Seasonal growth of Mugil liza Valenciennes, 1836 in a tropical estuarine system

Journal of

Applied Ichthyology J. Appl. Ichthyol. (2015), 1–6 © 2015 Blackwell Verlag GmbH ISSN 0175–8659

Received: July 24, 2014 Accepted: September 18, 2014 doi: 10.1111/jai.12704

Seasonal growth of Mugil liza Valenciennes, 1836 in a tropical estuarine system By M. F. Sousa, N. N. Fabr
e and V. S. Batista Instituto de Ci^ encias Biol ogicas e da Sa ude, Campus A. C. Sim~ oes, Universidade Federal de Alagoas, Macei o, AL, Brazil

Summary Seasonal changes in the abiotic factors and cyclical events – with the reproduction diagnosed by the gonadosomatic index and the energy status diagnosed by the condition index – were estimators of seasonal growth for juvenile and adult lebranche mullets, Mugil liza, during the rainy and dry seasons in a tropical estuarine system of the western south Atlantic. Fish were collected monthly by gillnets of different mesh sizes. Salinity, water temperature (°C) and dissolved oxygen (%) were measured using a multiparameter. Daily rainfall data (mm) were obtained from the National Institute of Meteorology. Seasonal rings were identified and counted in the sagittae otoliths. Relative Marginal Increment (RMI) was calculated to determine the seasonal growth rate. Gonadosomatic Index (GSI) and Condition Index (CI) were assessed separately for juveniles and adults by ANOVA, with time (month) M. liza growth is controlled by annual cycles of biological and abiotic processes. Using general linear models, the seasonal growth rate of M. liza was well predicted by body condition and rainfall for juveniles, and by salinity for adults. Seasonal variation was observed in the Munda
u Lagoon, with rainfall directly or indirectly being the main seasonal factor influencing the cyclical biological fish processes of M. liza. Introduction Fish growth is influenced mainly by local climate characteristics, availability of food resources, migratory dynamics, ~ez-Aguirre and Gallardo-Caand population density (Ib
an ~ez and bello, 1996; Fabr
e and Saint-Paul, 1998; Ib
an Guti
errez-Ben
ıtez, 2004; Whitfield et al., 2012). In temperate environments growth is strongly dependent on temperature, which causes distinct seasonal patterns of growth that are clearly visible as rings on hard body structures. In contrast, tropical environments have often been considered non-seasonal because the photoperiod and temperature are relatively stable throughout the year. Accordingly, early researchers assumed that fish growth in the tropics was relatively constant throughout the year (Panella, 1971). However, in recent decades this view has changed, as increasing numbers of fish otolith studies from different tropical environments have indicated cyclical growth patterns (e.g. Morales-Nin and Panfili, 2005; P
erez and Fabr
e, 2009).

In the absence of temperature and photoperiod effects, the rainfall cycle is a clear candidate to cause seasonal changes in the biology and ecology of fishes in tropical rivers and coastal systems. Precipitation levels influence the rhythm of seasonal migrations, affecting habitat use and, therefore, the dynamics of biological processes that can determine cyclical patterns of growth (Fabr
e and Saint-Paul, 1998; Cutrim and Batista, 2005; Panfili et al., 2006; P
erez and Fabr
e, 2009). This also is expected to be found in estuarine ecosystems where the drainage basin rainfall is the usual control, but here with the effect of tides (Blaber, 1997; Barletta et al., 2010). These environmental processes affect mainly salinity and nutrient availability, which are potential triggers of body growth and ring formation. Suggested by this concept, seasonal otolith rings in tropical environments have been related to factors affected by rainfall, such as photoperiodism, salinity food production, and reproductive cycles (Cardona, 2000; Marriott and Mapstone, 2006) becoming important tools in the evaluation of environment effects on estuarine and coastal fishes. Mullets (family Mugilidae) are fishes commonly caught by artisanal fishers (Blaber, 1997) who are typical users of tropical estuaries. Salinity has been indicated as an important determinant of mullet habitat selection and growth (Cardona, 2000). Responses may differ in adults and juveniles, as observed in the flathead grey mullet Mugil cephalus Linnaeus 1758, which has faster juvenile growth rates in oligohaline environments and faster adult growth rates in polyhaline waters (Cardona, 2006). Although widely distributed in the Atlantic Ocean from the Caribbean to northern Argentina (Menezes et al., 2010), the lebranche mullet Mugil liza Valenciennes 1836 was only accessed in subtropical and temperate ecosystems (e.g. Vieira and Scalabrini, 1991; Esper et al., 2001; Gonz
alez-Castro et al., 2009; Albieri and Ara
ujo, 2010; Garbin et al., 2014); no on-going assessments of the adaptive response to local environmental variability (spatial and temporal) were taking place in the tropics. The objective of this investigation was to investigate to what extent the growth rates are associated with cyclical biological events in lebranche mullet inhabiting a tropical estuary. In this sense, it is hypothesized that seasonal changes in the abiotic factors and cyclical events – reproduction diagnosed by the gonad-somatic index and energy status diagnosed by the condition index – to be estimators of seasonal growth for juvenile and adult mullets.

2

Materials and methods The study was carried out between July 2011 and June 2012 in the estuarine Munda
u Lagoon, Alagoas, Brazil (centroid at 9° 380 15″ S 35° 460 20″ W), in the western Atlantic (Fig. 1). Fish were collected monthly in 5 km2 sampling area in the lagoon and 0.2 km2 in the lagoon mouth by nylon monofilament gillnets (1600 m long and 3 m high) with different mesh sizes (40, 50 and 60 mm). Salinity, water temperature (°C) and dissolved oxygen (%) were measured using a Hanna HI 9828 multiparameter. Daily rainfall data (mm) was obtained from the National Institute of Meteorology – INMET. The rainy season was defined as May to August (rainfall 242.9  94.2 mm) and the dry season from September to April (100.7  63.9 mm). Of the 319 fish caught at about 30 fish/month (Table 1), the data collected was: total length (Lt) (cm), total weight (g) (Wobs), stomach weight (g) (Wstom), liver weight (g) (Wliv) gonad weight (g) (Wgon). Maturity stage was assigned using macroscopic gonadal examination. Stages were defined as: I – Immature, II; Developing virgin, III; Mature, IV; Postspawning and V, Spent or resting. Fishes were classified by gonadal maturation into juveniles (stages I and II) and adults (stages III, IV and V). Otoliths sagittae were extracted from each specimen, washed and stored for measurement. To observe seasonal rings (macro-increment), otoliths were first decalcified using alcohol (p.a.) for 72 h, and observed submerged in alcohol on a dark base under reflected light. An image of each otolith was digitally captured and stored. Seasonal rings were identified and counted by two independent observers. The distance (mm) between the nucleus and each ring (Rn) and the otolith edge (Rt) (Fig. 2) was measured using LEICA LAS EZ, v 2.1.0 software. Relative Marginal Increment (RMI) was calculated to determine the seasonal growth rate, RMI = (RtRn) (RnRn1)1, where Rt is the total radius of the otolith; Rn, the distance between the nucleus and last ring; and

M. F. Sousa et al.

Rn-1, the distance between the nucleus and next-to-last ring formed. This index and both the Gonadosomatic Index (GSI) and Condition Index (CI) were assessed separately for juveniles and adults by ANOVA (P = 0.05), with the month as a fixed factor. The monthly mean values of the indices (CI, GSI, RMI) were submitted to a Pearson correlation; the variables that correlated significantly (P < 0.05) with the RMI juveniles and adults were used to build the multiple regression model defined as: Y = a + b1X1 + b2X2+ . . . + bkXk. Where: Y is the dependent variable (RMI); a = intercept; b = regression coefficient that represent the independent contributions of each independent variable to the prediction of the dependent variable. Data was log-transformed where required after assumptions tests (Bartlett and Kolmogorov-Smirnov). Results Results indicate that GSI for juveniles remained low throughout the year (Fig. 3a), but was significantly higher (P < 0.05) for adults in August–September (Fig. 3b). CI for juveniles was low in the dry season of December and January (Fig. 3c), whereas adults had low CI in the reproductive rainy season – (Fig. 3d). RMI decreased sharply for juveniles from November to December–January (Fig. 3e); RMI in adults was lower from April to September, including a significant drop (P < 0.05) in August that coincided with the spawning period (Fig. 3f). Besides the CI effect, rainfall significantly explained the growth rates of juveniles (Table 2). In contrast, salinity significantly explained the growth rates for adults (Table 3). The condition index, which is an indicator of the energy status of the population, was then identified as a general predictor of biological cyclic processes. In the tropical Munda
u Lagoon both juvenile and adult M. liza were recorded in the rainy season. Adults in an

Fig. 1. Munda
u Lagoon, Northeastern Brazil. Sampled area indicated by an 2 u ellipse (5 km ) and in the Munda
Lagoon mouth (0.2 km2) from July 2011 to June 2012

Mugil liza growth in a tropical estuarine system

3

Table 1 Number of individuals collected (adults and young) in Munda
u Lagoon estuary, Northeastern Brazil, July 2011 to June 2012 Abundance (n)

Abiotic factors

Months

Adults

Young

Total

Temp

Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

14 23 20 15 18 26 14 12 7 9 8 10

14 6 8 45 43 9 10 11 17 15 22 16

28 29 28 60 61 35 24 23 24 24 30 26

29.9 28.4 27.4 25.2 28.8 29.7 25.7 27.9 29.6 25.6 28.4 26.9

           

DO(mg/L) 0.1 0.2 0.6 0.7 1.2 0.2 0.9 1.4 0.4 1.2 0.2 0.05

4.1 5.2 7.4 4.9 5.5 4.4 5.8 4.4 5.4 5.1 4.9 6.4

           

0.8 0.7 0.8 0.2 0.6 0.1 0.4 0.6 0.4 1.6 0.1 0.2

Salinity 5.6 0.3 2.6 12.4 7.0 12.3 10.8 10.0 10.4 13.8 7.0 7.5

           

Rainfall (mm) 1.0 0.9 1.8 6.9 6.1 1.7 0.3 0.8 0.5 0.3 5.4 1.0

366.2 209.1 111.2 48.2 112.3 14.2 189.9 67.8 172.2 56.6 85.3 285.4

Temp, temperature; DO, dissolved oxygen; Sal, salinity.

annual growth ring in December–January. Given that juveniles neither spawn nor migrate, the decreased growth must be related to other causes. The dissolved oxygen stability in the Munda
u Lagoon throughout the year (P = 0.233) makes it unlikely to affect M. liza growth. On the other hand, mean salinity is highly variable during the year (P < 0.001). The reproductive period occurs during the rainy season, when salinity is low (2.8  1.8). Discussion

Fig. 2. Sagitta otoliths of a Mugil liza with four rings showing distances measured to each ring (R1, R2, R3 and R4) and to the edge (Rt)

advanced stage of gonadal maturation are the only ones that migrate to the coastal zone. In the dry season, the highest proportion of juvenile M. liza was 4–5 months after spawning occurred in the Munda
u Lagoon. Juveniles form an

These results support the acceptance that the cyclical biological processes of fishes in tropical ecosystems are challenging because of the complex interactions between physiological (e.g. growth, reproductive cycles) and environmental factors (e.g. temporal patterns of habitat use, seasonal changes in abiotic habitat characteristics) (Blaber, 2002). Rainfall is particularly important as it increases the water volume and turbidity, light penetration and food availability, and reduces salinity during the wet season. These changes might favour the condition of fishes according to species and life history requirements. Heavy rains have been associated with increased run-off and better feeding conditions (Bergenius et al., 2005), but also with periods of intense energy expenditure (Bayley, 1988; Fabr
e and Saint-Paul, 1998). Seasonal variations were observed in the Munda
u Lagoon where rainfall directly or indirectly was the main seasonal factor influencing the M. liza cyclical biological fish processes. Quantitative seasonal indices related to biological cycles, such as body condition, reproduction and growth of juvenile lebranche mullet- were strongly associated with rainfall fluctuations. Mugilidae are often considered as estuarine dependent (Menezes and Figueiredo, 1985). However, considering the conceptual variation of ‘dependency’ (Able, 2005), it is preferable to focus on the functional significance of spatial and temporal patterns of growth and movement. The different grey mullet species have seasonal variations in abundance that appear to be related to their breeding seasons and

4

M. F. Sousa et al.

(b)

0

50

0

Jul

Jun

Apr

May

Mar

Months

Months

(c)

(d)

Young 1.4

400 Rainfall(R);

Mean(CI);

Adults 1.4

Mean ± 0.95 conf. interval (CI)

200 1.0 150 100 0.8

Condition index (CI)

250

Rainfall (mm)

300

1.2

400 Rainfall;

350

Mean(CI);

Mean ± 0.95 conf. interval (CI)

350 300

1.2

250 200 150

1.0

100 50 Jun

May

Apr

Mar

Feb

Jan

Dec

Nov

Oct

0.8

Months

Sep

0 Aug

Jun

May

Apr

Mar

Feb

Jan

Dec

Nov

Oct

Sep

Jul

Aug

50

Jul

Condition index (CI)

0

Rainfall (mm)

Jan

Feb

Dec

Oct

Nov

Sep

Jul

–0.05

Aug

50

100

Jun

0.00

150 2

Apr

100

May

150

200

4

Mar

0.05

250

Jan

200

300

Feb

0.10

350

6

Dec

250

Mean ± 0.95 conf. interval (GSI)

8

Oct

300

Mean (GSI);

Nov

0.15

400 Rainfall(R);

Sep

350

Gonadosomatic index (GSI)

Mean(GSI); Mean ± 0.95 conf. interval (GSI)

Rainfall (mm)

Gonadosomatic index (GSI)

Rainfall(R)

Adults 10

400

Rainfall (mm)

Young 0.20

Aug

(a)

0

Months

Months

Rainfall (mm)

100

0.2

50 0.0

Jun

Jun

Apr

May

Mar

Jan

Feb

Dec

Nov

Oct

0

Sep

50

0.0

Jul

0.2

150

May

100

Apr

150

0.4

200 0.4

Mar

0.6

250

0.6

Feb

200

300

Jan

250

0.8

350 0.8

Dec

1.0

300

Mean ± 0.95 conf. interval (RMI)

Nov

350

Mean(RMI);

Oct

1.2

400 Rainfall;

Sep

Mean ± 0.95 conf. interval (RMI)

Jul

Mean(RIM) ;

Rainfall (mm)

Rainfall;

Adults 1.0

Aug

400

Relative marginal increment (RMI)

(f)

Young

1.4

Aug

Relative marginal increment (RMI)

(e)

0

Months

Fig. 3. Variation by month of biological variables of young, left graphs (n = 143 individuals); adults, right graphs (n = 176 individuals) of M. liza, July 2011 to June 2012. Monthly mean  confidence intervals (a = 5%) and rainfall level (dotted lines) are also indicated in graphs of GSI (a) young and (b) adults, CI (c) young and (d) adults, RMI (e) young and (f) adults

spawning migrations (Wijeyaratne and Costa, 1987; Koutrakis, 2004). In the subtropical estuary of Lagoa dos Patos, M. liza juveniles are typically found in the lagoon during periods of high food availability, and in adjacent coastal waters after the migration of mature adults (Vieira and Scalabrini, 1991). The same pattern was found in the Ba
ıa de Paranagu
a subtropical estuary (Esper et al., 2001) and in the temperate Mar Chiquita Lagoon (Gonz
alez-Castro

et al., 2009). In both subtropical and temperate climate zones, M. liza is clearly a marine migrant species activated by cold temperatures and intense winds. During the reproductive period the adult M. liza showed reduced growth, forming an annual hyaline ring, as also observed in other mugilid species (Iba~ nez-Aguirre and Gallardo-Cabello, 2004; Espino-Barr et al., 2005; Gonz
alez-Castro et al., 2009; Santana et al., 2009). These seasonal growth patterns indicate

Mugil liza growth in a tropical estuarine system

5

Table 2 Multiple regression of dependent variable RMI for M. liza juveniles Variable

a

SE (a)

B

SE(B)

P

Intercept Gonadosomatic index Condition index Oxygen Salinity Rainfall

0.06 0.48 0.21 0.10 0.29

0.10 0.13 0.12 0.09 0.13

0.78 5.50 1.02 0.13 0.02 0.05

0.14 9.12 0.28 0.07 0.02 0.02