than 200 μm) to the phase prior to hydration (between 200 and 600 μm). The latter group, r which ranged from 700–900 μm, was composed of hydrated oocytes.
Fig. 7. Oocyte diameter distribution in spawning-capable phase of Hypleurochilus fissicornis in rocky intertidal of Mar del Plata. Sample N = 63.
Fecundity
The fecundity (BF) varied between 715 and 2800 hydrated oocytes, with an average value of 1657 for a total length of 68.57 mm. The estimated fecundity values showed a slight correlation with the size of females (R 2 = 0.38) (Figure 8).
Fig. 8. Number of hydrated oocytes in relation to fish length in Hypleurochilus fissicornis.
Length at first maturity
L50 values were 53.66 and 55.83 mm TL, for females and males, respectively (Figure 9), the χ2 statistical test did not show any significant difference (χ2 = 0.7616, df = 1, P = 0.368). One hundred per cent maturity was attained at 65 mm TL in females and males. Figure 9 shows the percentage of maturity observed in males and females for each size class, as well as the adjustment sigmoid logistic parametric model employed.
Fig. 9. Proportion of mature individuals observed for each length class of Hypleurochilus fissicornis in 2010. A = females, N = 223, B = males, N = 212.
Discussion
Blennies show an array of accessory structures including a testicular gland, testicular blind pouches, modified spermatic ducts and specialized fin glands that vary greatly among species (Patzner & Seiwald, Reference Patzner and Seiwald1987; Rasotto, Reference Rasotto1995; Neat, Reference Neat2001). The size of the testicular gland relates to the variety of blennies' testes (Lahnsteiner et al., Reference Lahnsteiner, Richtarski and Patzner1990). In H. fissicornis, the testicular gland was larger than the testicle. The spermatids develop within the testicular gland from early stages of maturation in November.
The occurrence of females showing mature oocytes or recent post-ovulatory follicles from December to April, delimit the spawning period of H. fissicornis. The presence of 50% of females showing advanced atretic stages (post-spawning maturity stage) in May, strongly suggested the end of the reproductive period. These results were supported by macroscopic and GSI data. Our results revealed that the breeding season for H. fissicornis spans five months as has been observed for some blennies that inhabit temperate zones, (e.g. Aidablennius sphinx, Carrassón & Bau, Reference Carrassón and Bau2003, Lipophrys pholis, Shackley & King, Reference Shackley and King1977; Carvalho et al., Reference Carvalho, Moreira, Queiroga, Santos and Correia2017 and Istiblennius enosimae, Sunobe et al., 1998).
The hepatosomatic index (HSI) of H. fissicornis was higher in winter for both females and males. Similar observations have been found in other species of blennies, such as Salarias pavo (Podroschko et al., Reference Podroschko, Patzner and Adam1985) and Parablennius parvicornis (Santos, Reference Santos1995). The liver acts as a storage organ for glycogen and lipids. The energy stored in the liver is reused in muscle activity and reproductive effort (Love, Reference Love1979). In some blennies, increase of the GSI is related to the decrease of the HSI, due to the energy of the liver being used in growth and gonad differentiation (Podroschko et al., Reference Podroschko, Patzner and Adam1985; Santos, Reference Santos1995). The same pattern has been observed here. The higher values of CF recorded during late spring and summer, could be explained by the availability of food during this period (Kittlein, Reference Kittlein1991; Albano, Reference Albano2009; Rumbold et al., Reference Rumbold, Nuñez Velázquez and Gancedo2014). The blenny H. fissicornis underwent seasonal shifts in relative abundance in the study area. Although numerous factors could account for this variability, the cause is likely seasonal temperature fluctuations and the thermal preference of the species. Thermal preference has been demonstrated to be important in determining the composition and structure of fish communities in intertidal habitats (Faria & Almada, Reference Faria and Almada2001; Patzner et al., Reference Patzner, Hastings, Springer, Wirtz, Gonçalves, Patzner, Gonçalves and Kapoor2009). For example, CPUE of H. fissicornis, typically considered a temperate species (Patzner et al., Reference Patzner, Hastings, Springer, Wirtz, Gonçalves, Patzner, Gonçalves and Kapoor2009), peaked in late spring, like the congener Hypleurochilus geminatus, in North America (Faria & Almada, Reference Faria and Almada1999; Patzner et al., Reference Patzner, Hastings, Springer, Wirtz, Gonçalves, Patzner, Gonçalves and Kapoor2009). This pattern, present in this genus, might suggest a stable adult component of blenny assemblages with observed variability due to the timing of spawning and subsequent recruitment, which is often linked to temperature (Faria & Almada, Reference Faria and Almada1999).
This study shows that oocyte development in H. fissicornis begins in October and increases in November when the ovary is dominated by oocytes in vitellogenesis and advanced atretic stages (post-spawning) dominate in May. Several species of blennies reproduce in summer, when warm waters facilitate faster egg development (Almada et al., Reference Almada, Gonqalves, Santos and Baptista1994; Carrassón & Baú, Reference Carrassón and Bau2003), as was also recorded for H. fissicornis in the present study.
The release of several batches of eggs at different times during a long spawning season may offer some advantages. Batch spawning avoids the risk of losing a progeny through a negative environmental event, and may also have implications for dispersion with different winds and currents on different days (Oliveira et al., Reference Oliveira, Gonçalves, Ros, Patzner, Gonçalves and Kapoor2009). Histological analysis and the oocyte diameter frequency distribution characterized H. fissicornis as a batch spawner. Some blennies can make up to eight spawns in the same breeding season (Shackley & King, Reference Shackley and King1977). Batch fecundity was on average 1763 vitellogenic/hyaline oocytes in females. Batch fecundity of H. fissicornis showed a slight correlation (R 2 = 0.38) with the size of fish, showing higher fecundity in larger females.
The reproductive cycle and spawning mode described for the present work are in agreement with the findings for other species of blennies (e.g. Alticus momocrus, Aidablennius sphynx, Salaria pavo; Patznet & Lahnsteiner, Reference Patzner, Lahnsteiner, Patzner, Gonçalves and Kapoor2009). The success of these species agrees with their reproductive strategy and anatomical characteristics (accessory glands, anal gland, testicular gland, among other features) of the reproductive system, mainly in males. In addition to the parental care provided by males, this species exhibits eggs and larvae with copious amounts of yolk (Delpiani et al., Reference Delpiani, Bruno, Díaz de Astarloa and Acuña2012).
This work shows how species present different life histories according to their given environment. Reproductive activity has differences in seasonality in relation to latitude (and related to temperature) within a species distribution range. For instance, individuals of H. fissicornis that inhabit warmer habitats at the northern zone (Brazil) reproduce between May and December (Possamai & Fávaro, Reference Possamai and Fávaro2015), while those occurring in temperate waters in the southern zone reproduce between December and April (this work).
Acknowledgements
The authors would like to thank Gustavo Macchi (CONICET-INIDEP) and Marta Estrada (INIDEP), for technical assistance in the histological procedures. We are especially indebted to two anonymous reviewers for their suggestions and corrections, which we believe greatly improved this manuscript.
Financial support
This research was supported by CONICET (Consejo Nacional de Investigaciones Científicas y Técnicas) (Project-FONDO IBOL), UNMdP 15/E525, EXA 577/20 and UNMdP 15/E619, EXA 669/14 grants. S.M.D., D.O.B., were supported by CONICET fellowships.
