Interannual variations in the hatching pattern, larval growth and otolith size of a sand-dwelling fish from central Chile
- 689 Downloads
The interannual variation (2010–2013) of larval abundance, growth and hatching patterns of the Chilean sand stargazer Sindoscopus australis (Pisces: Dactyloscopidae) was investigated through otolith microstructure analysis from samples collected nearshore (<500 m from shore) during austral late winter-early spring off El Quisco bay, central Chile. In the studied period, the abundance of larval stages in the plankton samples varied from 2.2 to 259.3 ind. 1000 m−3; larval abundance was similar between 2010 and 2011, and between 2012 and 2013, but increased significantly from 2011 to 2012. The estimated growth rates increased twice, from 0.09 to 0.21 mm day−1, between 2011 and 2013. Additionally, otolith size (radius, perimeter and area), related to body length of larvae, significantly decreased from 2010 to 2012, but increases significantly in 2013. Although the mean values of microincrement widths of sagitta otoliths were similar between 2010 and 2011 (around 0.6–0.7 μm), the interindividual variability increases in 2011 and 2013, suggesting large environmental variability experienced by larvae during these years. Finally, the hatching pattern of S. australis changed significantly from semi-lunar to lunar cycle after 2012.
KeywordsSagitta Chile Ichthyoplankton Dactyloscopidae
Several species of marine fish exhibit beach-spawning behaviour, including some silversides (Atherinopsidae), certain killifish (Fundulidae), a puffer (Tetraodontidae), a few smelts (Osmeridae) and a stickleback (Gasterosteidae). Placement of eggs at higher levels in the intertidal zone seems to be a risky behaviour, but it has numerous advantages, including increased incubation temperatures, increased oxygen availability and reduced aquatic predation (Martin and Swiderski 2001). In these species, hatching may be negatively correlated with the onshore wind-induced wave action which disturbed the beach (Frank and Leggett 1981; Griem and Martin 2000). On the other hand, several species that laid eggs in subtidal sand or rocky bottom show lunar or semi-lunar spawning cycles (Hay 1990; Contreras et al. 2013; Palacios-Fuentes et al. 2014). This tendency is assumed to be related to the timing of larval hatching: benthic eggs have considerably narrow diel hatching windows, and those windows coincide with tides (i.e. during neap and/or spring tides) that are appropriate for dispersal by the strongest tidal currents only during 2 short periods in each lunar month (Yamahira 1997).
A group of fishes that inhabits sandy nearshore or estuarine environments in warm temperate to tropical waters from North and South America are the sand stargazers of the family Dactyloscopidae (Dawson 1982; Doyle 1998). This family is composed of 9 genera, with 48 species that present elongated bodies, eyes on top of the head and cryptic coloration (Hastings and Springer 2009). They frequently bury themselves in sand bottoms, similar to some trachinoids, with only the mouth and eyes exposed, making it very difficult to detect them visually, since their coloration is very similar to that of the adjacent background (Wagner et al. 1976). However, unlike virtually all other teleosts, which normally pump water over the gills by alternately expanding and contracting the buccal and opercular cavities, they have evolved a branchiostegal pump that replaces the opercular pump. Fingerlike labial and opercular fimbriae probably function to prevent particles from clogging the branchial chamber (Nelson 2006).
Sand stargazers exhibit internal fertilization as males of most species have copulatory structures derived from anterior anal fin rays. They are oviparous, laying eggs of approximately 1 mm in diameter attached to each other by filaments; they have parental care, and most males carry and incubate the eggs (Watson 1996; Teixeira et al. 2013). In the latter case, eggs are held in place between the large pectoral fins and the recurved spines in the anterior anal fin of males. Little else is known of this behaviour including how eggs are transferred from the female to the male, the number of eggs and number of clutches that are carried by males and consequently how many females mate with single males (Hastings and Petersen 2010). Hatching seems to occur around 4 mm in the species from the Gulf of California and the California region (Brogan 1992; Watson 1996) and from the Humboldt current (Herrera et al. 2007), and less than 3.6 mm in the species from Belize, Central America (Cavalluzzi 1997).
The members of the family Dactyloscopidae have worldwide distribution, being the southernmost distribution in the Pacific Ocean is the Chilean sand stargazer Sindoscopus australis (Fowler and Bean 1923). It is the only Dactyloscopid that inhabit cold temperate waters and is known only from coastal waters of Chile at depths of 2 m or less about 23°–36°S (Dawson 1977). Larvae are pelagic occurring in nearshore waters, and undergo notochord flexion at larger sizes compared to those of other Dactyloscopids, between 8 and 9 versus 6 mm of body length (Watson 1996; Herrera et al. 2007). Also, transformation (i.e. migration of eyes to the upper position of the head) occurs at larger sizes (Herrera et al. 2007), suggesting large pelagic duration of S. australis in relation to other Dactyloscopids; however, there is no previous information about this subject for any Dactyloscopid fish, nor about growth rates and/or hatch dates.
A useful tool to reveal early life history traits of marine fishes is through otolith microstructure analysis, because otoliths record not only age and growth patterns, but also hatching times, settlement, metamorphosis, migration and condition (Victor 1987; Folkvord et al. 1997; Landaeta et al. 2012; Plaza et al. 2013; Zenteno et al. 2014). Through diel variations in calcium and protein deposition, bipartite structures known as daily growth increments often form at microstructural level of the otolith (Campana 1984). Also, the correlation between otolith size and somatic size enables growth trajectories and histories to be back-calculated (Campana 1990; Takasuka et al. 2008).
The aim of this study was to estimate larval age and growth rates, as well as hatch date patterns of the Chilean sand stargazer (S. australis). For these purposes, the morphometrics and microstructure analysis of sagitta otoliths of S. australis were carried out throughout larval development on an interannual basis in nearshore waters of central Chile. The daily periodicity of the growth increments for S. australis has not been validated, nor for any Dactyloscopidae; however, several species of the suborder Blennioidei has been validated for daily periodicity in coastal waters of central Chile and elsewhere (Hernández-Miranda et al. 2009; Kohn and Clements 2011; Mansur et al. 2013).
Materials and methods
Summary of dates, sampling effort, number of hauls with positive presence of larval Sindoscopus australis and larval abundance (ind. 1000 m−3) for all the study period, 2010–2013
No. of hauls
No. of positive hauls
The left and right sagitta otoliths were removed from well-preserved 174 larvae (3.93–18.21 mm SL) (Fig. 1b, c). The otoliths were embedded in epoxy resin on a glass slide. The daily age was estimated by counting the number of otolith increments with a Motic BA310 light microscope at 1000× magnification under oil immersion. The longest radius of a sagitta was measured three times, and the average was used. The perimeters and areas of the otoliths were then measured once using the Moticam Image Plus 2.0 software.
Three independent counts were performed on both the right and left sagittae. The counts were performed after a prominent hatch mark (HM, Fig. 1b). When the coefficient of variation (CV = standard deviation/mean × 100) of the increment counts among the three readings was <10 %, the arithmetic mean of the three counts was calculated and utilized for the analysis. When the CV was >10 %, the otolith reading was discarded (n2010 = 2; n2011 = 4; n2012 = 5; n2013 = 1). Once selection of the values was done, comparison of readings was carried out with a Wilcoxon matched-pairs test, testing the hypothesis that reading of the left sagitta is the same that right sagitta. Because null hypothesis of the same lecture in both otoliths cannot be rejected (W = 71.5; P = 0.512), we assume that both otoliths can be utilized for analyses.
Microincrement widths were measured for the older/larger individuals from each sampling year. For these individuals, each microincrement was measured three times and the average was utilized for further analysis.
Larval abundance was compared among years with nonparametric tests (Kruskal–Wallis H test) because the data departed for a normal distribution (Shapiro–Wilk test, W = 0.682, P < 0.001). Post hoc comparisons were made with mean ranks for all groups.
Least-squares linear regression analyses were performed between the sagittae morphometry (radius and perimeter) and body length, and exponential models were fitted between the sagitta area and larval length. Linear models among years (2010–2013) were compared with a one-way ANCOVA and multiple slope test (Zar 1999). For comparisons of the relationship between otolith area and larval length among years, data were ln-transformed, and then, linear regression models were fitted. Additionally, linear regressions between the microincrement counts (age) and larval lengths were adjusted. In this case, the slope corresponded to the population growth rate, and the intercept corresponded to the estimated hatch size. To compare the population growth rate during 2010–2013, the slopes were compared with one-way ANCOVA (Zar 1999), utilizing the same size range for all years (from 4.76 to 17.38 mm BL).
Comparisons in the microincrement width (MIW) between individuals and among days were carried out with repeated-measures ANOVA (RM ANOVA). The analysis was a mixed two-factor design, with year as an independent factor and increment width at the third, 10th, 20th and 40th days as a repeated levels (n = 14 for 40th day; for the rest, n = 20). All RM ANOVAs were achieved on balanced set of levels over age.
The hatching days were back-calculated and related to the lunar cycle. In order to compare distribution patterns, hatch day frequencies for each year (2010–2013) were converted to Julian days and then converted to angles (°) by dividing by 365 and then multiplying by 360° so that the data could be analysed using circular statistics. Similarly, for each sampling date, the days since the new moon were counted (DNM), and thereby assigned DNM values from 0 to 29 for each date, in which 0 represented the new moon. The DNM values were converted to angles (°) by dividing by 29 (the length, in days, of the lunar cycle) and then multiplying by 360°. To assess whether the hatching events showed lunar periodicity, the data were analysed using the Rao’s spacing test (Batschelet 1981). The Rao’s spacing test is more powerful and robust than many other circular goodness-of-fit tests and is able to analyse bi- and multimodal distributions, whereas other tests cannot, such as the Rayleigh test and Watson’s U2 (Bergin 1991). The Rao’s spacing test is robust even for small sample sizes but also shows a low frequency of type I errors when analysing data that display no pattern. The null hypothesis that the hatching events would be equally or randomly spaced throughout the lunar cycle was tested for each data set. Finally, comparisons between the annual and lunar hatching distributions among years were conducted using the nonparametric Mardia–Watson–Wheeler test (W) for equal distributions (Mardia 1972).
Temporal variability of larval abundance
Otolith morphometry and its interannual variation
Linear regression models between body length (BL, mm) and sagitta otolith size parameters
R versus BL
P versus BL
ln A versus BL
R versus BL
P versus BL
ln A versus BL
R versus BL
P versus BL
ln A versus BL
R versus BL
P versus BL
ln A versus BL
Interannual variability in larval growth rates
Linear regression models between microincrement counts (age, in days) and body length (mm)
Hatch size (mm)
Growth rate (mm day−1)
Microincrement width variations
Hatch date distributions
Summary of statistic results (W, Mardia–Watson–Wheeler test for equal distribution) of comparison of the hatching day distributions of Sindoscopus australis during 2010–2013 off central Chile
The early life traits of the sand stargazer S. australis, namely hatch patterns, larval abundance and growth rates, showed large variability at interannual scale off central Chile. During 2010–2011, sand stargazer showed a semi-lunar hatching pattern, slow larval growth rates and low larval abundance. The following year 2012, larval hatch occurred uniformly throughout the lunar month, and the larval abundance and its growth increased twofold, but it sagitta otoliths were the smallest of the time series. Finally, in 2013, hatch patterns changed to a lunar cycle with its peak near new moon (i.e. spring tides), the otolith increased it size (radius, perimeter and area), as well as the larval growth rates.
During this study, larval abundances were estimated for the upper 30 m depth of the water column during night sampling in nearshore waters. Larval sand stargazers are scarcely or completely absent in deeper waters of the shelf and slope off central Chile (Landaeta et al. 2008), but they are collected in low abundance during daytime off central Chile (7.96 ± 4.97 ind. 1000 m−3, Landaeta et al. unpublished results). Therefore, it is plausible that night sampling in nearshore waters may be a good predictor of larval S. australis abundance.
Semi-lunar and lunar hatching patterns have been recently described for other criptobenthic marine fishes from central Chile, such as clingfishes (Gobiesox marmoratus, Sicyases sanguineus, Contreras et al. 2013) and Chilean triplefin blennies (Helcogrammoides chilensis, Palacios-Fuentes et al. 2014). A hatching pattern related to first and/or last quarter moon (estimated for 2010 and 2011 for S. australis) is associated with neap tides, which implicate the reduced effects of tidal export from coastal waters to offshore (Robertson et al. 1990). Later, during 2013, the hatching pattern was more frequent near new moon, i.e. during spring tides, which may increase dispersion and population connectivity. This different temporal hatching pattern may be forced by some environmental stress. Similarly, interannual variations in the hatching pattern during the lunar month have been described in rockfish from Japan, which hatch around the new and full moon (Plaza et al. 2003).
Large variations in the estimated larval growth rates as well as larval density occurred among years. This result may be due to differences in the back-calculated hatch distribution, around winter in 2010 and 2011, and during late winter-early spring 2012 and 2013, which can trigger differences in the main growth controllers (i.e. temperature and productivity). Because all years the sampling design was the same (oblique tows from 20 m depth to surface), interannual variations in larval density may obey to a combination of differences in larval survival and meteorological/oceanographic forcing. At El Quisco bay, it occurs an increase in seawater temperature and chlorophyll concentration in subsurface waters from austral winter to spring (Hernández-Miranda et al. 2003), which may partially explain the detected differences between 2011 (large otoliths, slow growth) and 2012 (small otoliths, fast growth), but it does not explain differences in larval traits from individuals hatched during 2012 and 2013 (large otoliths, fast growth).
Uncoupling between otolith and somatic growth rates may occur (Hare and Cowen 1995). The effects of somatic growth rates on the otolith and somatic size relationships (the ‘growth effect’), if it exists, could be a possible source of bias in the back-calculation and somatic size estimation procedures (Takasuka et al. 2008). Also, slow-growing fish tend to have larger otoliths than fast-growing ones of the same length (Wright et al. 1990; Francis et al. 1993). On the other hand, Munk (1993), Folkvord et al. (2000) found no differences in the otolith size–fish size relationship estimated for larvae characterized by different growth rates. Fey (2001) suggested that in hyperoptimal temperatures or conditions of significant temperature fluctuations, otolith growth and increment widths are related more closely to temperature-affected metabolic rate than to somatic growth. Then, during the studied period, it could be occurred a period with uncoupled growth of otolith and soma in larval S. australis (2010–2011) and a period of coupling between otolith and somatic growth during 2012 and 2013.
The estimated larval growth rates (0.09–0.21 mm day−1) were similar to those estimated for other nearshore benthic fishes from central Chile such as clingfishes Gobiesox marmoratus (0.24 mm day−1) and Sicyases sanguineus (0.14 mm day−1, Contreras et al. 2013) and triplefin blennies (0.14–0.16 mm day−1, Palacios-Fuentes et al. 2014). Hence, it has been accumulating evidences that slow larval growth seems to be a generalized process for littoral fishes in the south-eastern Pacific Ocean, an ecosystem that is characterized by low temperatures associated with the cold-water Humbold Current System (HCS) but also characterized by a high productivity due the occurrence of upwelling events. Hence, it is reasonable to hypothesize that small variations in any of growth controllers would trigger great variations in growth rates in these fishes, triggering great changes in survival as well, such as it has been suggested for a number of studies for teleost marine fishes (Shepherd and Cushing 1980; Chambers and Leggett 1987; Miller et al. 1988; Leggett and DeBlois 1994; Takasuka et al. 2003; Hawn et al. 2005; Plaza and Ishida 2008). In the case of Chilean sand stargazer (S. australis), the higher growth rates particularly in 2012 and 2013 matched with the higher abundances. A similar pattern was detected in demersal and pelagic fish species from the Barents Sea (cod, haddock and herring, Ottersen and Loeng 2000), in Japanese mackerel (Scomberemorus niphonius) larvae in the Seto Inland Sea, Japan (Shoji and Tanaka 2003), and the sablefish off Oregon affected by extreme El Niño and La Niña events (Sogard 2011).
Natural hypoxic events may affect population dynamics of nearshore fishes in coastal waters of Chile, as it has been detected in toadfish Aphos porosus (Hernández-Miranda et al. 2012a). An intense event of natural hypoxia in January 2008 affected the entire resident community of shallow Coliumo Bay, causing widespread mortality of organisms and a mass stranding event, with fish being one of the most affected groups (Hernández-Miranda et al. 2010, 2012b). These authors reported a recovery in richness of the fish assemblage in a timescale of only 3 months; nevertheless, densities reached only about half their previous level after 2 years of the event. At a longer timescale, negative effects on the total density of the fish assemblage, including A. porosus, were also detected (Hernández-Miranda et al. 2012a), similarly to the reduction in larval abundance of the sand stargazer observed from 2010 to 2012.
Therefore, such event may have modified the population parameters of the sand-dwelling fish S. australis in central Chile, reducing its larval supply (i.e. density). Nonetheless, the sand stargazer showed a relatively rapid recovery after 3 years, increasing twice its larval growth, its otolith size and changing the hatching patterns in the lunar cycle.
We thank the field work of J.E. Contreras, F. Salas-Berríos, M.J. Ochoa-Muñoz, P. Palacios-Fuentes, C. González-Erazo and Dr. Randy Finke. Two anonymous reviewers improved with their comments and suggestions an early version of the ms. This research was funded by project Fondecyt Grant No 1100424, adjudicated to FPO, GP and MFL.
- Batschelet E (1981) Circular statistics in biology. Academic Press, New YorkGoogle Scholar
- Brogan MW (1992) Ecology of larval fishes around reefs in the Gulf of California, Mexico. Ph.D. Dissertation, University of ArizonaGoogle Scholar
- Cavalluzzi MR (1997) Larvae of Gillellus jacksoni, G. uranidea (Dactyloscopidae), Stathmonotus stahli tekla, and S. hemphilli (Chaenopsidae), with comments on the use of the early life history characters for elucidating relationships within the Blennioidei. Bull Mar Sci 60:139–151Google Scholar
- Dawson CE (1977) Studies on eastern Pacific sand stargazer (Pisces: Dactyloscopidae) 4. Gillellus, Sindoscopus new genus, and Heteristius with description of new species. Proc Cal Acad Sci 61:125–160Google Scholar
- Dawson CE (1982) Atlantic sand stargazers (Pisces: Dactyloscopidae), with descriptions of one new genus and seven new species. Bull Mar Sci 32:14–85Google Scholar
- Fowler HW, Bean BA (1923) Descriptions of eighteen new species of fishes from the Wilkes Exploring Expedition, preserved in the United States National Museum. Proc U. S. Natl Mus 63:1–27Google Scholar
- Francis MP, Williams MW, Pryce AC, Pollard S, Scott SG (1993) Uncoupling the otolith and somatic growth in Pagrus auratus (Sparidae). Fish Bull 91:159–164Google Scholar
- Hastings PA, Petersen CW (2010) Parental care, oviposition sites, and mating systems of Blennioid fishes. In: Cole KS (ed) Reproduction and sexuality in marine fishes. University of California Press, California, pp 91–116Google Scholar
- Hastings PA, Springer VG (2009) Systematics of the Blennioidei and the included families Chaenopsidae, Clinidae, Labrisomidae and Dactyloscopidae. In: Patzner RA, Gonçalves EJ, Hastings PA, Kapoor BJ (eds) The biology of the blennies. Science Publishers, Enfield, NH, pp 3–30Google Scholar
- Hawn AT, Martin GB, Sandin SA, Hare JA (2005) Early juvenile mortality in the coral reef fish Chromis cyanea (Pomacentridae): the growth-mortality hypothesis revisited. Bull Mar Sci 77:309–318Google Scholar
- Hernández-Miranda E, Veas R, Labra FA, Salamanca M, Quiñones RA (2012b) Response of the epibenthic macrofaunal community to a strong upwelling-driven hypoxic event in a shallow bay of the southern Humboldt current system. Mar Environ Res 79:16–28. doi:10.1016/j.marenvres.2012.04.004 CrossRefPubMedGoogle Scholar
- Herrera GA, Llanos-Rivera A, Landaeta MF (2007) Larvae of the sand stargazer Sindoscopus australis and notes on the development of Dactyloscopidae. Zootaxa 1401:63–68Google Scholar
- Mardia KV (1972) A multi-sample uniform scores test on a circle and its parametric competitor. J R Stat Soc Ser B 34:102–113Google Scholar
- Munk P (1993) Differential growth of larval sprat Sprattus sprattus across a tidal front in the eastern North Sea. Mar Ecol Prog Ser 99:17–27Google Scholar
- Nelson JS (2006) Fishes of the world. Wiley, New JerseyGoogle Scholar
- Shoji J, Tanaka M (2003) Larval abundance, growth, and recruitment of Japanese Spanish mackerel Scomberemorus niphonius in the Seto Inland Sea, Japan. The big fish bang. In: Browman HI, Skiftesvik AB (eds) Proceedings of the 26th annual larval fish conference. Institute of Marine Research, Bergen, pp 395–404Google Scholar
- Watson W (1996) Blennioidei. In: Moser HG (ed) The early stages of fishes in the California current region. CalCOFI Atlas No. 33, pp. 1148–1199Google Scholar
- Zar JH (1999) Biostatistical analysis. Prentice Hall, New JerseyGoogle Scholar